WO2023084447A1

WO2023084447A1 - Website content management, including generating recommendations for new content and website improvements

Info

Publication number: WO2023084447A1
Application number: PCT/IB2022/060843
Authority: WO
Inventors: Nicolai Munch Andersen; Angie LIM
Original assignee: Siteimprove A/S
Priority date: 2021-11-11
Filing date: 2022-11-10
Publication date: 2023-05-19

Abstract

Disclosed are systems and methods for providing content recommendations for a target website. The method can include receiving, at a computer system, a request for content recommendations for the target website, identifying user engagement data for third party websites that are related to the target website, determining content candidates that are currently popular or trending among users of the third party websites based on the user engagement data, accessing a performance model for the target website that models performance of content and webpage presentation details on the target website, determining projected performance scores and recommended presentation details for the content candidates by applying the performance model for the target website to the content candidates, selecting a portion of the of the content candidates based on the projected performance scores as content recommendations, and transmitting the content recommendations and corresponding recommended presentation details.

Description

WEBSITE CONTENT MANAGEMENT, INCLUDING GENERATING RECOMMENDATIONS FOR NEW CONTENT AND WEBSITE IMPROVEMENTS

TECHNICAL FIELD

[0001] This document generally describes devices, systems, and methods related to generating and implementing content recommendations for a website, such as through a content management system (CMS).

REFERENCE TO CO-PENDING APPLICATION

[0002] This application claims the benefit of priority under 35 U.S.C. § 119(e) to U.S. Provisional Patent Application No. 63/278,319, filed November 11, 2021 , and entitled WEBSITE CONTENT MANAGEMENT, INCLUDING GENERATING RECOMMENDATIONS FOR NEW CONTENT AND WEBSITE IMPROVEMENTS, and the entire disclosures set forth therein are incorporated herein by reference.

BACKGROUND

[0003] Websites can vary in their quality. The quality of a website can affect user experience of users who visit the website. For example, websites with broken links, misspellings, and other features that do not function as intended can be frustrating for users visiting a site. Additionally, websites that are not optimized for search engines (also referred to as “search engine optimization” or SEO) may have a low level of quality because users may not be able to locate relevant pages on the website (or locate the website more generally) using a search engine. For example, websites that do not have information formatted properly for search engines to retrieve and associate with other information on the page may have low levels of SEO, which may result in those websites not appearing as prominently in search results as they otherwise should for relevant search strings. Websites can also have a low level of quality if they are not readily accessible to all users regardless of impairment (also referred to as “website accessibility”). For example, if a website is not formatted properly, users who are seeing or hearing impaired may not be able to use website reader applications to review and navigate through the content. SUMMARY

[0004] The document generally describes technology for improving the content on a website through performance based ideation that leverages both the performance of content within the website, as well as the performance of content on other websites and on the internet more broadly. For example, the website content lifecycle (i.e., process by which content is initially created and improved on a website) can include various phases from initial ideation and creation of content, to the initial publication of the content, and then iterative and ongoing analysis, experimentation, and optimization of the content, such as through modifying content and assessing changes in performance (i.e., web traffic changes, ad conversions, SEO rank). The conventional content lifecycle can be lengthy and require multiple iterations, which may still lead to a sub-optimal results. The disclosed technology provides a platform through which high-performing content can be more quickly identified, created, and published on a website in a manner that optimizes performance, which can significantly shorten the content lifecycle and improve the overall website performance. For example, the disclosed technology, which can be provided in any of a variety ways (e.g., CMS system, plugin to CMS system, web content analysis system, API), offers actionable insights throughout this content lifecycle, including the creation of new content and improvements to existing content. [0005] The disclosed technology can include a variety of features to improve content creation, performance, and management, including generating and providing content recommendations that are specifically tailored to perform well on a specific website. For example, the audience for a particular website may be drastically different from the audience for another website. As a result, the same content (e.g., topics) and the same manner for presenting that content on a webpage (e.g., content configuration, page layout) may perform well on one website, but poorly on another website. The disclosed technology can identify and recommend content that will perform well on a particular website, as well as more presentation details for how to best present that content on a webpage so as to optimize the performance of the content. By doing this, the disclosed technology to minimize and/or reduce the content generation lifecycle by starting off with content recommendations that need little-to-no iterating and experimenting to perform well on a website. Additionally, the disclosed technology can assist with improving existing content on a website by identifying changes to content and the manner in which it is presented on the website so that it will perform better on the website.

[0006] As discussed throughout this document, website and web content performance can vary from website to website, and can be based on any of a variety of factors. For example, content on some websites may be considered to be performing well when it has a high number of page views and dwell time, yet with other websites performance may be measured at a rate of website visitors performing a conversion action on the website (e.g., purchase, account creation, sign-up for contact list). The disclosed technology can permit for website owners and managers to designate various criteria to assess website and web content performance, including designating different performance criteria for different parts of a website (e.g., different performance criteria for different groups of pages, different content types).

[0007] The disclosed technology can use multiple different signals to identify new and/or improved content for a specific website, including signals identifying content that has performed well on the specific website as well as signals indicating new content that is likely to be of interest to and perform well with the website’s audience. For example, the disclosed technology can combine internal signals that are generated from the existing content on a website and external signals that are generated from other websites. The internal signals can be generated, for example, by correlating the content and structural features on webpages with the performance of those pages to model one or more combinations of features that perform well. The external signals can be generated, for example, by identifying new content that is likely to perform well on the website based on content that is currently of interest on related third party websites (e.g., trending and/or popular content). Such external signals can be generated from, for example, external data indicating topics that are currently of interest to users, such as SEO data that identifies search queries and webpage rankings for various sectors and/or webpages, ads data that identifies user engagement with different ad content, and/or other external data. [0008] Although the disclosed inventive concepts include those defined in the attached claims, it should be understood that the inventive concepts can also be defined in accordance with the following embodiments.

[0009] Embodiment 1 is a method for providing content recommendations for a target website, the method comprising: receiving, at a computer system, a request for content recommendations for the target website; identifying, by the computer system, user engagement data for a plurality of third party websites that are related to the target website; determining, by the computer system, a plurality of content candidates that are currently popular or trending among users of the plurality of third party websites based on the user engagement data; accessing, by the computer system, a performance model for the target website that models performance of content and webpage presentation details on the target website; determining, by the computer system, projected performance scores and recommended presentation details for the plurality of content candidates by applying the performance model for the target website to the content candidates; selecting, by the computer system, a portion of the of the content candidates based on the projected performance scores as content recommendations; and transmitting, by the computer system, the content recommendations and corresponding recommended presentation details.

[0010] Embodiment 2 is the method of embodiment 1 , wherein the content request is for new content to be added to the target website, and the content recommendations include content candidates that are different from content presently included on the target website, that are presently popular or trending among the users of the third party websites, and that are projected to have at least a threshold performance level on the target website.

[0011] Embodiment 3 is the method of any one of embodiments 1 through 2, wherein the content request includes user input specifying subject matter for the new content to be added to the target website, and determination of the content candidates is further based on the subject matter specified by the user input, the content candidates being determined based, at least in part, on being related or similar to the subject matter specified by the user input.

[0012] Embodiment 4 is the method of any one of embodiments 1 through 3, the content request is for identification of improvements to preexisting content on the target website; and the content recommendations include content candidates that are different from content presently included on the target website, that are presently popular or trending among the users of the third party websites, and that are projected to have at least a threshold performance level on the target website.

[0013] Embodiment 5 is the method of any one of embodiments 1 through 4, wherein: the content request includes identification of a specific webpage on the target website for improvement, and determination of the content candidates is further based on existing content included on the specific webpage, the content candidates being determined based, at least in part, on being related or similar to the existing content included on the specific webpage.

[0014] Embodiment 6 is the method of any one of embodiments 1 through 5, wherein the plurality of third party websites are determined to be related to the target website based on the plurality of third party websites being competitors of the target website.

[0015] Embodiment 7 is the method of any one of embodiments 1 through 6, wherein the plurality of third party websites are determined to be related to the target website based on the plurality of third party websites appearing together with the target website in search engine results to one or more search queries.

[0016] Embodiment 8 is the method of any one of embodiments 1 through 7, wherein: determining the content candidates further includes determining popularity scores for each of the content candidates, and selecting the portion of the content candidates is further based on the popularity scores for each of the content candidates.

[0017] Embodiment 9 is the method of any one of embodiments 1 through 8, wherein the popularity scores are determined using a popularity model that correlates the user engagement data for the third party websites with a performance of content on the target website.

[0018] Embodiment 10 is the method of any one of embodiments 1 through 9, further comprising: training the performance model for the target website based on performance data for the target website and content for the webpages on the target website. [0019] Embodiment 11 is the method of any one of embodiments 1 through 10, wherein the training comprises: identifying content features and presentation features for the content on the webpages; determining performance scores for the webpages based on the performance data; and training the performance model using the content features, the presentation features, and the performance scores for the webpages.

[0020] Embodiment 12 is the method of any one of embodiments 1 through 11 , wherein the training further comprises: identifying previously implemented content recommendations on one or more updated webpages on the target website; accessing previously projected performance scores for the previously implemented content recommendations; determining current performance scores for the updated webpages based on the performance data; identifying content features and presentation features for the updated webpages; and training the performance model using the previously implemented content recommendations, the previously projected performance scores, the current performance scores, and the content features and presentation features for the updated webpages.

[0021] Embodiment 13 is the method of any one of embodiments 1 through 12, wherein the content recommendations are configured to be presented in a user interface with a selectable implementation option that, when selected, populates a webs content editing interface with content elements and presentation elements based on the content recommendations and corresponding presentation details. [0022] Embodiment 14 is the method of any one of embodiments 1 through 13, wherein the user interface comprises a content management system (CMS) user interface.

[0023] Embodiment 15 is the method of any one of embodiments 1 through 14, wherein the computer system is part of a CMS platform that is used for editing and managing the target website.

[0024] Embodiment 16 is the method of any one of embodiments 1 through 15, wherein the computer system provides an application programming interface (API) that is used by third party website management systems to provide content recommendations. [0025] Embodiment 17 is the method of any one of embodiments 1 through 16, wherein at least a portion of the third party website management systems include a CMS platform.

[0026] The devices, system, and techniques described herein may provide one or more of the following advantages. For example, the lifecycle for creating high- performing content on a website can be significantly reduced by the disclosed technology, which can identify content and manners for presenting the content that will result in high performance on the website. In another example, the disclosed technology can assist with more accurately and efficiently identifying new content that may of interest to a website’s audience, and for distinguishing that content from other globally popular content that may not be of interest to the website’s audience. For instance, previously a website owner/manager would need to experiment with different new content by iteratively publishing, reviewing results, and then adjusting the content to improve performance. This experimentation can take significant time and may not necessarily suggest how to actually improve performance when new content is underperforming. The disclosed technology resolves those issues by identifying and selecting specific new content that is popular/trending more broadly outside of the website, yet will be relevant to and well received by the website’s audience.

[0027] In another example, the disclosed technology can provide specific details and recommendations for not only the content that should be added to a website, but also the manner in which that content is presented on the website (e.g., format, layout). As a result, the disclosed technology can significantly improve the certainty with which a website owner/manager can develop and implement new, high- performing content.

[0028] In a further example, the disclosed technology can automatically and dynamically perform these new content determinations and recommendations, which can permit for website owners/managers to maintain high performance of the website even as the website’s audience interests and tastes may change over time. [0029] In another example, the disclosed technology can assist website owners/managers with improving the performance of existing website content, as well. For example, the disclosed technology can be used to identify pages for which the performance can be improved through updates to the content and/or its presentation on the website.

[0030] In a further example, the disclosed technology can additionally be extended to the performance of a website and its content on other systems, such as search engines performance, ads performance, online retail performance, and/or social media performance. For example, the disclose technology can be applied to generating and improving high-performing content on other systems, such as generating improved ad content for use on ads platforms and/or generating improved social media posts for use on social media platforms.

[0031] The details of one or more implementations are set forth in the accompanying drawings and the description below. Other features and advantages will be apparent from the description and drawings, and from the claims.

BRIEF DESCRIPTION OF THE DRAWINGS

[0032] FIGS. 1A-B are conceptual diagrams of example systems for performance-based ideation of new and improved website content.

[0033] FIG. 2 is an example system for generating, providing, and implementing performance-based recommendations for new and improved website content.

[0034] FIG. 3 is a flowchart of an example process for generating, providing, and implementing performance-based recommendations for new and improved website content for a target website.

[0035] FIG. 4 is a flowchart of an example process for identifying high performing webpages on a target website.

[0036] FIG. 5 is a flowchart of an example process for generating a performance model for a target website.

[0037] FIG. 6 is a flowchart of an example process for identifying trending and/or popular content on third party websites that is relevant to a target website.

[0038] FIG. 7 is a flowchart of an example process for identifying candidates for new and/or improved content on a target website.

[0039] FIGS. 8A-B are flowcharts of example processes for generating content recommendations for a target website. [0040] FIGS. 9A-C are screenshots of an example CMS user interface and process flow for presenting and implementing new content recommendations related to a target website.

[0041] FIGS. 10A-C are screenshots of an example CMS user interface and process flow for presenting and implementing content recommendations for improving existing website content on a target website.

[0042] FIGS. 11 A-C are screenshots of an example CMS user interface and process flow for presenting and implementing new content recommendations based on user input for a target website.

[0043] FIG. 12 is a schematic diagram that shows an example of a computing device and a mobile computing device.

[0044] Like reference symbols in the various drawings indicate like elements.

DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS

[0045] This document generally describes technology for generating new, high- performing content for websites, and improving the performance of existing content for websites. The disclosed technology, which can be provided in any of a variety of implementations (e.g., API, integration with CMS, integration with web analysis systems), can permit for users to easily implement data, insights, and recommendations for new and improved content into the content creation processes, for example, within a CMS system. For example, by analyzing relevant data related to the performance of content on a website, such as SEO, Ads, engagement, and experience data, content insights and recommendations can be generated and provided to website owners/managers. These recommendations can permit for a performance-based content ideation and creation/improvement of web content, resulting in improved website performance and increased business impact.

[0046] The disclosed innovation can generate more concrete recommendations and insights on new and improved web content, which can assist in identifying content gaps and opportunities. This can help website owners/managers create high-ranking and high performing content reaching the right audience and drive engagement and business outcome. Recommendations and insights can include, for example, planning new content (e.g., what to write, what to edit), creating new content (e.g., what to write), and/or editing existing content (e.g., optimizing existing webpages). Additional valuable cross discipline insights and recommendations to content ideation/creation can also be provided, such as for analytics insights (e.g., search terms, traffic and conversion insights, visitor profile and behavior), ads insights (e.g., validation of keyword performance through paid search, ads landing pages, ads performance metrics (i.e. , clicks, CTR, conversions etc.)), experience insights (e.g., insights into the provided website experience highlighting the elements that hinders the customer experience and the overall performance, such as readability, misspellings, broken links, accessibility issues, load speed, custom policies), and/or other improvements.

[0047] FIGS. 1A-B are conceptual diagrams of example systems 100 and 150 for performance-based ideation of new and improved website content. The example systems 100 and 150 can be used to generate and implement content recommendations on a target website that is being improved. The systems 100 and 150 can perform these same operations for multiple different websites, however. The phrase “target website” is simply used in this document to refer to a specific website to which operations pertain to illustrate the disclosed innovation.

[0048] Referring to FIG. 1A, the system 100 includes a web content performance system 102 (e.g., server system, cloud computing system) which can receive target website information 108 (step A) for a target website 104 and third party website information 110 (step B) for a plurality of third party websites 106 to determine and provide content recommendations for the target website 104. The target website 104 can be hosted by a server system (not depicted) and can include multiple different webpages P1-Pn. The content recommendations that are generated by the web content performance system 102 can be specific to the target website 104. The third party websites 106 can include other websites S1-Sn, which can include competitor websites (e.g., websites providing similar and/or related content, goods, or services as the target website 104) and non-competitor websites (e.g., websites providing dissimilar and/or unrelated content, goods, or services as the target website 104).

[0049] The target website information 108 can include any of a variety of details regarding the content on the target website 104 and its performance, such as the webpages P1-Pn themselves, web content quality data (e.g., SITEIMPROVE DCI score, accessibility scores), engagement data (e.g., page views, dwell time, SEO data, ads data, conversion data), and/or other data. The third party website information 110 may be more limited, and may include only information that is publicly accessible for the third party websites 106 (e.g., web content retrievable from the third party websites) and/or provided by a third party service analyzing publicly accessible information about the third party websites 106 (e.g., SEO information related to the third party websites 106, such as search rankings, terms; ads data related to keyword bidding).

[0050] The web content performance system 102 can use the target website information 108 and the third party website information 110 to generate one or more performance models for the target website 104, as indicated by step C (112). As discussed in greater detail below, a performance model for the target website 104 can model the performance of different types of content, as well as the different ways that it is presented, including the layout and configuration of the content (e.g., format, organization of content elements, page structure, menus), textual details (e.g., textual complexity, sentence length, word length, paragraph length, headings), image details (e.g., ratio of images to text, image size, image details), website structure (e.g., first level webpage, second level webpage), and/or other presentation details. The performance model can be trained using, at least, existing webpage content and performance information for that content from the target website information 108, and can be trained using any of a variety of machine learning techniques, such as neural networks, clustering and classification, regression analysis, natural language processing, and/or other machine learning techniques. The performance model can effectively model the behavior of the audience of users visiting the target website 104 - indicating preferences for particular types of content, manners of presenting that content, and/or other insights.

[0051] The web content performance system 102 can use the performance model to identify high performance candidates for the target website, as shown in step D (114). For example, the webs content performance system 102 can use the third party website information 110 to identify content that is currently popular and/or trending on third party websites 106, and can feed that information into the performance model to identify new content that is likely to perform well on the target website 104, along with recommendations for how to present that information on the target website 104 (e.g., format, layout, text details, image details, website structure). The performance model can additionally be used to identify improvements to existing content on the target website 104, such as changes to the content and/or manner in which is presented. For example, the performance model can provide a projected performance score, which is an indicator of the likely performance of corresponding content on the target website 104. A performance score for an existing page on the target website 104 can be determined and compared against a projected performance score for the content of the existing page, were it to be updated. At least a threshold difference between the actual and projected performance scores of an existing webpage can identify an opportunity to improve the performance of that page, which the output of the performance model can additional help guide through content and presentation insights.

[0052] New and improved content candidates that are processed through the performance model can be assessed based on their projected performance, and candidates that have at least a threshold projected performance (e.g., performance score above threshold level, ranked in top X% of candidates) can be provided by the web content performance system 102 as recommendations, as indicated by step E (116). The recommendations be provided and presented via any of a variety of pathways, such as being served directly to requesting client devices, provided as part of an API offered by the web content performance system 102, and/or integrated into one or more website management systems (e.g., CMS platforms). A presentation in an example CMS user interface 118 is depicted, which includes a section 120 providing new content recommendations and another section 122 for improvements to existing pages on the target website 104. As shown in the example user interface 118, the recommendations can identify the content (e.g., “dogs”) as well as the manner for presenting that content (e.g., layout, text to image ratio, and level similar to page P1) (122, 130), and can also identify a basis for the recommendation (e.g., “dogs” is trending topic, page P1 is high performing) (124, 132). The user interface 118 also includes features through which the user can act upon and implement the recommendations, as indicated by the selected elements 126 and 134, which can cause a user interface to be presented with at least a portion of the recommended content automatically generated along with guidance for the user to complete the remaining portion of the content in a manner consistent with the recommendation.

[0053] The recommended new/improved content can be implemented (published) on the target website 104, as indicated by step F (136), which can create a performance-based ideation loop to continually improve the performance of content on the target website 104 via the web content performance system 102. For example, new content can be recommended and implemented on the target website 104, and then the performance of that recommended and implemented content can be used to further refine and improve the performance model for the target website 104. Additionally, the performance loop can additionally be used to ensure that web content on the target website 104 is continually performing at a high level, including as user interests in content and manners of presenting the content change and evolve over time.

[0054] Referring to FIG. 1B, the example system 150 (similar to system 100) depicts a conceptual overview of training and using a performance model to generate content recommendations for the target website 104. As depicted, the target website 104 can have an audience of users 152 who visit and use the target website 104, such as through requesting webpages for the target website 104 from their client devices, uploading content to the target website 104, and/or purchasing products/services using the target website 104. The activities of the users 152 on the target website 104 can be tracked using various third party engagement tracking tools and data, such as GOOGLE ANALYTICS, third party SEO trackers, online ads platforms, and others. In the depicted example, some of the pages 154 on the target website 104 have high levels of engagement 158 — meaning high performance of the pages 154. High levels of engagement on the pages 154 can include, for example, higher page views, longer dwell times, and/or a higher ratio of conversions (e.g., create user account, purchase item). In this example, the target website 104 also includes some other pages 156 that have lower levels of user engagement 160 and, as a result, are lower performing. [0055] The web content performance system 102 can assess the performance level of the pages of the website 104, such as through generating performance scores for the pages, and can use that data in combination with the content of the webpages to train the target website performance model 170. For example, multiple different signals of performance can be ingested by the web content performance system 102, such as user behavior data (e.g., page views, dwell time, page navigation), webpage content quality data, SEO data, ads data, social media data, and/or other data, and applied against criteria specific to the target website 104 to determine the performance score for each page. Data 164 and 166 for both high and low performing pages (as determined based on the performance score), as well as other pages between low and high performing, can be used by the web content performance system 102 to train the performance model 170. The data 164 and 166 can include the page content itself, as well as the performance signals identified above. Training the model 170 using the high and low performing pages 154, 156 can permit for a more complete and robust model 170, which can differentiate more precisely between features that are well received by the users 152 and, as a result, perform well, and other features that do not. Other pages on the website 104 (outside of just high and low performing pages) can additionally and/or alternatively be used to train the model 170, including data for some or all pages regardless of performance score. The separation into high and low performing pages is simply provided here as an illustrative example.

[0056] The users 152 can additionally visit and use third party websites 106, which can include engagement 162 that is separate from the target website 104. The engagement data for the third party websites 106 that is accessible for the purposes of the target website 104 may be limited, however, as noted above. The engagement 162 can include information indicating content that is currently of interest to the users 152, who also visit the target website 104, such as search queries, ad conversions, and other details regarding engagement with the third party websites 106 that may be available for the target website 104. This data can be used to tease out popular and/or trending content 168, which the web content performance system 102 can process through the target website performance model 170 to identify new or improved content candidates for the target website 104 that are likely to be high performing on the target website 104 with regard to the users 152. For example, assume the target website 104 is a travel website that includes guides to different travel destinations, links to different travel services, and travel stories. A new travel destination that is not included on the target website 104 becomes popular. Without the web content performance system 102, the target website 104 will be unable to significant capture web traffic for this new destination, and would have to rely on manual monitoring of their website content relative to current trends to catch-up. However, with the web content performance system 102, this new destination can be automatically and promptly identified via activity with regard to third party websites 106 (e.g., search queries related to the new destination), and recommended as new content in a manner that will perform well on the target website.

[0057] The high performance content recommendations generated by the web content performance system 102 using the model 170 can be provided and implemented as part of the target website 104, as indicated by 172. As noted above with regard to FIG. 1A, this can provide a performance feedback loop that is used to continually improve the performance of content on the target website 104, and to train and update the performance model 170 for the target website 104.

[0058] FIG. 2 is an example system 200 for generating, providing, and implementing performance-based recommendations for new and improved website content. The example system 200 can be used to implement the processes, systems, and devices described throughout this document. The system 200 can perform the features described above with regard to FIGS. 1A-B, such as training performance models, generating new/improved content recommendation, and implementing those content recommendations on a target website.

[0059] The system 200 includes a web content performance system 202, similar to the web content performance system 102 described above with regard to FIGS. 1A-B. The system 200 also includes one or more web hosting systems 204 that can host the target website 214 and the third party websites 216, which can include competitor websites 218 of the target website 214. The system 200 includes client devices 206 (e.g., devices used by the users 152) that can request and obtain website content for the websites 214-218 from the hosting systems 204. The client devices 206 can additionally interface with various systems that feed into and provide website performance data sources 208, such as online ads platforms 220, web analytics systems 222 (e.g., GOOGLE ANALYTICS), search engine systems 224, social media platforms 226, and online retail platforms 228. These systems 220-228 can provide services that are embedded in or adjacent to the websites 214- 218, such as search and ads services that provide links to the websites 214-218. As a result, the client devices 206 can provide data to the systems 220-228 that are indicative of user engagement and/or performance of the websites, such as user ads behavior 242 (e.g., click through rates), user website behavior (e.g., dwell time, page views), user search behavior (e.g., search queries, selected results), user social media behavior (e.g., social media posts), and user retail behavior (e.g., product and services purchases).

[0060] The website performance data sources 208, which can include additional and/or alternative data sources to the examples mentioned here, can provide the web content performance system 202 with target website data 252 pertaining to the target website 214 and third party website data 254 pertaining to the third party websites 216, including the competitor websites 218. The web content performance system 202 can include a performance modeling system 232 that is configured to generate one or more performance models for the target website 214 using, at least, the target website data 252 and website content 256 for the target website 214. The performance modeling system 232 may additionally include other information to generate the one or more performance models, including using the third party website data 254, website content 256 for the third party websites 216, and/or web content analysis for the target website 214 and/or third party websites 216 as generated by the web content analysis system 230, which can generate assessments of website quality and can identify issues for correction, such as broken links, misspellings, and accessibility issues.

[0061] The web content performance system 202 can additionally include a prospective content identification system 234 that is configured to identify content candidates for new and/or improved content on the target website 214 using the models generated by the performance modeling system 232 and/or the third party website data 254, which can assist in surfacing trending and/or popular content for the target website 214. The web content performance system 202 can further include a content recommendation system 236 to select particular content candidates to recommend on the target website (e.g., high probability of high performance on the target website) and can generate content recommendations, including both the content and manner of presentation being recommended. The web content performance system 202 can further include a recommendation performance tracking system 238 to track the performance of previously recommended and implemented recommendations, which can be used to further train and improve the performance model generated by the performance modeling system 232 for the target website 214.

[0062] The web content performance system 202 can provide performance information and content recommendations 258 to, in the depicted example, an example CMS 210, which can incorporate the content recommendations into web content creation and editing features 260 that are provided on website owner client devices 212. For example, the web content performance system 202 can provide an API that is called by the CMS 210 related to the target website 214, and which can provide the content recommendations to the CMS 210. Additional and/or alternative configurations are also possible, including the system 202 being part of the CMS 210, the system 202 interfacing directly with the client devices 212, and/or the system 202 interacting with and/or being part of another system different from a CMS that is used by the devices 212 to manage the content on the target website 214, such as the website hosting systems 204. Based on interactions between the CMS 210 and the devices 212, the recommended content can be implemented and published 262 on the website hosting systems 204.

[0063] FIG. 3 is a flowchart of an example process 300 for generating, providing, and implementing performance-based recommendations for new and improved website content for a target website. The example process 300 can be performed on any of a variety of systems, such as the web content performance systems 102 and/or 202, as described above with regard to FIGS. 1A-B and 2.

[0064] The performance of existing pages on the target website can be determined (302). An example process for performing those determinations is described in greater detail below with regard to FIG. 4. One or more performance models for the target website can be generated, modeling at least high performing features on the target website, based on the performance assessment for at least a portion of the pages on the target website (304). An example process for performing the model generation is described in greater detail below with regard to FIG. 5. [0065] Trending and/or popular content that may be relevant to the target website can be identified based on data for third party websites (306). An example process for performing the model generation is described in greater detail below with regard to FIG. 6. Content candidates for new content and/or content improvements on the target website can be identified based on the trending and/or popular content (308). An example process for performing the model generation is described in greater detail below with regard to FIG. 6.

[0066] In some instances, an initial focus for the content recommendations can be identified based on, for example, user input (e.g., keyword) and/or selection of one or more existing web pages on the target website, such as selection of pages with low performance seeking improvement (310). For example, a user may want to draft a new page related to “dogs,” and can provide the keyword “dogs” (or other related keywords) as input that is used to subsequently identify content and a manner of presenting the content to achieve high performance on the target website. In another example, a user may be able to view a list of low performing pages on the target website (i.e. , ranked by performance score), and may select one or more of those pages for improvement. In another example, the system may automatically select pages with performance scores below a threshold level for improvement. The content on the pages selected for improvement can be used to provide the initial focus that is used to identify content recommendations. An example of user input to effectively “seed” the content recommendation is described below with regard to FIGS. 11A-C, and an example of recommendations being provided for a specific existing page on the target website are described below with regard to FIGS. 8B and 10A-C.

[0067] Content recommendations for the target website can be generated based on the performance model, the content candidates for new and/or improved content, and/or the initial focus for the content recommendations (312). An example process for generating content recommendations for new and/or improved content is described below with regard to FIG. 8A, and example user interfaces for providing the same are described below with regard to FIGS. 9A-C and 10A-C. An example process for generating recommendations new and/or improved content based on an initial focus for the content, such as through user input and/or selection of an existing webpage, is described below with regard to FIG. 8B, and example user interfaces for providing the same are described below with regard to FIGS. 11A-C.

[0068] Content recommendations for the target website can be transmitted and presented in a user interface for review and implementation by a website owner and/or manager (314). For example, as described above with regard to FIGS. 1A-B and 2, the content recommendations can be integrated into and/or provided as part of a CMS platform, which can present the content recommendations to a website owner and/or manager, who can select features to implement the recommendations. [0069] User selection of the recommendations can be received, and new and/or improved content can be initialized into a user interface and guidance can be provided to the user to complete the content according to the selected content recommendation (316). For example, a user may be presented with multiple different content recommendations for new content to be added to the target website. Each of the recommendations can include information identifying the type of content to be generated, along with the manner in which the content should be presented/formatted on the website in order to optimize the content’s performance. Selection of a recommendation can cause an interface for creating the content to be pre-populated with portions of the recommended content formatted/configured according to the recommendation, along with placeholders and other guidance for the user to flesh out the remainder of the content.

[0070] Once published on the target website (implemented), the performance of the recommended content can be tracked and that performance can further be incorporated into the performance model for the target website (318). For example, the performance model can be continually updated and improved based on previous content recommendations that, at the time of the recommendation, were projected to achieve a projected level of performance. These recommendations, their projected performance, and the actual performance of the content once implemented on the target website can be used to further train and improve the performance model. [0071] FIG. 4 is a flowchart of an example process 400 for identifying high performing webpages on a target website. The example process 400 can be performed on any of a variety of systems, such as the web content performance systems 102 and/or 202, as described above with regard to FIGS. 1A-B and 2. The process 400 can be performed as part of the process 300, for example, at step 302. [0072] Data for the webpages on the target website can be accessed (402), along with performance criteria for the target website (404). The data for the webpages can include data indicating engagement of users with the webpages, such as the data 242-252 described above with regard to FIG. 2. The performance criteria can be criteria that is tailored for the target website, such as criteria that is designated by the website owner and/or configured for the type of target website. The criteria can define which data related to user engagement are indicators of high performance of a webpage on the target website. The criteria can be different across different websites. For example, retail-based websites may focus more on whether content results in products sales, whereas other websites providing information to users may focus instead on page views, shares, and dwell time on the web pages.

[0073] Performance scores for each of the webpages on the target website can be determined based on the data and the performance criteria (406). For example, the criteria can be applied to the data for the webpages to generate performance scores for each of the webpages, which can be numeric values along a range (e.g., 0.0 - 1.0, 0-100), enumerated values indicating broader groupings of performance (e.g., low performance group, moderated performance group, high performance group), and/or other values.

[0074] Although the performance scores can be based on data for the webpages themselves, they can optionally be additionally determined based on the performance of competitor webpages (408). For example, the performance of a webpage on the target website can be assessed within the context of the other webpages on the target website, which can be helpful and indicative of which pages are performing well and which ones are not, but it may fail to take into consideration how performance scores stack up when compared against the performance of peer/competitor pages. For instance, one page on the target website may perform poorly relative to other pages on the target website. However, when compared against performance scores for competitor webpages with similar/analogous content, the target webpage may actually be determined to be performing well (i.e., the content on the target webpage and the competitor pages may simply be content that is of less interest to users relative to other content on the site). Accordingly, in some instances, the performance scores for the webpages on the target website can be augmented based on their relative performance against comparable competitor webpages.

[0075] As part of this optional process 408, the competitor webpages can be identified (410), data for the competitor webpages can be accessed (412), and performance scores for the competitor webpages can be determined (412). The competitor webpages can be predetermined by the website owner/manager, determined based on third party industry groupings of webpages, and/or dynamically based on webpages appearing in similar data appearances between the websites, such as appearing in the same results for search queries. As discussed above with regard to FIGS. 1A-B and 2, the data available for competitor webpages may be limited, but in some instances may be more broadly accessible. The performance scores for the competitor webpages can be determined similar to the performance scores for the target website (step 406), and can use the same criteria as the target website. The performance scores between the target webpages and the competitor webpages can be compared (416), and modifications to the target webpage performance scores can be modified based on the comparisons (418). For instance, in a simple example, if the performance of the target webpage is greater than the performance of the competitor webpages, the performance score of the target webpage may be increased, and if the performance of the target webpage is less than the performance of competitor webpages, then the performance score for the target webpage may be decreased.

[0076] The performance scores for the webpages on the target website can be compared (e.g., compared against each other, compared against threshold performance score values) (420), which can permit for groups of pages to be identified and designated, such as high performing pages (422) and low performing pages (424). Additional and/or alternative groupings are also possible. [0077] FIG. 5 is a flowchart of an example process 500 for generating a performance model for a target website. The example process 500 can be performed on any of a variety of systems, such as the web content performance systems 102 and/or 202, as described above with regard to FIGS. 1A-B and 2. The process 500 can be performed as part of the process 300, for example, at step 304. [0078] In the depicted example process 500, high and low performing webpages can be retrieved from the web server hosting the pages (502), and data for the pages along with their performance scores (e.g., as determined using process 400) can be accessed (504). Although only high and low performing webpages are mentioned, additional and/or alternative groupings, as well as all webpages on the target website may be included in the process 500. Features of the webpages can be identified, such as content features (506), layouts and configurations (508), structural context of the pages within the website (510), and/or external and/or temporal features (512).

[0079] The content features (506) can include, for example, the substantive content of the webpages, such as topics, words, phrases, images, titles, subject headings, videos, audio, and/or other content that is present on the webpage. The content features can additionally include assessments and relationships regarding the content, such as complexity of the text on the webpage, the ratio of text to images, a number of different content elements on the page (e.g., number of words, number of images), a quality of images (e.g., resolution of images), and/or other values. The layouts and configurations (508) can include the visual layout, design, and organization of the webpage, including the presence of header areas, menus, and/or other structural aspects of the manner in which the webpage is presented to a user. The structural context (510) can include the positioning of the webpages within the link, navigation, and page structure of the webpage, such as first level webpages, second level webpages, etc. The external and/or temporal features (512) can include information identifying an external presence of the webpages outside of the website, including links to the webpage on third party sites, promotions of the webpage in online ads, social medial promotion and engagement with regard to the webpage, and/or other external identification of the webpage that can influence the performance of the page. The external and/or temporal features can additionally timing for the various external activities, which can potentially be correlated against changes in website performance over time to identify which and what external features impacted website performance (and which ones did not).

[0080] Performance information for previous implemented content recommendations can be accessed (514), as discussed above with regard to FIG. 3. [0081] One or more performance models can be trained (516), for example, using the data for the webpages (504), the features identified for the webpages (506-512), and/or based on the performance or previous recommendations. As discussed above, any of a variety of techniques can be used to train a performance model, such as neural networks, clustering, classifying, regression analysis, and/or others. The model can be trained by correlating website features (506-512) to the performance scores determined with process 400, which can then be used to project the performance of various content items and to identify an optimal manner in which to present the content items to maximize performance on the target website. Once trained, the performance model can be output (518).

[0082] FIG. 6 is a flowchart of an example process 600 for identifying trending and/or popular content on third party websites that is relevant to a target website. The example process 600 can be performed on any of a variety of systems, such as the web content performance systems 102 and/or 202, as described above with regard to FIGS. 1A-B and 2. The process 600 can be performed as part of the process 300, for example, at step 306.

[0083] Competitor websites can be identified (602). For example, as discussed above with regard to FIG. 4, competitor websites may be predesignated/identified by the website owner/manager, and/or may be provided by one or more third parties based on industry groupings of websites. Additionally and/or alternatively, other third party websites that may have overlapping users and/or content with the target website can be identified (604). Identifying such overlapping websites can include identifying third party websites that appear in search results with the target website (606), identifying third party websites that are linked to the target website (608), identifying third party websites in user navigation paths with the target website (610), and/or identifying third party websites mentioned in media posts with the target website, such as social media posts, news articles, and/or other media content (612). The determinations for steps 604-612 can be performed using data sources that include aggregation and other broader web traffic and usage information across multiple different websites, such as the user search behavior 246, as well as data that may be available from crawling websites and web content 256, and from user website behavior 244 for the target website. Other data sources can also be used. [0084] SEO and other engagement data for the identified websites can be accessed (614) and trending/popular content can be identified based on analysis of that data (616). For example, the SEO data 246 and other engagement data, such as social media data 248, can be analyzed to identify presently trending keywords, topics, and/or content elements based on any of a variety of techniques, such as statistical analysis and/or semantic analysis. Identifying competitor websites and other third party websites that have overlapping users and/or content with the target website, and then using data associated with those sites can help to identify content and/or topics that are presently of interest to the users of the target website, even if they are not presently known to the owners/managers of the target website. The trending and/or popular content can be output (618).

[0085] FIG. 7 is a flowchart of an example process 700 for identifying candidates for new and/or improved content on a target website. The example process 700can be performed on any of a variety of systems, such as the web content performance systems 102 and/or 202, as described above with regard to FIGS. 1A-B and 2. The process 700 can be performed as part of the process 300, for example, at step 308. [0086] Trending/popular content topics can be received (702) and popularity scores for the trending/popular content topics can be determined (704). Popularity scores can be a quantification of how popular the content topics are with regard to the users associated with the competitor and other related third party websites (identified in process 600). The popularity scores can be based on the SEO and engagement data the content topics, such as the search volume for a keyword identifying a content topic. The popularity score can be based on one data source, and/or it can combine indicators popularity across multiple different data sources (e.g., SEO data, social media data). The popularity scores may additionally be determined based on one or more trained models that model popularity based on these data sources. Such a model may additionally be trained specifically for the target website - correlating SEO and engagement data to specific resulting popularity of that corresponding content on the target website.

[0087] The target website can be accessed and existing content topics on the target website can be identified (706), which can be used to determine which of the trending/popular content topics are already present on the target website (and potential candidates for improvement) and which ones are not, and would be candidates for new content to be added to the target website.

[0088] The steps 708-720 can be performed for each of the content popular/trending content topics. Starting at step 708, a content topic can be selected and a determination can be made as to whether or not the content topic is already present on the target website (710). If the content topic is already present, then the webpage(s) on the target website with the selected content topic can be considered a candidates for possible improvement. The performance score for the target webpage with the content topic can be determined, as discussed above, and a determination can be made as to whether the performance score for the target webpage is below a threshold performance level (712). If the target webpage is below the threshold level, then the target webpage and the selected can be added as an improvement candidate (716). If the target webpage is not below the threshold level — meaning the target webpage is already performing well and does not present a significant need for improvement — then a next content topic can be selected and evaluated (720 looping back to 708).

[0089] Returning back to step 710, if the selected content topic is not already present on the target website, then the popularity score of the selected content can be compared against a threshold level (714). If the content topic has a popularity score above the threshold level — meaning the content is sufficiently popular so as to warrant potential addition to the target website — then the selected content can be added as a new content candidate (718). Although not discussed above, step 712 may additionally consider the popularity of the selected content element in combination with the performance score of the target website when making the determination at step 712 — permitting improvement candidates to only be added when the webpage is performing poorly and the content topic is sufficiently popular so as to warrant consideration for improvement. If more content topics are available for consideration (720), then the steps 706-720 can be repeated until all popular/trending content topics have been evaluated. The new content candidates and the improvement content candidates can be output (722).

[0090] FIGS. 8A-B are flowcharts of example processes 800 and 850 for generating content recommendations for a target website. The example processes 800 and 850 can be performed on any of a variety of systems, such as the web content performance systems 102 and/or 202, as described above with regard to FIGS. 1 A-B and 2. The processes 800 and 850 can be performed as part of the process 300, for example, at step 312.

[0091] Referring to FIG. 8A, the example process 800 is generally directed to determining recommendation scores for new and improvement content candidates. As part of the process 800, new content candidates can be received (802) and projected performance scores and implementation details can be determined by applying each of the new content candidates to the trained performance model discussed above (804). The implementation details can include, for example, formatting, layout, configuration, webpage structure, external features, and/or timing aspects for the recommended new content candidates. The projected performance score can estimate an expected level of performance for the new content being added to the website with the implementations details using the performance model, which can incorporate the historical performance of other content using a variety of implementation details on the target website. Recommendation scores for each of the content candidates can be determined based on the projected performance scores and the content popularity scores, as discussed above (806). The recommendation score can combine both the projected performance score and the content popularity score. For example, a new content topic that appears to be highly popular and will have modest projected performance on the target website may have a recommendation score that is greater than another new content topic that has low popularity yet a higher projected performance on the target website. The recommendation score can seek to provide a quantifiable way to rank and sort content recommendations based an overall impact and importance of the new content to the target website. [0092] Steps 808-812, similar to steps 802-806, can be performed for improvement candidates. The recommendation score for improvement candidates at step 812 may additionally be based on the current performance score for the target webpage that would be improved as part of the improvement candidate. For example, a recommendation score may be increased if there is a significant increase from the current performance score to the projected performance score for the target website. Similarly, the recommendation score may be decreased if there is only a nominal increase (or a decrease) from the current performance score to the projected performance score for the target website.

[0093] The new and improvement content candidates can be ranked based on their recommendation scores (814) and can be output for presentation to and potential implementation by the website owner/manager (816).

[0094] Referring to FIG. 8B, the example process 850 is generally directed to determining recommendation scores for new and improvement content candidates when using an initial content focus, such as an initial content focus based on user input and/or content that is present on an existing target webpage. As part of the process 850, new and improvement content candidates are received (852) along with an initial content focus (854). For example, a user can provide a keyword or other input that can be used to effectively seed the content recommendation and implementation details that are provided back to the user - effectively refining the user’s initial thoughts about new content to add to the target website so as to optimize the performance the content that is added to the website. Similarly, a user and/or the system may select one or more target webpages on the website (or portions thereof) to seed the content recommendation that is provided, including for providing specific recommendations for improving the content on those pages.

[0095] One or more related content candidates can be identified based on a semantic comparison of the initial content focus and the content candidates identified in, for example, process 700 (856). For example, content topics that are similar and/or related to the initial content focus can be identified so as to provide recommendations that are substantively related to the initial focus. Any of a variety of techniques can be performed to identify related and/or similar content topics, such as through semantic analysis, machine learning techniques, subject matter taxonomies, and/or other techniques. For each of the related content candidates, projected performance scores and implementation details can be determined (858) and recommendation scores can be determined (860), similar to the discussion above with regard to steps 804-806. The related content candidates can be ranked based on the recommendations scores 862, and output for consideration, evaluation, and implementation by the owner/manager of the target website (864).

[0096] FIGS. 9A-C are screenshots of an example CMS user interface 900 and process flow for presenting and implementing new content recommendations related to a target website. The example CMS user interface 900 can be presented on any of a variety of client devices, such as the website owner client devices 212, based on interaction with a CMS (e.g., CMS 210) and a web content performance system, such as the such as the web content performance systems 102 and/or 202, as described above with regard to FIGS. 1A-B and 2.

[0097] Referring to FIG. 9A, an example CMS user interface 900 is depicted for creating a new page. In the example, the fields for the new page are blank. A selectable element 902 is presented through which a user can obtain content recommendations for the target website.

[0098] Referring to FIG. 9B, in response to receiving user selection of the element 902, the user interface 900 submits a request for, receives, and presents content recommendations 910, which can be generated using the systems, processes, and devices described above. The content recommendations 910 include new page content recommendations 912 and improved page content recommendations 914. Each of the content recommendations can include a variety of details, including the content topic 916a-c, the recommended content 918a-c, the recommended page layout and configuration for the new content 920a-c, the recommended webpage structure for the new content 922a-c, the recommendation score 924a-c, the projected performance score 926a-c, and/or the content popularity score 928a-c. The improvement content recommendation can additionally include information identifying the current page (“page P6”) along with information identifying how the page is recommended to be changed (e.g., “decrease word count...”), and a comparison of the current performance of the page 932 with the projected performance score 926c, which can assist a user in assessing the impact of the proposed changes. Additional and/or alternate details can be provided. Each of the content recommendations can be provided with selectable elements 930a-c that the user can select to readily implement the content recommendations, such as in the CMS interface 900.

[0099] Referring to FIG. 9C, in response to receiving user selection of the element 930a for the new content recommendation for “dogs” 916a, the fields in the user interface 900 for adding new content to the website can be automatically populated with content and implementation details from the recommendation (940- 950). For example, the content, layout, structure, and configuration of the new projected content can be automatically populated into the fields, along with guidance for the user to fill out a remaining portion of the new page content, of the user interface 900.

[0100] FIGS. 10A-C are screenshots of an example CMS user interface 1000 and process flow for presenting and implementing content recommendations for improving existing website content on a target website. The example CMS user interface 1000 can be presented on any of a variety of client devices, such as the website owner client devices 212, based on interaction with a CMS (e.g., CMS 210) and a web content performance system, such as the such as the web content performance systems 102 and/or 202, as described above with regard to FIGS. 1A-B and 2.

[0101] Referring to FIG. 10A, an example CMS user interface 1000 for editing an existing webpage 1008 on a website is presented. The interface 1000 includes a navigation pane 1002 for editing different pages on the website and options for a selected page 1004, including an option 1006 to improve the page via content recommendations as described throughout this document.

[0102] Referring to FIG. 10B, in response to selecting the option 1006, a pane 1010 with content recommendations and implementations 1012-1020 for the webpage can be presented for the webpage. The recommendation details 1012- 1020 can be similar to those described above with regard to FIG. 9B. The pane 1010 can additionally be provided with a selectable element 1022 that a user can select to implement and auto-populate the content recommendation on the webpage. [0103] Referring to FIG. 10C, in response to selecting the element 1022, the recommended content improvements 1030 can be automatically populated on and added to the webpage 1008, along with guidance for completing the content.

[0104] FIGS. 11A-C are screenshots of an example CMS user interface 1100 and process flow for presenting and implementing new content recommendations based on user input for a target website. The example CMS user interface 1100 can be presented on any of a variety of client devices, such as the website owner client devices 212, based on interaction with a CMS (e.g., CMS 210) and a web content performance system, such as the such as the web content performance systems 102 and/or 202, as described above with regard to FIGS. 1A-B and 2.

[0105] Referring to FIG. 11A, an example CMS user interface 1100 is presented for content recommendations 1102, depicted with example new content recommendations 1104 and 1106 similar to those presented in the user interface 900 described above with regard to FIG. 9B. The user interface 1100 additionally includes features through which a user can customize the content recommendations 1108, including through the use of a textual input field 1110. Additional and/or alternate input mechanisms are also possible, such as through selection of existing webpages and/or other existing content elements.

[0106] Referring to FIG. 11 B, in response to receiving the example user input “ice cream” in the input field 1110, the user interface 1100 can request, receive, and present customized content recommendations 1120 based on the user input. The customized content recommendations can be determined using, for example, the process 850 described above with regard to FIG. 8B, and can result in the presentation of content recommendations 1122 and 1124, which are related to the user input (“ice cream”).

[0107] Referring to FIG. 11 C, in response to the user selecting to implement one of the recommendations, the user interface 1100 can be automatically populated with features 1140-1150 from the customized recommendation, similar to FIG. 9C. [0108] Although the content recommendations and disclosed technology are generally described above with regard to webpages and webpage content, the disclosed technology can be applied to content elements generally, including mobile app content and content elements that are presented on other websites, such as ad content evaluated and presented on other websites via an online ad platform, product information presented in online retail stores, social media content that is presented on social media platforms, and others.

[0109] FIG. 12 shows an example of a computing device 1200 and an example of a mobile computing device that can be used to implement the techniques described here. The computing device 1200 is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The mobile computing device is intended to represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smart-phones, and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed in this document. [0110] The computing device 1200 includes a processor 1202, a memory 1204, a storage device 1206, a high-speed interface 1208 connecting to the memory 1204 and multiple high-speed expansion ports 1210, and a low-speed interface 1212 connecting to a low-speed expansion port 1214 and the storage device 1206. Each of the processor 1202, the memory 1204, the storage device 1206, the high-speed interface 1208, the high-speed expansion ports 1210, and the low-speed interface 1212, are interconnected using various busses, and can be mounted on a common motherboard or in other manners as appropriate. The processor 1202 can process instructions for execution within the computing device 1200, including instructions stored in the memory 1204 or on the storage device 1206 to display graphical information for a GUI on an external input/output device, such as a display 1216 coupled to the high-speed interface 1208. In other implementations, multiple processors and/or multiple buses can be used, as appropriate, along with multiple memories and types of memory. Also, multiple computing devices can be connected, with each device providing portions of the necessary operations (e.g., as a server bank, a group of blade servers, or a multi-processor system).

[0111] The memory 1204 stores information within the computing device 1200. In some implementations, the memory 1204 is a volatile memory unit or units. In some implementations, the memory 1204 is a non-volatile memory unit or units. The memory 1204 can also be another form of computer-readable medium, such as a magnetic or optical disk.

[0112] The storage device 1206 is capable of providing mass storage for the computing device 1200. In some implementations, the storage device 1206 can be or contain a computer-readable medium, such as a floppy disk device, a hard disk device, an optical disk device, or a tape device, a flash memory or other similar solid state memory device, or an array of devices, including devices in a storage area network or other configurations. A computer program product can be tangibly embodied in an information carrier. The computer program product can also contain instructions that, when executed, perform one or more methods, such as those described above. The computer program product can also be tangibly embodied in a computer- or machine-readable medium, such as the memory 1204, the storage device 1206, or memory on the processor 1202.

[0113] The high-speed interface 1208 manages bandwidth-intensive operations for the computing device 1200, while the low-speed interface 1212 manages lower bandwidth-intensive operations. Such allocation of functions is exemplary only. In some implementations, the high-speed interface 1208 is coupled to the memory 1204, the display 1216 (e.g., through a graphics processor or accelerator), and to the high-speed expansion ports 1210, which can accept various expansion cards (not shown). In the implementation, the low-speed interface 1212 is coupled to the storage device 1206 and the low-speed expansion port 1214. The low-speed expansion port 1214, which can include various communication ports (e.g., USB, Bluetooth, Ethernet, wireless Ethernet) can be coupled to one or more input/output devices, such as a keyboard, a pointing device, a scanner, or a networking device such as a switch or router, e.g., through a network adapter.

[0114] The computing device 1200 can be implemented in a number of different forms, as shown in the figure. For example, it can be implemented as a standard server 1220, or multiple times in a group of such servers. In addition, it can be implemented in a personal computer such as a laptop computer 1222. It can also be implemented as part of a rack server system 1224. Alternatively, components from the computing device 1200 can be combined with other components in a mobile device (not shown), such as a mobile computing device 1250. Each of such devices can contain one or more of the computing device 1200 and the mobile computing device 1250, and an entire system can be made up of multiple computing devices communicating with each other.

[0115] The mobile computing device 1250 includes a processor 1252, a memory 1264, an input/output device such as a display 1254, a communication interface 1266, and a transceiver 1268, among other components. The mobile computing device 1250 can also be provided with a storage device, such as a micro-drive or other device, to provide additional storage. Each of the processor 1252, the memory 1264, the display 1254, the communication interface 1266, and the transceiver 1268, are interconnected using various buses, and several of the components can be mounted on a common motherboard or in other manners as appropriate.

[0116] The processor 1252 can execute instructions within the mobile computing device 1250, including instructions stored in the memory 1264. The processor 1252 can be implemented as a chipset of chips that include separate and multiple analog and digital processors. The processor 1252 can provide, for example, for coordination of the other components of the mobile computing device 1250, such as control of user interfaces, applications run by the mobile computing device 1250, and wireless communication by the mobile computing device 1250.

[0117] The processor 1252 can communicate with a user through a control interface 1258 and a display interface 1256 coupled to the display 1254. The display 1254 can be, for example, a TFT (Thin-Film-Transistor Liquid Crystal Display) display or an OLED (Organic Light Emitting Diode) display, or other appropriate display technology. The display interface 1256 can comprise appropriate circuitry for driving the display 1254 to present graphical and other information to a user. The control interface 1258 can receive commands from a user and convert them for submission to the processor 1252. In addition, an external interface 1262 can provide communication with the processor 1252, so as to enable near area communication of the mobile computing device 1250 with other devices. The external interface 1262 can provide, for example, for wired communication in some implementations, or for wireless communication in other implementations, and multiple interfaces can also be used. [0118] The memory 1264 stores information within the mobile computing device 1250. The memory 1264 can be implemented as one or more of a computer- readable medium or media, a volatile memory unit or units, or a non-volatile memory unit or units. An expansion memory 1274 can also be provided and connected to the mobile computing device 1250 through an expansion interface 1272, which can include, for example, a SIMM (Single In Line Memory Module) card interface. The expansion memory 1274 can provide extra storage space for the mobile computing device 1250, or can also store applications or other information for the mobile computing device 1250. Specifically, the expansion memory 1274 can include instructions to carry out or supplement the processes described above, and can include secure information also. Thus, for example, the expansion memory 1274 can be provide as a security module for the mobile computing device 1250, and can be programmed with instructions that permit secure use of the mobile computing device 1250. In addition, secure applications can be provided via the SIMM cards, along with additional information, such as placing identifying information on the SIMM card in a non-hackable manner.

[0119] The memory can include, for example, flash memory and/or NVRAM memory (non-volatile random access memory), as discussed below. In some implementations, a computer program product is tangibly embodied in an information carrier. The computer program product contains instructions that, when executed, perform one or more methods, such as those described above. The computer program product can be a computer- or machine-readable medium, such as the memory 1264, the expansion memory 1274, or memory on the processor 1252. In some implementations, the computer program product can be received in a propagated signal, for example, over the transceiver 1268 or the external interface 1262.

[0120] The mobile computing device 1250 can communicate wirelessly through the communication interface 1266, which can include digital signal processing circuitry where necessary. The communication interface 1266 can provide for communications under various modes or protocols, such as GSM voice calls (Global System for Mobile communications), SMS (Short Message Service), EMS (Enhanced Messaging Service), or MMS messaging (Multimedia Messaging Service), CDMA (code division multiple access), TDMA (time division multiple access), PDC (Personal Digital Cellular), WCDMA (Wideband Code Division Multiple Access), CDMA2000, or GPRS (General Packet Radio Service), among others. Such communication can occur, for example, through the transceiver 1268 using a radio-frequency. In addition, short-range communication can occur, such as using a Bluetooth, WiFi, or other such transceiver (not shown). In addition, a GPS (Global Positioning System) receiver module 1270 can provide additional navigation- and location-related wireless data to the mobile computing device 1250, which can be used as appropriate by applications running on the mobile computing device 1250. [0121] The mobile computing device 1250 can also communicate audibly using an audio codec 1260, which can receive spoken information from a user and convert it to usable digital information. The audio codec 1260 can likewise generate audible sound for a user, such as through a speaker, e.g., in a handset of the mobile computing device 1250. Such sound can include sound from voice telephone calls, can include recorded sound (e.g., voice messages, music files, etc.) and can also include sound generated by applications operating on the mobile computing device 1250.

[0122] The mobile computing device 1250 can be implemented in a number of different forms, as shown in the figure. For example, it can be implemented as a cellular telephone 1280. It can also be implemented as part of a smart-phone 1282, personal digital assistant, or other similar mobile device.

[0123] Various implementations of the systems and techniques described here can be realized in digital electronic circuitry, integrated circuitry, specially designed ASICs (application specific integrated circuits), computer hardware, firmware, software, and/or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0124] These computer programs (also known as programs, software, software applications or code) include machine instructions for a programmable processor, and can be implemented in a high-level procedural and/or object-oriented programming language, and/or in assembly/machine language. As used herein, the terms machine-readable medium and computer-readable medium refer to any computer program product, apparatus and/or device (e.g., magnetic discs, optical disks, memory, Programmable Logic Devices (PLDs)) used to provide machine instructions and/or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term machine-readable signal refers to any signal used to provide machine instructions and/or data to a programmable processor.

[0125] To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0126] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a client computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0127] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

[0128] While this specification contains many specific implementation details, these should not be construed as limitations on the scope of the disclosed technology or of what may be claimed, but rather as descriptions of features that may be specific to particular embodiments of particular disclosed technologies. Certain features that are described in this specification in the context of separate embodiments can also be implemented in combination in a single embodiment in part or in whole. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Moreover, although features may be described herein as acting in certain combinations and/or initially claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination may be directed to a subcombination or variation of a subcombination. Similarly, while operations may be described in a particular order, this should not be understood as requiring that such operations be performed in the particular order or in sequential order, or that all operations be performed, to achieve desirable results. Particular embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims.

Claims

CLAIMS WHAT IS CLAIMED IS:

1 . A method for providing content recommendations for a target website, the method comprising: receiving, at a computer system, a request for content recommendations for the target website; identifying, by the computer system, user engagement data for a plurality of third party websites that are related to the target website; determining, by the computer system, a plurality of content candidates that are currently popular or trending among users of the plurality of third party websites based on the user engagement data; accessing, by the computer system, a performance model for the target website that models performance of content and webpage presentation details on the target website; determining, by the computer system, projected performance scores and recommended presentation details for the plurality of content candidates by applying the performance model for the target website to the content candidates; selecting, by the computer system, a portion of the of the content candidates based on the projected performance scores as content recommendations; and transmitting, by the computer system, the content recommendations and corresponding recommended presentation details.

2. The method of claim 1 , wherein: the content request is for new content to be added to the target website, and the content recommendations include content candidates that are different from content presently included on the target website, that are presently popular or trending among the users of the third party websites, and that are projected to have at least a threshold performance level on the target website.

38 The method of claim 2, wherein: the content request includes user input specifying subject matter for the new content to be added to the target website, and determination of the content candidates is further based on the subject matter specified by the user input, the content candidates being determined based, at least in part, on being related or similar to the subject matter specified by the user input. The method of claim 1, wherein: the content request is for identification of improvements to preexisting content on the target website; and the content recommendations include content candidates that are different from content presently included on the target website, that are presently popular or trending among the users of the third party websites, and that are projected to have at least a threshold performance level on the target website. The method of claim 5, wherein: the content request includes identification of a specific webpage on the target website for improvement, and determination of the content candidates is further based on existing content included on the specific webpage, the content candidates being determined based, at least in part, on being related or similar to the existing content included on the specific webpage. The method of claim 1 , wherein the plurality of third party websites are determined to be related to the target website based on the plurality of third party websites being competitors of the target website. The method of claim 1 , wherein the plurality of third party websites are determined to be related to the target website based on the plurality of third party websites appearing together with the target website in search engine results to one or more search queries.

39 The method of claim 1 , wherein: determining the content candidates further includes determining popularity scores for each of the content candidates, and selecting the portion of the content candidates is further based on the popularity scores for each of the content candidates. The method of claim 8, wherein the popularity scores are determined using a popularity model that correlates the user engagement data for the third party websites with an performance of content on the target website. The method of claim 1 , further comprising: training the performance model for the target website based on performance data for the target website and content for the webpages on the target website. The method of claim 10, wherein the training comprises: identifying content features and presentation features for the content on the webpages; determining performance scores for the webpages based on the performance data; and training the performance model using the content features, the presentation features, and the performance scores for the webpages. The method of claim 11 , wherein the training further comprises: identifying previously implemented content recommendations on one or more updated webpages on the target website; accessing previously projected performance scores for the previously implemented content recommendations; determining current performance scores for the updated webpages based on the performance data; identifying content features and presentation features for the updated webpages; and

40 training the performance model using the previously implemented content recommendations, the previously projected performance scores, the current performance scores, and the content features and presentation features for the updated webpages. The method of claim 1 , wherein the content recommendations are configured to be presented in a user interface with a selectable implementation option that, when selected, populates a webs content editing interface with content elements and presentation elements based on the content recommendations and corresponding presentation details. The method of claim 13, wherein the user interface comprises a content management system (CMS) user interface. The method of claim 1 , wherein the computer system is part of a CMS platform that is used for editing and managing the target website. The method of claim 1 , wherein the computer system provides an application programming interface (API) that is used by third party website management systems to provide content recommendations. The method of claim 16, wherein at least a portion of the third party website management systems include a CMS platform.