Product structure and rule engine of marketing activity platform design

Editor's lead: From the perspective of product operation, the importance of marketing activities is self-evident, so how to promote the growth of activities through platform-based marketing design? Based on the core content of the product, the overall architecture of the platform and the rule engine in the core design are analyzed. Let's have a look! The last article told you the general background of the marketing platform and the problems encountered in the process of product construction. Today, this article mainly talks about the core content of the product, which is divided into two parts: First, the overall architecture of the platform is 1. Product architecture Let's talk about product architecture first. The detailed product architecture diagram is considered to contain company information, so it will not be done for the time being. From the perspective of interactive stratification, the framework of the marketing system is as follows: (1) The presentation layer is mainly the front-end activity page. (2) The interactive layer is mainly active play, such as lottery and answering questions. And interact with the participating users; It also includes touch forms, such as SMS, push, etc. (3) The rule engine at the bottom of the public * * * rule layer includes general logic and conditional strategy library (judging whether a new person is a new person, whether he has participated in an activity, recommending other contents after completing a specific path, etc.). ) and action collection library (click on the lottery to trigger the lottery process, scan the code to trigger, etc. ). (4) Activity rewards at the rights level, such as cash-filled red envelopes, video rights, coupons, etc. 2. Looking at the overall situation and focusing on the core, we all know that each activity chain is a closed loop formed by the target users of upstream activities and the downstream rights prizes. For example, newcomers (target users of activities) are guided to participate in newcomers' courtesy activities through landing pages, and 5 yuan cash-filled red envelopes (equity award) will be issued if they meet the requirements. There are still many rules to deal with, such as the judgment of newcomers' conditions, activity logic, prize distribution interface, docking with existing payment interface, activity data transformation monitoring and so on. The flow of the whole activity chain is very simple, and we know it very well. But before any product is launched, we need to think about its upstream and downstream, whether it constitutes a closed loop and so on. Where is the product made and what problems need to be solved? This is the boundary of the product mentioned in the last article. The core of the marketing activity platform is to quickly complete the activity creation and marketing through the activity engine. As shown in the above figure, the closed-loop path solved by the campaign marketing platform is: creating campaign-configuring campaign rules-selecting delivery channels-monitoring campaign data-monitoring asset consumption-monitoring system performance, and designing campaign templates and systems to feed back campaign monitoring data. (1) The activity center selects corresponding templates according to the activity requirements, such as Jiugongge lottery, sign-in and answer questions. (2) Activity configuration Configure the logic of this activity according to the activity rules, such as lottery activities: lottery number distribution, winning probability, winning probability, whether it is related to the task, etc. (3) Channels mainly include push, landing page, WeChat and H5. (4) Effect insight mainly includes activity data statistics, participant statistics, participant statistics, activity conversion user statistics (funnel chart), prize use conversion statistics, etc. (5) Asset loss monitoring is mainly used to monitor the number of participating users and the number of prizes distributed, and whether there is over-delivery or missing delivery. (6) System monitoring focuses on the concurrent pressure monitoring of system performance, bearing pressure and peak activity. 3. The product expandability defines the core competence of the product, and at the same time, we need to consider the product expandability, which is what we call low coupling and high cohesion. It can be roughly divided into the following two points: (1) the ability to combine upstream and downstream products. As mentioned above, the upstream of the activity is the user group. For activity users, marketing activities should support basic user management, such as user basic information, participation records, award-winning records and so on. As input factors of rules, these information are mainly used in activity judgment logic. Additional functions can support tagged users, and can be used to distribute activity scenes and launch activities for specified user groups. Secondly, considering the products of major customers, 20% of customized services will be retained. Big customers have their own user database, and their user data is more perfect than the user management provided by our own products, such as economic subsystem, big data user center and so on. At this time, we provide a general user interface, through which we can obtain the target user group. Because user data is sensitive, most of them are provided by customers. We need to design interface encryption methods and user storage methods to ensure the data sensitivity requirements of large customers. In fact, it is the ability of product compatibility, and activities are only part of the whole marketing product line. (2) Micro-service module design General products can also be combined into different products through module configuration and provided to different customer groups with different needs. Accordingly, the design requirements of each module are higher, not only product design, but also technical design, which requires low coupling. The product side needs to constantly add and subtract various functional modules, do a good job of product iteration in time, and meet 80% of customer demand in the market in time. The technical side needs to reduce the strong correlation between functions and interfaces in design. Second, the core design-rule engine 1. Why do you want to be a rule engine? Business code often contains a large number of cases, and casebycase is full of judgments and choices on conditions. When these conditions such as if-else/switch are increasing, the code becomes difficult to maintain, and the following problems will occur: the existing business logic cannot be expressed intuitively, and it is difficult for newcomers to start. Add&; Logic is difficult to change and expansion is extremely difficult; Generally, the processing cost is high. Every time you change the logic, you need to go through a complete R&D- test-release-retest-gray scale, which is inefficient and costly. Isolate this inevitable business decision logic, make the logic clear and can be maintained independently. 2. The rule engine defines the abstract business logic judgment process: data stream input = output corresponding results according to rules (logic judgment is treated as a black box) =. The rule engine is a processor, which receives dynamic data inflow and obtains decision results according to internal rules. In order to separate business logic and ensure its independent maintenance and dynamic update. Input: specific values of various conditions such as user id, attribute value and mobile phone number. Output: The result of decision may be bool (logical value, true/false) or concrete value, and these results can be used as a new set of data to make decisions. Rule engine service is usually based on the core rule engine, adding some services, such as facade service at execution time (facade pattern can be used to encapsulate the underlying implementation of the system, hide the complexity of the system, and provide a set of more simple and easy-to-use and higher-level interfaces), visual rule creation, multi-rule engine support, more systematic rule management body, calling logic flow, additional data support and so on. 3. Application Scenarios of Rule Engine Generally speaking, the rule engine abstracts repetitive and standardized business scenarios into simple or responsible logic, and outputs corresponding results by inputting data and judging rules. Common application scenarios: wind control system, power distribution &; Recommendation scenario, fund decision scenario, data label scenario, activity scenario, etc. In this section, we will focus on the application in the activity scene. (1) Draw different people &; Different scenes correspond to different prize pools (different winning probabilities, different prize sets), and common games include roulette, Jiugongge, and smashing golden eggs. (2) Task acceptance rules and task completion indicators can be dynamically configured (different people have different tasks, and indicator conditions can be dynamically configured &; Combination), common gameplay: answering questions, game activities. Tandem Play: Events match user paths. All the events that need to be correlated (connected in series) are matched by the source events. According to the user's participation activities, the events that need to be triggered and the corresponding trigger values are obtained through time filtering and partial dynamic calculation. For example, lottery and task can also be played in series, and the number of lottery times can be obtained after completing the task, which increases the probability of lottery. Example: After the user enters the activity, the task is assigned according to certain rules, the target user participates in the lottery, and the user reaches multiple combined indicators, so that the task is completed after the consumption record is recorded in the current month. After the task is completed, the user will be given a chance to draw a lottery (several times) or a direct reward according to the reward received by the user, and decide whether to send a private message to keep. After the user gets the chance to draw a lottery, he will draw a lottery for the prize pool such as cash, and the winning probability is high, which is calculated according to the characteristics of the user. Ps: There are configurable content rules in it. (3) In general incentive model, different user characteristics correspond to different incentive degrees (different people have different perceptions of rewards in different scenarios, such as prizes for new users and old users). Common gameplay: sign in and punch in, and bargain together. (4) The general reach model distinguishes the copied contents. Common gameplay: answer test, personal year-end report, etc. Knowing the application of the rule engine in the activity scene, we can look at the common activity logic and think about whether we can regularize a certain process. Because products come from life. References:

/p/37 183 12 14