TW201432595A - Online marketplace to facilitate the distribution of marketing services from a marketer to an on-line merchant - Google Patents

Abstract

Methods and systems are provided for use, for example, in e-commerce. E-commerce data from a merchant may be received regarding interactions of customers with a website of the merchant. Characteristics of the e-commerce data may be identified. Based at least in part on the one or more identified characteristics, a plurality of marketing actions may be selected, or automatically selected, that are potentially effective for the merchant. The selected marketing actions may be ranked based at least in part on one or more of a quality score and one or more bids associated with each of the selected marketing actions. A ranked list of the selected marketing actions may be provided, or recommended, to the merchant.

Description

A line listing that is used to facilitate the distribution of marketing services from a single seller to an online merchant.

Some implementations are generally related to the field of e-commerce systems and methods, and more particularly, to providing a line listing set for facilitating advertisements from a marketer to an online merchant and displaying it in software (eg, a software application) and/or Or marketing services (for example, sales).

Advertising behavior and marketing behavior are central to any business activity, especially for online business practices that do not interact directly with customers. Advertising behavior usually begins with a product and an advertisement for that product. Traditional advertising practices include television commercials, bulletin boards, magazine advertisements, and other sources that may be viewed or viewed by the masses.

However, online advertising is not the same. Online merchants who trade or sell products must have more than one product and potential customers. Must also have online visibility, that is, the online identity of potential customers must be known and visible. In today's digital world, people spend more time on the Internet. Therefore, the most effective source of product transactions and services for online merchants is usually a passive audience on the World Wide Web. E-commerce businesses that have identified a product, a target audience, and have acquired a website for their business activities are now millions, and if not millions, have quickly searched the Internet for billions of dollars, Through such as Google ^TM , Yahoo! ^{The TM} and Bing ^TM popular search engines create lead search for keyword searches.

There are many ways in which an online e-commerce merchant can find a target customer.

First, the customer may know the identity of the online merchant website and the corresponding global resource locator (URL, "uniform resource locator"). For example, a user familiar with Amazon ^TM shopping sites may be listed simply enter "www.Amazon.com" network address in a browser, you can direct customers to the leadership of the e-commerce site.

Second, the e-commerce merchant can appear in an in-place ad link of a search result or another web page. In order to display the search results as soon as possible, the e-commerce merchant can become such as Google Ads or Yahoo! Partners in services such as Ads, and purchase certain keywords related to their business, to create an automatic sequential correlation between a search keyword (speech, etc.) and the corresponding e-commerce merchant, thus ensuring the electronic Business merchants appear in the results of a search engine web page. These services typically use a "click-to-click" payment method, and each time an Internet user clicks on the e-commerce merchant's URL link, the e-commerce merchant is charged a specific fee.

Furthermore, an e-commerce merchant can choose to purchase "online real estate", which is a product through a number of popular third-party websites such as Facebook ^TM , Groupon ^TM , and homepages of websites such as CNN.com, NYTimes.com, etc. Service advertising campaign. These websites may be partners of the e-commerce merchant and allow placement of e-commerce advertisements on their websites or pop-up advertisements when a user visits the website. This often takes a significant amount of time from the marketing department or employee to identify potential partner sites and enter into contracts with the third party to display advertisements on their websites.

Finally, use an e-commerce merchant such as Google or Yahoo! The above method can be used when the service purchases a keyword and defines a relevant regional, product and target customer for its products and services, and also establishes a partnership with a third-party website to obtain online advertising real estate for e-commerce commercial advertising. Some combinations. However, even with the options currently available to e-commerce merchants, e-commerce merchant owners must spend a lot of time, resources, and capital to build an advertising campaign, research the appropriate search engines that can be used, and define many such keywords and goals. Complex variables of audience, regionality, product classification, product, product attributes, etc., and monitoring and collecting statistics on customer behavior, and determining which type of user and what type of website is appropriate and effective for advertising.

This problem is compounded by the continuous development and change of user behavior, products, website content and keywords and their associated back-end search engines. Small businesses often don't have the time or resources to hire graphic designers who create advertisements, marketing professionals who hire research and learn effective resources and targeted advertising methods, and hire familiar users like Google, Yahoo! Professional information technology (IT, "information technology") personnel with Bing ^TM third-party advertising system.

After a clear understanding of the problem described above, is located in Mountain View, California (Mountain View) of Lexity ^TM company has started to provide software coding marketing actions for e-commerce merchants, when installed on a merchant's site, the site can generate for businesses Advertising/marketing behavior and optimizing advertising/marketing behavior. Lexity also provides software for coding marketing actions that, when installed on the merchant's computer system, facilitates "behind-the-scenes" work, manages advertising and provides information to the merchant (for example, through the Lexity user interface). The marketing actions can be presented by a software application (often referred to simply as "Apps"). Lexity has developed a full suite of such e-commerce applications, each of which is designed to optimize the merchant website in a particular manner and/or to configure the merchant's computer system in a particular manner (eg, to manage the merchant) Advertising, providing feedback to the merchant, etc.). The inventors expect that in the near future, companies other than Lexity (ie, "third-party marketers") will begin to design software applications that have similar purposes to support online merchant marketing and advertising needs. When more and more applications are being marketed from Lexity and third-party marketers, merchants need assistance in selecting one or more marketing actions (eg, software applications) that are best suited for websites and/or computer systems (eg, Help the e-commerce merchant to maximize profits). To address this need, the inventors developed a marketplace that integrates marketers with merchants. An important feature of the marketplace is a recommendation engine that automatically suggests certain e-commerce applications to a merchant.

In one embodiment, a system is designed to facilitate appropriate marketing actions by the e-commerce merchant based on the complete e-marketing information of an e-commerce merchant. Specifically, the e-commerce material can include any and all of the materials accumulated in an e-commerce store operation, including orders, website traffic data, the store products, and the like. Statistical methods and heuristic calculus generated by human experts can be applied to this data to determine which marketing action has the best chance of increasing the profit of the merchant.

Marketing actions can take many forms, including providing instructions to be executed by the merchant and a marketing software application. Suggested marketing actions for the merchant through various channels, including direct contact via email (targeted merchants listed in the salesperson database), and e-commerce marketing application users installed in the merchant Various visual presentations in the interface. In addition, third-party marketers can bid on locations in these various pipelines, and through these pipelines, marketing recommendations can be made. The bid price can be made by other bids, It is determined by the bid position and by a quality score (i.e., the more the system can determine the desired offer of the merchant, the lower the bid price can be).

100‧‧‧ computer system

102‧‧‧ busbar

104‧‧‧Processor

106‧‧‧ main memory

108‧‧‧Read-only memory

110‧‧‧Storage device

112‧‧‧ display

114‧‧‧Input device

116‧‧‧ cursor control device

118‧‧‧Communication interface

200‧‧‧ computer system

202‧‧‧ hardware layer

204‧‧‧Operating system

206‧‧‧Application software layer

208‧‧‧Networks and Protocols

210‧‧‧ graphical user interface

300‧‧‧E-commerce system

302‧‧‧Markers

304‧‧‧Steps

306‧‧‧Markers

308‧‧‧Recommended engine

310‧‧‧Business Website

312‧‧‧Marketing

314‧‧ businesses

316‧‧‧Customer

350‧‧‧Description method

352‧‧‧Recommended engine

354‧‧‧Steps

356‧‧‧Steps

358‧‧‧Steps

360‧‧‧Steps

362‧‧‧Steps

400‧‧‧ Flowchart

402‧‧‧Steps

404‧‧‧Steps

405‧‧‧Steps

406‧‧‧Steps

408‧‧‧Steps

410‧‧‧Steps

412‧‧‧Steps

500‧‧‧flow chart

502‧‧‧Steps

504‧‧‧Steps

506‧‧‧Steps

700‧‧‧Flowchart

702‧‧‧Steps

704‧‧‧Steps

706‧‧‧Steps

708‧‧ steps

710‧‧ steps

712‧‧‧Steps

714‧‧‧Steps

716‧‧ steps

800‧‧‧ Flowchart

802‧‧ steps

804‧‧‧ steps

806‧‧‧Steps

808‧‧‧Steps

810‧‧‧Steps

812‧‧‧ steps

814‧‧‧Steps

900‧‧‧Flowchart

902‧‧ steps

904‧‧‧Steps

906‧‧‧Steps

908‧‧‧Steps

910‧‧ steps

912‧‧ steps

914‧‧‧Steps

916‧‧‧Steps

918‧‧ steps

The specific embodiments can be fully understood from the following detailed description of certain embodiments and the drawings, wherein:

The first and second figures are examples of computer architectures of computer systems constructed in accordance with specific embodiments.

The third figure depicts the components of an e-commerce system in which specific embodiments are implemented.

Figure 3A depicts a recommendation engine in which specific embodiments are implemented.

The fourth figure depicts a flow diagram that can be performed by a recommendation engine in accordance with an embodiment.

The fifth figure depicts a flow diagram that includes steps that can be performed in the generation of a rule repository, in accordance with an embodiment.

The sixth diagram depicts an exemplary table that forms a set of rules to map recognized features of e-commerce material to an e-commerce application, in accordance with an embodiment.

The seventh figure depicts a flow chart for performing a sorting marketing action by a recommendation engine, in accordance with an embodiment.

The eighth figure depicts a flow diagram for distributing (e.g., selling) marketing services to a merchant, performed by a third party marketer, in accordance with an embodiment.

The ninth diagram depicts a flow diagram for receiving (e.g., purchasing) a marketing service from a merchant by a merchant, in accordance with an embodiment.

Specific embodiments are discussed below with reference to the first to ninth figures, which are The illustration of some specific embodiments of the invention is not intended to limit the scope of the claimed invention.

The first figure depicts an example of a computer system 100 on which any of the methods and systems of various embodiments can be implemented. The computer system 100 can represent any of the computer systems and physical components required to perform the computerized methods discussed with respect to the third to ninth figures, and more particularly, can represent a server , a client or other computer system on which an e-commerce server, website, web browser and/or web analytics application can be implemented. The computer system 100 can include a bus bar 102 or other communication mechanism for information communication, and a processor 104 coupled to the bus bar 102 for information processing. The computer system 100 also includes a main memory 106, such as a random access memory (RAM) or other dynamic storage device coupled to the bus bar 102 for storing information and instructions executed by the processor 104. The main memory 106 can also be used to store temporary variables or other intermediate information during execution of instructions by the processor 104. The computer system 100 can further include a read only memory (ROM) 108 or other static storage device coupled to the bus bar 102 for storing static information and instructions executed by the processor 104. A storage device 110, such as a hard disk, can also be provided and coupled to the busbar 102 for storing information and instructions.

The computer system 100 can be coupled to a display 112 via the busbar 102 to display information to the user. However, in the case of a server, the display may not be provided, and all management of the server may be through a remote client. The end is carried out. Similarly, an input device 114 (including alphanumeric and other typing) can also be coupled to the busbar 102 for information communication and command selection of the processor 104, but the server configuration may not provide the device. Another type of user input device is a cursor control device 116, such as a mouse, trackball or cursor direction key, for direction information communication and command selection of the processor 104, and for controlling cursors on the display 112. mobile. The input device may or may not be provided in a server configuration.

The computer system 100 can also include a communication interface 118 coupled to the busbar 102. The communication interface 118 provides wired and/or wireless two-way data communication to/from the computer system, for example, via a local area network (LAN) or other network (including the Internet). The communication interface 118 can also transmit and receive electronic, electromagnetic or optical information that carries digital data representing various forms of information and instructions. For example, two or more computer systems 100 may be networked in a conventional manner, with each computer system using a different communication interface 118.

It should be understood that the recommendation engine and merchant website can be implemented in the illustration of the individual computer system 100, which can utilize a client machine, a server machine, or a server, client known to those of ordinary skill in the art. Some combinations of the end with other network devices.

The various databases described herein are computer-based record keeping systems. In other words, each of these databases is a combination of computer hardware and software, and its overall actions allow for the storage and retrieval of information (data). Accordingly, it can be a similar computer system 100, and its features are often stored as a storage medium capable of accommodating a significant amount of information.

The second figure depicts a computer system 200 from the perspective of a software structure in accordance with an embodiment of the present invention. The computer system 200 can be a server or a group of servers or computers. The various hardware components of the computer system 200 are presented in a hard layer 202. An operating system 204 extracts the hardware layer 202 and acts as the master of various applications in the application layer 206. Networks and protocols such as Hypertext Transfer Protocol (HTTP), Secure Hypertext Transfer Protocol (HTTPS), and Secure Sockets Layer (SSL) are used to communicate data with other networks over the Internet. System components, such as the recommendation engine and merchant website, can be implemented in a computer system, such as a computer System 200.

As depicted in the third diagram, a recommendation engine 308 interfaces with marketers 302, 306 (or more generally, advertisers, sales consultants, web analytics engineers, etc.) and one or more merchants 314 (although only described herein) a business). The two-marketer form described herein, one type incorporates the recommendation engine 302, and one type does not incorporate the recommendation engine 306 (hereinafter, referred to as a "third-party marketer"). In other words, there is a proposal to host the recommendation engine and design a company/organization of marketing actions, and then perform marketing actions by the recommendation engine. There are also other companies/organizations that design marketing actions but do not control the recommendation engine (in other words, third-party marketers). In order for its marketing actions to be more prominently displayed to a merchant, more likely to suggest to a merchant, to have a higher ranking in the list of marketing actions presented to a merchant 312, etc., third party advertisers may recommend to the merchant The engine makes a bid (or other incentive). The bidding process will be described in more detail below.

As described in the third figure, the recommendation engine can directly interface with the merchant website 310 and can receive e-commerce information about the customer interacting with the merchant website.

E-commerce data can generally be classified into "traffic-side" e-commerce data and "back-end" e-commerce data. The traffic side e-commerce data may include which web page the customer views, when to watch, how long to watch, and the like. The traffic side e-commerce data may also include what actions the customer performs on the merchant website (eg, which product the customer views, what items the customer places in a virtual shopping cart, etc.). The traffic side e-commerce data may also include whether the customer is reflowing and how many times the customers have returned. The traffic side e-commerce material may also include where the customer came from (eg, where the customer introduced the website to the merchant website), and a query to the search engine to direct the customer to the merchant website. More generally, the recommendation engine takes into account the entire behavior of a customer on the merchant's website, and effectively guides a first-time visitor. Go to the merchant's website, or make any returning customer back to any of the merchant's website.

The back-end e-commerce data can include what items the customer orders, who orders, the shipping origin, the shipping destination, and the way the shipment is requested (eg, land, express, etc.). It is important to note that the information is not always available. For example, a software program can be downloaded and therefore does not have shipping information associated with it. Back-end e-commerce materials can also include a merchant inventory (eg, what inventory products a store has, how long the merchant's inventory changes over time, whether certain products are out of stock, too much inventory, etc.). More generally, three important categories of e-commerce data include customer orders, products offered by the website, and website traffic.

How to monitor the customer's implementation of interaction with a merchant's website is a matter of practice, not the focus of the application of the present invention. One such implementation may include storing a web cookie file (e.g., a two-digit number) in a user web browser so that a merchant can associate a browsing history with the user. For example, the merchant can use a web cookie file to identify a customer as a returning customer or a first visited customer. The merchant website may also store commands (eg, cursor control events, keyboard events, etc.) that are executed by the user when visiting the merchant website. The merchant can also monitor user activity even when the user interacts with a website that is not affiliated with the merchant. The monitoring may involve the use of a keystroke program and/or a JAVA language script that the user can choose to join before allowing the monitoring.

In order to make the e-commerce data easier to analyze, the recommendation engine can perform statistical analysis to concentrate/compress the data. For example, the recommendation engine can calculate the average sales of the website every quarter of an hour, every week, every day, and the like. The recommendation engine can list the number of customers visiting the merchant's website in which days. From the concentrated e-commerce material or from the original e-commerce material, the recommendation engine can identify certain characteristics of the e-commerce material, such as the customers Purchase behavior (for example, the most common purchases on Thursday nights, returning customers is easier to purchase than first-time customers), or the demographic attributes of such customers (for example, most televisions are forty-year-old males) Most perfumes are purchased by a 20-year-old woman.)

Based on the identified features of the e-commerce material, one or more marketing actions suggested to a merchant can be automatically selected using a human expert heuristic algorithm. The heuristic algorithm may be retrieved in a rule database (ie, mapping certain features of the e-commerce material to a set of rules for one or more marketing actions).

For example, the recommendation engine may identify from the e-commerce material that there are only a few return customers, but those who are actually returning customers are more likely to purchase than the first-time customers. In fact, the inventors have observed that this feature is very common for merchant websites that sell expensive items, such as merchant websites that sell jewelry or real estate. In order to respond to this identified feature of the e-commerce material, the recommendation engine can suggest a marketing action (eg, suggesting a software application) that specifically targets the returning customer (in other words, also known as retargeting). For example, the redirect marketing action can provide a discounted price to the returning customer, provide more product descriptions, and/or provide a more in-depth product comparison. The redirect marketing action also presents the customers with products/items that were previously visited by the merchant website.

Other identified features of the e-commerce material may include traffic spikes. For example, a merchant's website that sells bouquets may experience traffic spikes on web traffic before Valentine's Day and Mother's Day. As another example, some merchants may experience traffic spikes on Wednesday or weekends. In order to respond to the described characteristics of the e-commerce material (ie, traffic spikes), the recommendation engine can suggest a quick talk application like the merchant that allows the customer to ask questions directly to the merchant website employee (eg, using a keyboard) , tablet, smart phone, voice recognition software, etc.) The employee may then provide an immediate response to the questions.

Other recognized features of the e-commerce material may include keywords entered into a search engine that direct the customer to select a link or advertisement (e.g., banner) for the merchant website. In order to respond to the described characteristics of the e-commerce material, the recommendation engine can suggest a merchant bidding on those keywords to increase similar customer traffic. More system, the recommendation engine may suggest encoded in a software application in marketing action, action to increase the marketing business for Adwords ^TM in a bid price or the frequency of those keywords (Adwords ^TM for the Google of Mountain View, CA provided A program that, when a potential customer on the Google search engine enters certain text, allows the merchant to bid on the text so that the business advertisement can be displayed next to the search result or as a sponsored search result).

In simple terms, the recommendation engine classifies what happens in a merchant store (eg, on a multi-axis such as products supplied by the merchant's website, customer orders, and website traffic), and is inspired by a human expert. A calculus-based rule that maps the classification to one or more suggested marketing actions. If the multiple characteristics of the e-commerce material have been identified, the recommendation engine can provide one or more marketing actions in response to each feature.

As described in the third figure, in response to the e-commerce material, bids regarding third party marketing actions and other materials, the recommendation engine may suggest one or more marketing actions to the merchant. The recommendation engine may suggest a single marketing action at a time (eg, suggesting the first highest ranking action in the recommendation sequence, and suggesting the last lowest ranking action in the recommendation sequence). Alternatively, the recommendation engine may suggest multiple marketing actions at a time, in which case the sorting list may provide such marketing actions. In a specific embodiment, the suggested marketing actions may be provided to the merchant along with certain promotional or quoting methods. For example, as a marketing application from marketer A Reward can provide a half-price (50%) discount for the merchant to purchase a marketing app from Futures A.

For the recommendation engine, there are a number of communication pipes that can transmit suggested marketing actions to a merchant. The recommendation engine can send an email to the merchant listing the suggested marketing actions in sorted order. In addition, when the merchant interacts with the marketing application, it can suggest (through banner ads, etc.) that the merchant installs a further marketing application.

Regardless of the precise presentation of such suggested actions, the ordering of such marketing actions can be determined in the following manner. The ranking of the one-line pin action can be specified based on a quality score indicating the estimated validity of the marketing action. The estimated validity is based on the average effectiveness of the marketing action of the merchant (eg, the amount of sales increased for each advertising fee, etc.).

The ranking can also be based on a bid offered by a third party marketer. For example, the marketing action that was originally ranked fourth may be ranked second because of the increase in the bid of a third party marketer. The price of the bid can be determined by other bids, the order in which it is to be improved, and the quality score. For example, for a marketing action with a higher quality score, the bid that raises the ranking to the first position will be lower than the marketing action with a lower quality score. The ordering of the sales actions is not infinitely increased with the increased bid price. There can be a quality score threshold, regardless of the individual bid price, marketing actions below this threshold are not recommended. If this threshold is not met, the recommendation engine will lose credibility when it suggests a clear and unrelated marketing action.

In addition to providing a relative ranking of such marketing actions, the recommendation engine can provide the merchant with a numerical score for each marketing action that indicates an estimated validity of the marketing action. The value score may be the same as the quality score or may be a quality score weighted by the third party bid price. In addition to the ranking and the numerical score, the recommendation engine also The manner in which the merchant can provide e-commerce data on which the recommendations are based (eg, traffic spikes, inventory, etc.). In addition, the recommendation engine can provide instructions for the merchant as to why the marketing action is recommended. In the case of a discounted price marketing action suggested on Friday, the recommendation engine can indicate to the merchant that the traffic spike of the merchant customer appears on Friday, and for similar merchants, the discount price offered at the peak of the customer can be increased. Sales.

In addition, these suggested marketing actions can be categorized into many categories, such as search engine marketing applications, redirect applications, and the like. In other words, the suggested applications can be classified according to their functionality. Therefore, if a marketer intends to provide an application that can provide search engine marketing, these marketers may compete with each other (for example, if only one search engine marketing application is recommended, then all of these marketers will compete with each other to Can be like a merchant suggesting the marketer's search engine marketing application). Accordingly, the marketer can bid on a particular application type. If the one-line vendor application is preferred in that type (eg, has a higher quality score), the application basically facilitates receiving a "handicap" in the type. If multiple applications are categorized into one type, the recommendation engine can also suggest multiple applications, but present the applications in a sorted list.

As described in the third figure, in order to respond to a suggested marketing action, the merchant may select one or more of the marketing actions. Once one or more of the suggested marketing actions are selected, the software code of the selected marketing action can be transmitted from the seller to the merchant website (or more precisely, to the server hosting the merchant website) . The software can then be installed on the merchant website.

After installing a software that encodes marketing actions on a merchant website, the software can monitor certain key metrics (eg, customers) of the merchant website during a particular time period or indefinitely Flow, product list, etc.). The metric can be reported to the merchant, so the merchant can evaluate the validity (or invalidity) of the marketing action. The metrics may also be provided to the recommendation engine, so the recommendation engine may adjust future quality scores for such marketing actions (or similar marketing actions).

In a specific embodiment, the recommendation engine periodically suggests marketing actions to the merchant without requiring the merchant to make an explicit re-request. In another embodiment, the merchant may also make an explicit marketing request for the recommendation engine. For example, the merchant can request the recommendation engine to provide marketing actions that can assist the merchant in increasing customer traffic, increasing brand recognition, increasing profits, increasing customer loyalty, increasing the number of customers returning, increasing the click rate of the advertisement, etc. (or more specifically Providing a software application). In order to respond to the marketing request, the recommended marketing action of the return may be based only on the merchant's marketing request, or may be based on both the analysis of the electronic commerce data and the marketing request.

If there is a discrepancy between the e-commerce data and the marketing demand determined by the recommendation engine based on the marketing request received from the customer, the recommendation engine can directly notify the customer of the difference. For example, the merchant may request a change to increase customer traffic for the merchant website. In response, the recommendation engine can notify the customer that the customer traffic is not actually a major key, but the primary key is to increase the number of customers, the difference being who has visited the website and who is making the purchase.

In another embodiment, both the merchant's marketing actions suggested in response to the e-commerce data analysis and the marketing actions suggested in response to a clear marketing request from the merchant may be provided. In some embodiments, the latter marketing actions may have a higher ranking than the former, while in other embodiments, the former marketing actions may have a higher than the latter Higher sorting.

FIG. 3A depicts a recommendation engine 352, which may be stored, for example, on a non-transitory computer readable storage medium or medium that stores a computer program executable by a computer processor, in accordance with an embodiment. logic. The recommendation engine 352 includes a variety of logic or program logic to implement aspects of certain embodiments, including the description method 350, and each of its steps 354-362. Specifically, at step 354, e-commerce material relating to customer interaction with the merchant website is received from a merchant. Next, at step 356, the characteristics of the e-commerce material are identified. Next, at step 358, a marketing action is selected based at least in part on the one or more identified features, or a marketing action is automatically selected, the marketing actions having a potential effect on the merchant. Then, at step 360, an ordering of the selected marketing actions is presented, the ranking of the selected marketing actions being based at least in part on one or more quality scores and one or more bids, wherein the one or more quality scores and the one Or multiple bids have each of the selected marketing actions. Finally, as described, at step 362, a ranked list of the selected marketing actions is provided or suggested to the merchant.

It is noted that in some embodiments, the ranking may not be performed, and one or more marketing actions may be selected and provided or suggested to the merchant without ranking.

In the following description, the fourth to eighth figures provide a further description of the procedures performed by each component of the system in the third figure.

According to a specific embodiment, the fourth figure depicts the steps performed by the recommendation engine. First, the recommendation engine can receive e-commerce material about a customer's interaction with the website from a merchant website. The recommendation engine can identify features related to the e-commerce material from the e-commerce material. The feature can be identified based on a repeating pattern in the e-commerce material. Such features may include customer behavior, customer demographics, and the like. More specifically, the e-commerce can be executed first. Statistical analysis of data to reduce the e-commerce data. Then, the type of the reduced data can be identified from the original e-commerce data.

Based on the identified features, the recommendation engine can automatically select one or more marketing actions suggested to the merchant. If more than one marketing action is selected, the marketing actions can be further sorted. Finally, the recommendation engine transmits one or more of the suggested marketing actions to the merchant and may be provided in a sorted order.

As mentioned above, the recommendation engine can receive an explicit marketing request from a merchant, and the suggested marketing actions can be based on the e-commerce material and the explicit marketing request.

According to a specific embodiment, the fifth figure depicts the steps associated with generating the rule database, the rule database mapping the characteristics of the e-commerce material to the marketing application. First, the recommendation engine can receive and/or can generate a narrative of each marketing action (eg, step 304 of the third figure). The following includes an example marketing application from Lexity and its related narrative:

Lexity Live: Stay sensitive to your business rush and instantly view what your customers are browsing, checking and buying!

Live Archive: Analyze the traffic changes of your website over time. Track the actions of your top customers page by page.

MailChimpROI: Analyze changes over time from your MailChimp email campaign.

Pinterest Report: Track most of your pinned products, competition and influential picks on Pinterest.

Quick Chat: Chat immediately with any visitor on your site and see exactly what they see and do.

eCommHub: Automatic inventory management with bookbinding wholesalers, fulfillment centers and other stores Processing with orders.

Shopping Feeds: Make your products automatically listed on the shopping comparison engine, including Google, Bing, and TheFind.

MixRank: Show your competitors' battles to show you exactly what is working for them.

Google Shopping: Make your product listed on Google's new paid list - Google Shopping instead of Google Product Search.

Retargeting: Respond to lost sales by showing redirected display ads to visitors who have not yet purchased.

Second, the recommendation engine can receive and/or generate a list of e-commerce material features:

Many first-time customers

Very few return customers

Many impulsive shoppers

Many shoppers who just can't buy

Very few bargain shoppers

Many shoppers who buy at the last minute

Business websites visited by many 40-year-old men

Business websites visited by many 20-year-old women

Finally, for each e-commerce material feature, the recommendation engine can determine one or more marketing actions that have potential effects on a merchant website. The decision can be based on heuristic algorithms, empirical research, and the like. The set of exemplary rules is described in the table of the sixth figure.

According to a specific embodiment, the seventh figure depicts that once the subset of marketing actions has been potentially effected by applying the rule database, the steps relating to the ranking of the marketing actions are Step. First, the recommendation engine can determine a quality score for each marketing action that indicates the potential effectiveness of the marketing action for the merchant. The quality score may be determined based on one of the e-commerce materials and the specific marketing action. For example, if the e-commerce profile is "very few reflow customers" and the application is redirected, the quality score can be 90 (0 is the lowest quality score and 100 is the highest quality score). Alternatively, the quality score may also be a function of the actual e-commerce data and may be adjusted from a preset value given for a particular combination of e-commerce material features and marketing actions. In addition, the quality score can be determined by heuristic algorithms, empirical studies, and the like.

The recommendation engine can then generate a preliminary ranking of the marketing actions based on the quality scores and can transmit the preliminary ranking to the third party marketers whose marketing actions have been determined to have potential effects. The recommendation engine can prompt the third party marketers (if any) to bid to increase the ordering of their individual marketing actions. If no bids are accepted, the preliminary sort is the final sort and the sort step ends. Otherwise, if a bid is received, the recommendation engine may generate an updated ranking of the marketing actions based on the quality scores and the bids of the third party advertisers. The update order can be provided to the third party marketers, and the third party marketers can be prompted to change their bids (eg, increase, decrease). If some third-party marketers do decide to change their bids, they will generate an update order as previously described. Otherwise, the sorting process ends.

Additionally, if the recommendation engine has received an explicit marketing request from the merchant, the ranking may additionally be based on how close the suggested marketing action is to the explicit marketing request.

In a specific embodiment, the metric used to rank the marketing actions (in other words, the ranking metric) can be determined based on the product of the quality score and the bid price: Quality score) * (1 + bid price). The product can be calculated for each suggested marketing action, and the marketing actions can be sorted according to the increased product value (from lowest order to highest order). In another embodiment, the ranking metric is determined based on the sum of the quality score and the bid price: quality score + bid price. The summation may be calculated for each suggested marketing action and its corresponding bid, and the marketing actions may be sorted according to the increased sum total value (from lowest order to highest order). (The mathematical equations are provided by way of example, and other mathematical equations can be used.)

As discussed earlier, marketing actions with a quality score below a certain value may not be recommended regardless of the bid price. For the purposes of discussion, it is assumed that the quality score ranges from 0 to 100, with 0 indicating the lowest quality and 100 indicating the highest quality. Further assume that in an implementation, the recommendation engine does not recommend marketing actions with a quality score below 20. The ranking metric of the row pin action (calculated by the product) can be modified as follows: (quality score) * u (quality score -20) * (1 + bid price), where u (x) is a unit step function defined as follows, x When 0 is u(x)=1, and when x<0, u(x)=0. The ranking metric for the one-line pin action (calculated by summing up) can be modified as follows: (quality score + bid price) * u (quality score -20), where u(x) is also the unit step function. (The mathematical equations are provided by way of example, and other mathematical equations can be used.)

According to a specific embodiment, the eighth figure depicts the steps performed by a third party marketer. First, the third party marketer can submit a narrative to each of the marketing actions that the recommendation engine wants to distribute (eg, sell) to the merchant. Once the marketing action of the third party marketer determines that a merchant has potential effects, the third party marketer can be notified of the preliminary ranking of its marketing actions. Next, the third-party marketer decides whether to propose a bid price to increase the ranking of its marketing actions. If you want to improve the sort, the marketer submits a bid price to the recommendation engine, and An update ordering of marketing actions is received from the recommendation engine. Alternatively, the third party marketer can propose the desired sorting position, and the recommendation engine can notify the third party marketer to raise the ranking of the marketing actions to the corresponding fee required for the sorting. If you don't want to improve sorting (that is, the opportunity for the marketer to cancel the sorting of marketing actions), the marketer waits to see if any merchants choose their marketing actions.

If a merchant selects the marketing action of the third party marketer, the third party marketer can provide the marketing action (such as the encoding of the same software application) to the merchant. Finally, the third party marketer can receive the fee for providing the marketing action to the merchant. At certain locations in the program, the third party marketer also pays the entity/organization that is hosting the recommendation engine equal to the total amount of its bid price (not depicted). Although the above description for the sake of brevity only describes the marketing actions of the third party marketer, it should be clear how the program is generalized to handle situations in which the third party marketer attempts to distribute two or more marketing actions.

According to a specific embodiment, the ninth figure depicts the steps performed by a merchant. First, a merchant can decide whether to allow the recommendation engine to analyze e-commerce data related to customer interaction with the merchant website. If the merchant agrees, the e-commerce material is transmitted to the recommendation engine from the merchant website (or more specifically, from a server hosting the merchant website). Otherwise (ie, when the merchant does not agree), the program is terminated. Alternatively, a merchant can provide an explicit marketing request to the recommendation engine. The merchant can then receive from the recommendation engine one or more marketing actions that are expected to improve the merchant's website marketing capabilities. The merchant may select one or more of the suggested marketing actions. The merchant then receives one or more marketing actions (eg, such as the encoding of a software application) from a line of sellers (a third party marketer or a marketer that incorporates the recommendation engine). Or, if the recommendation engine has accessed the database in which the selected marketing action is stored, To receive the (etc.) marketing action from the recommendation engine. The merchant can install the received software on the merchant website to implement the selected marketing action. If a selected marketing action has an associated price, the merchant may additionally transmit a payment amount to the marketer to compensate for the marketer providing the marketing action.

It will be apparent from the foregoing discussion that various embodiments can be implemented with the aid of computer implemented programs or methods (i.e., computer programs or routine functions) or any programmable or dedicated hardware implemented digital logic. The program may be embodied in any computer language, including but not limited to object oriented programming languages, combination languages, markup languages and other similar languages, and may be similar in such things as CORBA (Common Object Request Broker Architecture), ^JavaTM and others. Implemented in an object-oriented environment, or can be implemented on a similar complex programmable logic chip (CPLD), field programmable logic gate array (FPGA), and any other similar programmable logic hardware.

It should also be understood that the detailed description provided by the computer implementation program and symbolic representation of data manipulation in computer memory is actually a better means available to those skilled in the computer science field to other technicians in the field. The most effective way to convey the essence of its work. In all cases, the programs executed by the computer system require physical manipulation of the physical quantities. Such computer-implemented programs are typically presented in electronic or electromagnetic information (eg, bits) for storage (eg, stored on a computer-readable storage medium), transmitted (eg, via a wired or wireless communication link), Combine, compare or operate. It has been proven that these signals are, in principle, considered as bits, values, elements, symbols, keys, numbers, etc., for the general purpose of use. However, it should be understood in the mind that all of these and similar terms are associated with the appropriate amount of entity, and that the terms are merely labels that facilitate the application of these quantities.

Unless specifically stated otherwise, it should be understood, such as processing, estimation, calculation, The words used in determining, displaying, etc. refer to the actions and procedures of a computer system or similar electronic computing device that will be presented by means of physical (electronic) quantities in the computer system register, memory or other storage medium. The information is used and converted into other data in the computer system memory, scratchpad or other storage medium for the same presentation in physical quantities. Particular embodiments may be implemented with the apparatus to perform the operations recited herein. The apparatus may be specially constructed for the desired purpose, or may be suitably programmed, or selectively activated and reconfigured by computer readable instructions stored in or on a computer readable storage medium (such as stored in a soft Any disc form of disc, CD, hard drive, CD-ROMs and magnetic discs, or read-only memory (ROM), random access memory (RAM), erasable programmable read-only memory ( EPROM), electronically smeared programmable read only memory (EEPROM), magnetic or optical card, or any form of media suitable for storing computer readable instructions, but is not limited thereto to perform such operations. Of course, the programs presented herein are not limited to being implemented by computer readable instructions, but may also be implemented in appropriate circuits, such as a specific purpose integrated circuit (ASIC), a programmed field programmable logic. Gate array (FPGA) and more.

It should be understood that the specific embodiments described above are illustrated by way of example only, and the specific embodiments are not limited to Rather, the present invention is to be construed as a combination of the various features and sub-combinations of the various features described herein.

100‧‧‧ computer system

102‧‧‧ busbar

104‧‧‧Processor

106‧‧‧ main memory

108‧‧‧Read-only memory

110‧‧‧Storage device

112‧‧‧ display

114‧‧‧Input device

116‧‧‧ cursor control device

118‧‧‧Communication interface

Claims (20)

A system comprising one or more processors and a non-transitory storage medium, the non-transitory storage medium comprising program logic for execution by the one or more processors, the program logic comprising: a recommendation engine, The following actions may be performed: receiving, from a merchant, e-commerce data relating to a customer interacting with a website of the merchant; identifying one or more features of the e-commerce material; selecting at least in part based on the one or more characteristics identified a marketing action that has a potential effect on the merchant; sorting the selected marketing actions based at least in part on one or more quality scores and one or more bids, wherein the one or more qualities The score and the one or more bids are for each of the selected marketing actions; and a ranked list of the selected marketing actions is provided to the merchant. The system of claim 1 wherein the selection comprises automatic selection. The system of claim 1, wherein the recommendation engine ranks the selected marketing actions based at least in part on a quality score and one or more bids for each action of the selected marketing actions. The system of claim 1, wherein the one or more bids comprise a plurality of bids from a plurality of third party marketers. The system of claim 1 wherein the bids are bids for locations related to marketing actions. The system of claim 1, wherein a quality score indicates that the estimate of the one-line pin action is valid Sex. The system of claim 1, wherein the e-commerce material includes materials accumulated in an operation of the e-commerce store, and includes order materials, website traffic data, and e-commerce store product materials. The system of claim 1, wherein the ranking of a first marketing action is based at least in part on an estimated profit that the merchant can obtain from the first marketing action. The system of claim 1, wherein the identified features include a customer's purchase behavior and demographic attributes of the customers. The system of claim 1, wherein the ranked list of the selected marketing actions is provided to include a suggested marketing action. The system of claim 1, wherein the providing a ranked list of the marketing actions comprises recommending marketing actions based at least in part on the manner of human experts and heuristics retrieved from the one or more rule databases. A method of suggesting a marketing action, the method comprising: receiving, from a merchant, e-commerce data relating to a customer interacting with the merchant-site; identifying one or more features of the e-commerce material; at least in part based on the one or more identified a feature, automatically selecting a plurality of marketing actions, the marketing actions having a potential effect for the merchant; sorting the selected marketing actions based at least in part on one or more quality scores and one or more bids, wherein The one or more quality scores and the one or more bids are for each of the selected marketing actions; and a ranked list of the marketing actions selected for the merchant. The system of claim 12, comprising the recommendation engine for selecting the marketing action based at least in part on one or more of a quality score and one or more bids for each of the selected marketing actions Sort. A system as claimed in claim 12, comprising receiving the one or more bids, wherein the one or more bids comprise a plurality of bids from a plurality of third party marketers. The system of claim 12, comprising receiving the one or more bids, wherein the bids are bids for locations related to marketing actions. The system of claim 12, comprising using a quality score that indicates an estimated validity of the one-line pin action. The system of claim 12, comprising using a quality score for indicating an estimated effectiveness of a first marketing action for a merchant associated with the first marketing action, the estimated validity being based at least in part on a similar merchant The average effectiveness of this marketing action. The system of claim 12, comprising using the e-commerce material, which is data accumulated in an operation of an e-commerce store, and includes order materials, website traffic data, and e-commerce store product materials. The system of claim 12, wherein the ranking of the first marketing action is based at least in part on an estimated profit that the merchant can obtain from the first marketing action. A non-transitory computer readable storage medium or medium tangibly storing computer program logic, the computer program logic being executable by a computer processor, the program logic including a recommendation engine for receiving a customer from the merchant An e-commerce material interacting with a website; identifying one or more characteristics of the e-commerce data; Automatically selecting a plurality of marketing actions based at least in part on the one or more identified features, the marketing actions having potential effects for the merchant; at least in part based on one or more quality scores and one or more bids, Sorting the selected marketing actions, wherein the one or more quality scores and the one or more bids are for each of the selected marketing actions; and providing a ranked list of the selected marketing actions Give the merchant.