CN116307791A

CN116307791A - Method and system for constructing online operation system

Info

Publication number: CN116307791A
Application number: CN202211107269.0A
Authority: CN
Inventors: 郭然然; 丁树同; 张利朋; 王媛媛
Original assignee: Shandong City Commercial Banks Alliance Co ltd
Current assignee: Shandong City Commercial Banks Alliance Co ltd
Priority date: 2022-09-09
Filing date: 2022-09-09
Publication date: 2023-06-23

Abstract

The invention provides a method and a system for constructing an online operation system, wherein the method comprises the steps of determining a quantization index; the quantization index is a data index of the service in each stage based on the AARRR model; setting an equipment ID for executing the service as a user identifier, configuring a full buried point type to be opened when a client accesses, recording an activation event and an end user behavior event under the premise of recording a tracking event activation after successful login, and adding a custom attribute for the end user behavior event under the premise of event activation; and outputting a log file of the final user behavior event, analyzing the log file to obtain a value of the quantization index, and optimizing the business process after obtaining the value of the quantization index. Based on the method, an online operation system construction system is also provided. According to the invention, different guest groups are subdivided according to different user behaviors, a detail-level user list of a specific behavior is obtained, and real-time monitoring, analysis and touch of guest groups of different levels are satisfied.

Description

Method and system for constructing online operation system

Technical Field

The invention belongs to the technical field of big data analysis in the financial industry, and particularly relates to an online operation system construction method and system.

Background

The prior banks, especially urban businesses, currently perform limited-scale motion marketing activities on the basis of the basic image of the client, which is constructed based on the basic attributes, the asset information and the like of the traditional user, lack accumulation of behavior data, and expose the problems of low efficiency, few strategies and difficult decision in the actual online operation process, so that the actual demands of the user can not be effectively and timely observed. In the aspect of data analysis, the key activities, the blocking points of products, insufficient insights of leakage reasons and insufficient effective effects of activity conversion effects are diagnosed, and the layering grouping operation capacity based on the user behavior insights is not formed yet; in the aspect of operation closed loop, at present, key links and nodes are scattered in each system, and records of full-journey behaviors of users, such as marketing activities, are lacked, so that the planning of activities is realized at present, but response feedback tracking of users after the activities are carried out, effect evaluation of activity input-output ratio is lost, and the 'wool-like' effect of the activities is higher than actual benefit conversion expectation, so that activity cost is high, and loss of the users is brought due to insufficient behavior insight of the users; in the aspect of operation strategies, active batch and single-line service strategies are taken as the main, the priority, flexibility and timeliness are poor, decisions are taken as the main, and an intelligent high-efficiency decision mechanism is lacked.

The disadvantages of the prior art include: firstly, due to the lack of insight analysis on user behavior data, user portraits are incomplete, behavior data are lost, the user portraits are incomplete, the actual demands of users cannot be timely, accurately and effectively insight, and cross-terminal and cross-channel user behavior data are not opened, so that a data island is formed; the front-end behavior data, the back-end business data and the third-party data are not communicated, and the accuracy and the data quality of the data analysis model are required to be improved. Secondly, the strategy lacks systemization, and the layering grouping operation capability based on user behavior insight is weak, so that the communication response rate is low, and the operation cost is high. The existing operation strategies are mainly active batch operation strategies, and mainly single-line service operation strategies, the strategies are not systematic and intelligent, the communication response rate is low, the priority is difficult to measure, the decision is not intelligent, and the operation cost is high.

Disclosure of Invention

In order to solve the technical problems, the invention provides an on-line operation system construction method and system, and an output systemization strategy scheme, which can meet the differentiation requirements of different service scenes.

In order to achieve the above purpose, the present invention adopts the following technical scheme:

an on-line operation system construction method comprises the following steps:

determining a quantization index affecting on-line operation; the quantization index is a data index system of the business in each stage based on an AARRR model;

setting an equipment ID for executing the service as a user identifier, configuring a full buried point type to be opened when a client accesses, and setting event attributes of all events; after the user logs in successfully, recording an activating event and an end user behavior event under the premise of tracking event activation, and adding a custom attribute for the end user behavior event under the premise of event activation;

and outputting a log file of the end user behavior event, and analyzing the log file to obtain the value of the quantitative index.

Further, the method further comprises the step of disassembling the business processes in combination with the full life cycle path of the user after obtaining the value of the quantization index, and optimizing each process.

Further, the data index system of the service in each stage based on the AARRR model specifically comprises:

a first data indicator at the acquisition stage; the first data index comprises the number of newly-increased users and the registration conversion rate of the newly-increased users;

the second data index in the activation stage comprises the number of newly-increased binding card users, the number of active users in a preset time period, average single-use duration and DAU/MAU;

a third data indicator at the retention stage, the third data indicator comprising a retention rate and a churn;

a fourth data index in the value stage, wherein the fourth data index comprises a successful financial purchasing user number, a successful financial purchasing conversion rate, a financial expiration loss rate, a successful deposit purchasing user number, a deposit successful purchasing conversion rate, a deposit expiration loss rate, a loan application successful user number and a loan application successful conversion rate;

a fifth data indicator at the propagation stage; the fifth data indicator comprises a K factor; the K factor may bring up how many new users a user initiating a recommendation may be.

Further, the process of setting the device ID of the executing service as the user identifier includes:

after the software development kit is initialized, generating a device ID as a user identifier;

after the user registration and login are successful, if the login ID can be obtained, the client actively calls the login ID; and when the user logs off, continuing to use the equipment ID as a user identification or reinitializing the equipment ID.

Further, the setting the event attribute of all events further includes: the event attribute is set to a common attribute.

Further, the method for opening the fully buried point type comprises the following steps: configuring a fully buried point type through a setAutoTrackEventType () method;

the method for recording the activation event comprises the following steps: recording an activation event by calling a trackAppInstalll ();

the method for tracking the event comprises the following steps: user behavior events are tracked through the track () method.

Further, the outputting the log file of the end user behavior event further includes obtaining by keyword screening the log file: event data output by the software development tool when the embedded point event trigger is successful, error reasons output by the software development tool when the embedded point event trigger is failed, effective events output by the software development tool when the event data report is successful, and ineffective events and error reasons output by the software development tool when the event data report is failed.

Further, the process of optimizing each flow is as follows:

formulating core indexes, wherein the core indexes comprise an index MAU of a full life cycle of a user and an asset management scale of a business line;

carrying out association disassembly business processes on the full life cycle path of the user and the core index; and (5) formulating an operation target to optimize each flow.

The invention also provides an on-line operation system construction system which comprises a determination index module, a configuration record module and an output analysis module;

the determining index module is used for determining a quantization index affecting on-line operation; the quantization index is a data index system of the business in each stage based on an AARRR model;

the configuration recording module is used for setting the equipment ID of the execution service as a user identifier, configuring the type of the full buried point to be opened when the client is accessed, and setting the event attribute of all events; after the user logs in successfully, recording an activating event and an end user behavior event under the premise of tracking event activation, and adding a custom attribute for the end user behavior event under the premise of event activation;

the output analysis module is used for outputting a log file of the end user behavior event and analyzing the log file to obtain the value of the quantitative index.

Further, the system also comprises an optimization module; the optimizing module is used for combining the full life cycle path dismantling business processes of the user after obtaining the value of the quantization index, and optimizing each process.

The effects provided in the summary of the invention are merely effects of embodiments, not all effects of the invention, and one of the above technical solutions has the following advantages or beneficial effects:

the invention provides a method and a system for constructing an online operation system, wherein the method comprises the steps of determining a quantization index affecting online operation; the quantization index is a data index system of the business in each stage based on the AARRR model; setting an equipment ID for executing the service as a user identifier, configuring a full buried point type to be opened when a client accesses, and setting event attributes of all events; after the user logs in successfully, recording an activating event and an end user behavior event under the premise of tracking event activation, and adding a custom attribute for the end user behavior event under the premise of event activation; and outputting a log file of the end user behavior event, and analyzing the log file to obtain the value of the quantitative index. And after obtaining the value of the quantization index, combining the full life cycle path disassembly business processes of the user, and optimizing each process. Based on an on-line operation system construction method, an on-line operation system construction system is also provided. The invention outputs the systematic strategy scheme, and can meet the differentiation requirements of different service scenes. Different guest groups can be subdivided specifically according to different user behaviors, a detail-level user list of a specific behavior is obtained, and real-time monitoring, analysis and access of guest groups of different levels are met.

The invention makes policy formulation for the clients with and without financial holding, and makes popularization and strengthening purchase guidance for the clients with financial holding, thereby improving the conversion rate of producing the slices; the method aims at the promotion of high conversion products and active service of non-financial and warehouse-holding customers, so that the customers are greatly increased, sales are greatly increased, and the conversion of the non-financial and warehouse-holding customers is improved.

Drawings

Fig. 1 is a flowchart of an on-line operation system construction method according to embodiment 1 of the present invention;

FIG. 2 is a schematic illustration showing the relationship between preset events and preset attributes according to embodiment 1 of the present invention;

FIG. 3 is a schematic diagram of the core index formulation according to embodiment 1 of the present invention;

FIG. 4 is a diagram of a full life cycle path of a user according to embodiment 1 of the present invention;

fig. 5 is a flowchart illustrating a full life cycle path splitting business process combined with a user according to embodiment 1 of the present invention;

fig. 6 is a schematic diagram of an on-line operation system construction system according to embodiment 2 of the present invention.

Detailed Description

In order to clearly illustrate the technical features of the present solution, the present invention will be described in detail below with reference to the following detailed description and the accompanying drawings. The following disclosure provides many different embodiments, or examples, for implementing different structures of the invention. In order to simplify the present disclosure, components and arrangements of specific examples are described below. Furthermore, the present invention may repeat reference numerals and/or letters in the various examples. This repetition is for the purpose of simplicity and clarity and does not in itself dictate a relationship between the various embodiments and/or configurations discussed. It should be noted that the components illustrated in the figures are not necessarily drawn to scale. Descriptions of well-known components and processing techniques and processes are omitted so as to not unnecessarily obscure the present invention.

Example 1

The embodiment 1 of the invention provides an online operation system construction method, which is based on an OSM model and an AARRR model, and according to banking business purposes, the method surrounds business targets or iterative strategies, determines quantitative indexes, and according to the indexes, utilizes the AARRR model, disassembles the whole life cycle journey of a user through behavior insight of the user, combs associated business strategies, constructs an index system, sets early warning values of each index for each link, develops channel conversion funnel analysis, event analysis and loss reason analysis, gives out the optimal strategy and optimal touch mode of each different link through perception of the user and behavior thereof, and directly takes the user as timely and accurate operation action based on decision.

Wherein the OSM model is a business analysis framework. The AARRR model core is a funnel model, which respectively represents five stages of the full life cycle of a user: acquisition, activation, persistence, benefits, recommendations. Two core points of the AARRR model: firstly, taking a user as a center and taking a complete life cycle of the user as a clue; secondly, the cost/income relationship of the whole control product is that the life cycle value (LTV) of the user is far greater than the sum of the acquisition cost (CAC) of the user and the operating cost (COC) of the user, which means the success of the product.

Fig. 1 is a flowchart of an on-line operation system construction method according to embodiment 1 of the present invention.

In step S100, determining a quantization index affecting on-line operation; the quantization index is a data index system of the business at each stage based on the AARRR model.

The AARRR model corresponding to the mobile banking APP comprises five stages:

a guest acquisition stage: this is the first step of online operation, and users are obtained through various popularization channels and various means, so that the investment strategy is more reasonably determined, and the acquisition user cost (CAC) is minimized. The indexes that need to be focused on at this stage include: the number of newly increased users and the conversion rate of newly increased user registration.

Wherein the number of newly added users is defined as: the number of devices that install and first launch an application in a time frame.

The newly added number of users is used for representing: the user share of channel contribution, analysis for next step confirm put strategy, whether there are a lot of junk users.

The newly added user registration conversion rate is defined as the number of successful users registered/the number of users to launch/download the application.

The newly added user registration conversion rate is used for reflecting the conversion effect of the activity; analyzing a registered leakage link; and registering transformation analysis.

An activation stage: the newly added user is converted into an active user through precipitation. At this time, attention needs to be paid to the number of active users, liveness, online use time length, use frequency and the like. The indicators that need to be of interest at this stage include: the number of newly added binding card users, the number of daily active users, the number of Zhou Huoyue users, the number of monthly active users, the average single use duration and the DAU/MAU.

The number of the newly added binding card users is defined as: and the number of users successfully binding the line card/class II/III accounts in the newly increased users.

The number of newly added binding card users is used for reflecting: analyzing the core user scale and the binding card conversion rate of the line;

the daily active user number is defined as the user number for starting the application today, and the calculation is not repeated after multiple times of starting;

the daily active user number is used for representing: core user scale and binding card flow leakage link analysis; decomposing active users; user activity rate;

zhou Huoyue the number of users is defined as: the number of users accessing the application through each access mode (Mobile App/Mobile H5/PC Web/applet) in the last week daily (including 7 days today) in the time range, and the application is not repeatedly metered for multiple accesses.

Zhou Huoyue the number of users is used to embody: periodic user scale; the periodical change trend is mainly the comparison of a popularization period and a non-popularization period;

the number of monthly active users is defined as: the number of users accessing the application through each access mode (Mobile App/Mobile H5/PC Web/applet) in the last month daily (30 days containing today) in the time range, and the application can not be repeatedly metered for multiple accesses.

The number of monthly active users is used for representing: user scale stability; evaluating popularization effect; overall user size varies.

The average single use time period is defined as the time period of one start use.

Average single use time length is used to represent: analyzing the viscosity of APP; observing the average use time length of different time dimensions to know habits of different user groups; and (5) reserving the basis of the loss analysis.

The DAU/MAU is defined as: it can be seen from the DAU/MAU what the average number of days the user has access to the App per month is, such as: an App has 50-100-thousand DAUs with a DAU/MAU ratio of 0.5, i.e. the average monthly access time of the user is 30 x 0.5 = 15 days. This is a relatively important indicator for evaluating user viscosity.

The higher the DAU/MAU value, the more viscous the App, indicating that more users are willing to use the App; otherwise, if the value of DAU/MAU is low, multidimensional analysis is required to be performed by combining multiple conditions of product attributes, time considerations (working day/holiday, etc.), version update, operation activities, ARPU values of user dimensions, and the like.

And (3) a retention stage: after a period of time, the newly added user still continues to use the application and is regarded as a retention user, and the proportion of the newly added user in the part of the users is the retention rate. The method is used for analyzing the participation or activity of the user and mainly examining how many people can perform subsequent behaviors in the user after performing initial behaviors.

The examination objects are divided into 'newly added visitor' and 'active visitor', and the examination objects support the retention rate of different dimensions of the query (including (equipment attribute, event attribute, page parameter, user group, single event and event group of the query embedded point after the event embedded point).

The retention rate includes: +n daily retention, +n Zhou Liucun, +n month retention. The proportion of the number of newly increased users per day+N days/N weeks/N months of registered users per newly increased users.

The retention was used to represent: APP quality assessment; user quality assessment; user scale measurement;

loss: and in the statistical time interval, the user leaves the APP in different periods.

Value stage: defining a life cycle value (LTV), the value a user creates over the life cycle, breaking down into APPs, i.e. the retention of the user AUM (asset size).

The value phase includes: successful purchase user number of financing, successful purchase conversion rate of financing, expiration rate, successful purchase user number of deposit, successful purchase conversion rate of deposit, expiration rate of deposit, successful user number of loan application, and successful conversion rate of loan application

The value stage is used for reflecting the contribution period of the user income, the leakage link of the key product and the contribution condition of the core user.

Propagation phase definition: users share the APP or functions/products within the APP to other users. Among these are the important index K factors. The main calculation mode of the K factor is as follows:

k= (number of invitations each user sent to his friends) × (conversion rate of the person receiving the invitation to a new user).

Suppose that on average each user will send out invitations to 20 friends, with an average conversion of 10%, k=20×10% =2.

When K >1, the user group increases like a snowball.

When K <1, the user group stops growing by self-propagation to a certain scale.

In step S200, setting a device ID for executing a service as a user identifier, configuring a full buried point type to be opened when a client accesses, and setting event attributes of all events;

an Event model (Event model) is adopted as a basic data model. The Event model contains two core entities, event and User, to record all the User's behavior, which is the core basis for the subsequent analysis of all the interface and functional designs.

And generating a unique identifier for the user according to a certain rule based on the distict_id. The method uses a scheme of associating equipment ID and login ID, and user identification and tracking of clients

The client access implementation method comprises the following steps:

client access refers to burying points by using an iOS/Android/JavaScript and other SDKs, and the specific call flow is as follows:

after the initialization of the SDK is completed, a device ID is automatically generated as a user identification.

2) The client actively invokes a login interface when the user registers successfully, logs in successfully, initializes the SDK (if a login ID can be obtained).

3) When the user logs off, tracking continues using the previous user identification. If there is no special case, it is generally recommended to choose this approach. For JavaScript SDKs, a logo (true) method may also be invoked that reinitializes the device ID in addition to clearing the login ID.

Preset events and preset attributes. An example of a relationship between preset events and preset attributes is given in fig. 2. The preset events protected in the present invention are not limited to the attributes listed in fig. 2, and other events to be set and event attributes may be set according to the needs.

The fully buried point type is configured by a setAutoTrackEventType () method. The attribute that needs to be added for all events is registered as a common attribute through the register SuperProperties () after initializing the SDK.

In step S300, after the user logs in successfully, recording an activation event and an end user behavior event under the premise of recording a tracking event activation, and adding a custom attribute for the end user behavior event under the premise of event activation;

when the user registration is successful or login is successful, calling a logic () method of the SDK;

after the SDK is initialized, tracking user behavior events through a track () method, and adding custom attributes for the events:

after initializing the SDK, opening a log output function of the SDK by the following method to debug the user behavior event;

in S400, a log file of the end user behavior event is output, and the log file is analyzed to obtain the value of the quantization index.

Screening SA. Keywords in Logcat:

when the trigger of the buried point event is successful, the SDK outputs event data at the beginning of the track event

When the trigger of the embedded point event fails, the SDK outputs the corresponding error reason

When the event data is successfully reported, the SDK outputs event data at the beginning of the valid message field

When the event data reporting fails, the SDK outputs event data at the beginning of an invalid message field and outputs an error reason.

The method comprises the following specific steps of introducing back-end data into a data analysis platform in real time by using Logagent:

first, download log agent deployment package and decompress:

second, checking the running environment;

thirdly, editing the configuration file;

fourth, checking the configuration file;

fifthly, starting the LogAgent;

sixth, verify data using LogAgent

The data format, field type are checked using the LogAgent's DebugSender. The method comprises the following specific steps:

the LogAgent locally verifies whether the data format is legal, such as JSON format, whether the necessary fields (disttinct_id, etc.) exist, etc.

And sending the data passing the internal verification to a remote end for verification of the data content, such as field type, track_sign and the like.

And performing data quality check on the acquired data by two modes of buried point management, importing and real-time checking.

For buried point management, several scenarios for data verification are as follows:

and if the meta event has a corresponding event name or attribute name and accords with the data acquisition format, the meta event is normally displayed.

The event is not created, and event data which does not exist in the meta event is reported;

the code receives the error report returned by the service: can not create undefined event, 'xxxx' with normal token.

If the custom event is created in the meta event and the $character is added, the event time error is reported.

If the original system reserved field is included in the created attribute name in the metadata, the attribute is created by error.

The reported data contains attribute names which do not exist in meta-event, and the code receives the report error returned by the service: won't create undefined properties: { "xxx": "xxx" }.

For each piece of display data, the data fields are more intuitively and conveniently seen through a formatting method.

In step S500, after obtaining the value of the quantization index, the method further includes disassembling the business processes in conjunction with the full life cycle path of the user, and optimizing each process.

And adopting an APP bus and service line 1+N mode to design and comb an index system, and disassembling the index according to a service target.

The core index is formulated, the core index of the whole life cycle of the APP user is not MAU, the core index of the service line is AUM, as shown in fig. 3, which is a schematic diagram of the formulation of the core index in embodiment 1 of the present invention.

Around the core index, the full life cycle journey of the user is combed and associated with the core index, as shown in fig. 4, which is a schematic diagram of the full life cycle path of the user in embodiment 1 of the invention; fig. 5 is a flowchart illustrating a full life cycle path splitting business process of combining users according to embodiment 1 of the present invention.

Formulating operation targets

First stage, user acquisition stage: the main aim of the stage is to promote the user to download and register the mobile phone APP and log in;

a second stage, user activation stage: the main aim of this stage is to facilitate the user to bind the card;

third stage, user retention stage: the main aim of this stage is to increase the user's viscosity, promoting the user's next month's retention;

fourth stage, user value stage: the main aim of the stage is to promote the user to reach the business line and trade in the business line, and to promote the conversion rate of the business line;

fifth stage, user propagation sharing stage: the main aim of this stage is to promote the user's willingness to share the mobile phone APP.

Based on the index system in the step S100, a data billboard is built, wherein the data billboard comprises large disc data, a user structure, service lines, activity effects, product performance monitoring and functions.

Large disk data: service personnel pay attention to core service indexes of new users, including the number of newly increased users, the number of active users, the use duration, the retention rate, the access of active pages, the number of registered page access people/times and the like, so as to meet the daily data monitoring requirement;

the user structure: knowing the structure of the new user, including: age, asset, etc.;

transformation analysis: diagnosing the conversion rate of the registration process, insight the breakpoint of the user, and evaluating the registration process;

the activity effect is as follows: activity flow conversion and effect evaluation, and promotion of activity strategy iteration and optimization;

product performance monitoring and functional diagnosis: and monitoring register crash, loading failure and core business process conversion, which are used for optimizing and increasing similar functions of the guiding small assistant and improving conversion rate.

The user observation comprises positioning a long time consuming link based on the service handling time index, evaluating whether the service handling time is in a reasonable range, and positioning a key link of the time consuming.

The whole process is time-consuming: and comparing the business handling time consumption and the historical time consumption average value aiming at the insight analysis of the method, and judging whether the business time consumption is normal or not.

The key links are time-consuming: and particularly monitoring the time consumption of each link, positioning the service link with longer time consumption, positioning the node with longer conversion time according to each subdivision dimension, and performing diagnosis analysis.

The idea is as follows: screening card holding users which are not registered in a mobile phone bank and touch users which have opened APP from the stock retail users to wake up, and attracting the users to download APP and register and log in through rights and interests; and in addition, users are attracted through a micro-signal public number and other flow platforms.

And (3) making a strategy:

guest group 1: a counter card opening user; age of screening (judging its operational ability); asset condition (judging its value); guest group 2: a card-free user opens the APP once, but does not log in; public domain traffic subscribers.

Boot behavior: for the first stage, the H5 page can guide the registration to obtain the mobile phone number for the external flow user by registering gift activities;

triggering time: touching the users meeting the conditions for 3 rounds in batches;

touch the passageway: for the user in the stage one, short messages and telephones are taken as main materials, and offline friend circles, other channel flows and the like; for the users in the second, third, fourth and fifth stages, short messages, telephones and PUSH can be used;

and (3) Offer design: for the stage one user, mainly guiding download registration at present, the core document is developed around the highest authority, such as 30-element telephone fee (with the highest rewards highlighted), and a download page link is attached, and the link is directly connected with a download link.

For the users in the second stage, the result page after successful registration is linked with an activity page, the activity of binding cards can be further promoted directly after successful registration, after successful activity participation, the activity/interest of promoting the guiding value is increased on the activity result page, the users participating in the activity are required to be clustered, the structure of the users is analyzed through clustering, the layering subdivision of the users is further carried out according to different user structures, different interests are matched for different subdivision users, after successful rights acquisition, the sharing link is increased, the sharing propagation value is promoted, and the sharing rate of the fifth stage is promoted.

Effect analysis and iteration strategy:

the first stage, the user acquisition stage, i.e. registration and login stage

Analysis insight: the newly increased user number registered with gift activity guiding conversion is 10507, the registration conversion rate is 56%, and the funnel analysis of the key conversion links of the activity shows that the number of missing links of the client for inputting the short message verification code is more.

Improvement: the short message verification code is directly popped up on the registration page, and the input time of the verification code is prolonged from 60 seconds to 120 seconds.

Password login:

analysis insight: hundreds of thousands of users run off, the password login failure is a main reason, and the quick login conversion rate is insufficient.

Improvement: and (3) adding a popup window and an artificial customer service entrance, optimizing and quickly logging in, highlighting the popup window and strengthening guidance.

WeChat login:

analysis insight: the conversion rate is close to 100%, and the conversion rate is higher.

Improvement: optimizing a login prompt button text and an icon, guiding a user to select a WeChat login mode, and guiding the user to click;

face/fingerprint login:

analysis insight: the conversion rate is higher, but the function is not good;

improvement: error correction [ network busy ] and [ request timeout ];

the second stage, user activation, namely a card binding stage;

analysis insight: the users are few, the conversion is few, the number of the users in the class II/III is less than thousands, and the conversion rate of primary user account opening is less than three;

improvement: optimizing the [ unbinding card ] button to a function prompt page commonly used by a user, such as [ My ] and [ financial resources ], adding a [ unbinding card ] lifting frame when clicking any one of the channel pages, checking whether an account exists when clicking the [ unbinding card ], and directly jumping to the binding card page for the existing account.

Third stage, user retention stage

Analysis insight: the number of the newly increased users and the retention rate are lower, especially the month retention rate is lower.

Improvement: for the crowd, the viscous activity, the high-benefit financial accounting and the solicitation of the targeted rights are increased, the response conditions of the users are monitored by isolating time, such as 2 hours, 1 day and 1 week, and for the users which do not respond, the solicitation can be performed by intervening a remote bank/telephone.

Fourth stage, user value stage

Analysis insight: the newly added users have fewer users with account and business line transactions within one month, and the conversion rate of the business lines is lower.

Improvement: for two strategies, no account is generated within one month and 100 days, aiming at user structure analysis, targeted benefit matching is carried out, in analysis insights, more generation users in the newly added users are found, aiming at the passenger group, targeted generation and generation exclusive financial management, exclusive large deposit and exclusive activities are carried out, and on-site activity introduction and offer are carried out.

Secondly, each link of financial, loan and deposit key business lines is disassembled, the conversion rate of each link is analyzed, and the conversion rate in the face recognition and risk assessment links is found to be lower.

In contrast, when the face recognition is not passed twice, the intervention of manual customer service of a remote bank is added, and the manual pass is performed; in the risk assessment link, keywords are highlighted, and the assessment link is simplified.

Fifth stage, user propagation sharing stage

Analysis insight: the user is found to have higher number of active sharing users and sharing rate of transfer and sharing task type tasks, and the sharing transmission number and sharing rate of other service lines are lower.

Improvement: after the financial product is purchased successfully, the rights and interests guidance of the shared gift is added on the successful result page, and after the sharing is successful, the rights and interests can be directly acquired. For loans, the guidance of sharing the good gift is added on the result page after the loan application is successful, and rights and interests can be directly acquired after the sharing is successful.

The method for constructing the online operation system provided by the embodiment 1 of the invention outputs a systemized strategy scheme, so that the differentiated requirements of different service scenes can be met. Different guest groups can be subdivided specifically according to different user behaviors, a detail-level user list of a specific behavior is obtained, and real-time monitoring, analysis and access of guest groups of different levels are met.

According to the method for constructing the online operation system, provided by the embodiment 1, the policy formulation is carried out for the clients with the financial accounting and the clients without the financial accounting, the promotion and the reinforced purchase guidance are carried out for the orientations of the clients with the financial accounting and the clients with the financial accounting, and the conversion rate of the produced sheets is improved; the method aims at the promotion of high conversion products and active service of non-financial and warehouse-holding customers, so that the customers are greatly increased, sales are greatly increased, and the conversion of the non-financial and warehouse-holding customers is improved.

Example 2

Based on the method for constructing an on-line operation system provided in embodiment 1 of the present invention, embodiment 2 of the present invention provides an on-line operation system construction system, as shown in fig. 6, which is a schematic diagram of an on-line operation system construction system in embodiment 2 of the present invention, where the system includes a determining index module, a configuration recording module and an output analysis module;

the determining index module is used for determining a quantization index affecting on-line operation; the quantization index is a data index system of the business in each stage based on the AARRR model;

the output analysis module is used for outputting a log file of the end user behavior event and analyzing the log file to obtain the value of the quantization index.

The system also comprises an optimization module; and the optimization module is used for carrying out optimization on each process by combining the full life cycle path disassembly business process of the user after obtaining the value of the quantization index.

In the determining index module, the data index system of the business in each stage based on the AARRR model specifically comprises:

The process of setting the device ID of the execution service as the user identifier in the configuration record module comprises the following steps:

Setting event attributes of all events further includes: the event attribute is set to a common attribute.

The method for opening the fully buried point type comprises the following steps: configuring a fully buried point type through a setAutoTrackEventType () method;

The output analysis module further comprises: the method comprises the following steps of screening keywords from log files to obtain: event data output by the software development tool when the embedded point event trigger is successful, error reasons output by the software development tool when the embedded point event trigger is failed, effective events output by the software development tool when the event data report is successful, and ineffective events and error reasons output by the software development tool when the event data report is failed.

The process executed by the optimization module comprises the following steps: formulating core indexes, wherein the core indexes comprise indexes MAU of the full life cycle of a user and asset management scale of a business line;

The description of the relevant parts in the system for constructing an online operation system provided in the embodiment of the present application may refer to the detailed description of the corresponding parts in the method for constructing an online operation system provided in embodiment 1 of the present application, which is not repeated herein.

It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements is inherent to. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element. In addition, the parts of the above technical solutions provided in the embodiments of the present application, which are consistent with the implementation principles of the corresponding technical solutions in the prior art, are not described in detail, so that redundant descriptions are avoided.

While the specific embodiments of the present invention have been described above with reference to the drawings, the scope of the present invention is not limited thereto. Other modifications and variations to the present invention will be apparent to those of skill in the art upon review of the foregoing description. It is not necessary here nor is it exhaustive of all embodiments. On the basis of the technical scheme of the invention, various modifications or variations which can be made by the person skilled in the art without the need of creative efforts are still within the protection scope of the invention.

Claims

1. The method for constructing the online operation system is characterized by comprising the following steps of:

2. The method according to claim 1, further comprising, after obtaining the value of the quantization index, further comprising dismantling the business processes in conjunction with the full life cycle path of the user, and optimizing each process.

3. The method for constructing an online operation system according to claim 1, wherein the data index system of the service at each stage based on the AARRR model specifically comprises:

4. The method for constructing an on-line operation system according to claim 1, wherein the process of setting the device ID of the executing service as the user identifier comprises:

5. The method for constructing an online operation architecture according to claim 1, wherein the setting event attributes of all events further comprises: the event attribute is set to a common attribute.

6. The method for constructing an on-line operation system according to claim 1, wherein the method for opening the fully buried point type is as follows: configuring a fully buried point type through a setAutoTrackEventType () method;

7. The method for constructing an online operation system according to claim 1, wherein the step of outputting the log file of the end user behavior event further comprises the step of performing keyword screening on the log file to obtain: event data output by the software development tool when the embedded point event trigger is successful, error reasons output by the software development tool when the embedded point event trigger is failed, effective events output by the software development tool when the event data report is successful, and ineffective events and error reasons output by the software development tool when the event data report is failed.

8. The method for constructing an online operation system according to claim 2, wherein the process of optimizing each flow is:

9. The system for constructing the online operation system is characterized by comprising a determination index module, a configuration record module and an output analysis module;

10. An on-line operations architecture building system according to claim 9, further comprising an optimization module; the optimizing module is used for combining the full life cycle path dismantling business processes of the user after obtaining the value of the quantization index, and optimizing each process.