Marketing strategy acquisition method and device, electronic equipment and storage medium

Abstract

The invention provides a marketing strategy obtaining method and device, electronic equipment and a storage medium, and relates to the technical field of product marketing, and the method comprises the steps: comparing first prediction values of conversion indexes of different candidate marketing strategies under the condition of excluding the influence of a historical marketing strategy; the method comprises the following steps: obtaining a candidate marketing strategy, namely an optimal candidate marketing strategy, with the best individual effect, and then determining the degree of influence of a historical marketing strategy on the marketing effect of the optimal candidate marketing strategy by utilizing a conversion index predicted value under two conditions of considering the influence of the historical marketing strategy and excluding the influence of the historical marketing strategy. Therefore, the final possible marketing effect is attributed, namely the contribution degree of the historical marketing strategy and the optimal candidate influence strategy to the final marketing effect is determined. The final marketing strategy is formulated by referring to the attribution result, it is guaranteed that the to-be-executed marketing strategy has a good marketing effect, and the marketing efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of marketing technology, and in particular to a method, device, electronic device and storage medium for acquiring a marketing strategy. Background Art

[0002] In the new Internet consumer finance sector, user marketing is an important part of the business operation process. Marketing efficiency optimization is of great significance. By improving user marketing efficiency, we can better meet customer needs, improve marketing conversion rate and customer satisfaction. For the target customer group, quickly determining the optimal marketing strategy setting plan is a feasible direction for marketing efficiency optimization, which can shorten the marketing plan formulation cycle and reduce or avoid the cost increase and resource waste caused by blind trial and error.

[0003] Optimizing marketing strategy setting plans is a systematic task. The commonly used technology is based on a priori judgment combined with comparative experiments. The target customer group is divided into multiple experimental groups, and then different preset marketing strategies or strategy combinations are executed on the users of each experimental group. The effect indicator results of the marketing plan within a certain period of execution are observed, and finally the best marketing strategy setting plan is selected based on factors such as cost.

[0004] However, the reliability of comparative experiments is also limited. For example, some users in the target audience may be affected by marketing strategies outside the experiment during the experiment, making it difficult to determine which marketing strategy produced the marketing effect on these users, resulting in distorted experimental results. Summary of the invention

[0005] Based on this, it is necessary to provide a method, device, electronic device and storage medium for obtaining a marketing strategy in response to the above technical problems, so as to attribute the marketing effects, determine the best marketing strategy setting plan based on the attribution results, and improve marketing efficiency.

[0006] A method for acquiring a marketing strategy, comprising: acquiring first user information and second user information of each target user in a target customer group, wherein the first user information is user information without the influence of a historical marketing strategy, and the second user information is user information under the influence of the historical marketing strategy, and the first user information and the second user information respectively include behavioral parameters of the target user; inputting the first user information and information of a candidate marketing strategy into a prediction model corresponding to a conversion indicator to obtain a first prediction value of the conversion indicator, wherein the conversion indicator is used to characterize the degree of numerical change of the behavioral parameter after the candidate marketing strategy is executed; based on the first prediction value of the conversion indicator corresponding to the candidate marketing strategy, selecting an optimal candidate marketing strategy from a plurality of the candidate marketing strategies; inputting the second user information and information of the optimal candidate marketing strategy into the prediction model to obtain a second prediction value of the conversion indicator; based on the first prediction value and the second prediction value, predicting the marketing effect of the optimal candidate marketing strategy and the degree of influence of the historical marketing strategy on the marketing effect; and determining a target marketing strategy to be executed with reference to the marketing effect and the degree of influence.

[0007] In an embodiment of the present application, the information of the candidate marketing strategies includes marketing moments, marketing reach channels and marketing rights. Before inputting the first user information and the information of the candidate marketing strategies into the prediction model corresponding to the conversion indicator, the method also includes: constructing different candidate marketing strategies by adjusting the parameter configuration of at least one of the marketing moments, the marketing reach channels and the marketing rights to obtain different candidate marketing strategies.

[0008] In an embodiment of the present application, the obtaining of the first user information and the second user information of each target user in the target customer group includes: obtaining the user information of the target user within a preset time period before the marketing moment of the historical marketing strategy as the first user information; and obtaining the user information of the target user within a preset time period after the marketing moment of the historical marketing strategy as the second user information.

[0009] In an embodiment of the present application, the first user information and the second user information further include marketing information, respectively, and the step of obtaining the first user information and the second user information of each target user in the target customer group includes: obtaining the user information of the target user within a preset time period after the marketing moment of the historical marketing strategy as the second user information; setting the time series data set of each parameter under the marketing information in the second user information to an empty set, and setting the statistical data of each parameter to zero, to obtain the first user information.

[0010] In an embodiment of the present application, the prediction model is trained in the following manner: historical marketing records are obtained under the target marketing scenario, the historical marketing records include the executed marketing strategies, the marketing moments when the marketing strategies are started, the users corresponding to the marketing strategies, the user information of the users within a preset time after the marketing moments, and the user information of the users within a preset time before the marketing moments; a training sample set of the conversion indicators is constructed using the historical marketing records, and the prediction model is trained based on the training sample set of the conversion indicators.

[0011] In an embodiment of the present application, predicting the marketing effect of the optimal candidate marketing strategy and the degree of influence of the historical marketing strategy on the marketing effect based on the first prediction value and the second prediction value includes: calculating a first difference between the first prediction value of the optimal candidate marketing strategy and the first prediction values ​​of the other candidate marketing strategies, and determining a maximum value of the first difference, wherein the maximum value is used to characterize the marketing effect; calculating a second difference between the first prediction value of the optimal candidate marketing strategy and the second prediction value of the optimal candidate marketing strategy, wherein the second difference is used to characterize the degree of influence of the historical marketing strategy on the marketing effect.

[0012] In an embodiment of the present application, the target marketing strategy to be executed is determined by referring to the marketing effect and the degree of influence, including: if the maximum value exceeds a first preset threshold and the second difference does not exceed a second preset threshold, the target marketing strategy is determined according to the optimal candidate marketing strategy; if the maximum value does not exceed the first preset threshold and the second difference exceeds the second preset threshold, the target marketing strategy is determined according to the historical marketing strategy; if the maximum value exceeds the first preset threshold and the second difference exceeds the second preset threshold, the target marketing strategy is determined according to a combination of the historical marketing strategy and the optimal candidate marketing strategy.

[0013] A device for acquiring a marketing strategy comprises: an acquisition module for acquiring first user information and second user information of each target user in a target customer group, wherein the first user information is user information without the influence of a historical marketing strategy, and the second user information is user information under the influence of the historical marketing strategy, and the first user information and the second user information respectively include the behavior parameters of the target user; a first prediction module for inputting the first user information and candidate marketing strategy information into a prediction model corresponding to a conversion index to obtain a first prediction value of the conversion index, wherein the conversion index is used to characterize the degree of numerical change of the behavior parameter after the candidate marketing strategy is executed; a screening module for screening out an optimal candidate marketing strategy from a plurality of candidate marketing strategies based on the first prediction value of the conversion index corresponding to the candidate marketing strategy; a second prediction module for inputting the second user information and information of the optimal candidate marketing strategy into the prediction model to obtain a second prediction value of the conversion index; a third prediction module for predicting the marketing effect of the optimal candidate marketing strategy and the degree of influence of the historical marketing strategy on the marketing effect based on the first prediction value and the second prediction value; and a determination module for determining a target marketing strategy to be executed with reference to the marketing effect and the degree of influence.

[0014] In an embodiment of the present application, the information of the candidate marketing strategies includes marketing moments, marketing reach channels and marketing rights. Before inputting the first user information and the information of the candidate marketing strategies into the prediction model corresponding to the conversion indicator, the device is also used to: construct different candidate marketing strategies in the following way: adjust the parameter configuration of at least one of the marketing moments, the marketing reach channels and the marketing rights to obtain different candidate marketing strategies.

[0015] In an embodiment of the present application, the acquisition module is used to: obtain user information of the target user within a preset time period before the marketing moment of the historical marketing strategy as the first user information; and obtain user information of the target user within a preset time period after the marketing moment of the historical marketing strategy as the second user information.

[0016] In an embodiment of the present application, the first user information and the second user information further include marketing information, respectively, and the acquisition module is used to: obtain the user information of the target user within a preset time period after the marketing moment of the historical marketing strategy as the second user information; set the time series data set of each parameter under the marketing information in the second user information to an empty set, and set the statistical data of each parameter to zero to obtain the first user information.

[0017] In an embodiment of the present application, the device is also used to train the prediction model in the following manner: obtaining historical marketing records under the target marketing scenario, the historical marketing records including the executed marketing strategies, the marketing moments when the marketing strategies were started, the users corresponding to the marketing strategies, the user information of the users within a preset time after the marketing moments, and the user information of the users within a preset time before the marketing moments; using the historical marketing records to construct a training sample set of the conversion indicators, and training the prediction model based on the training sample set of the conversion indicators.

[0018] In an embodiment of the present application, the third prediction module is used to: calculate a first difference between a first prediction value of the optimal candidate marketing strategy and first prediction values ​​of other candidate marketing strategies, and determine a maximum value of the first difference, wherein the maximum value is used to characterize the marketing effect; calculate a second difference between the first prediction value of the optimal candidate marketing strategy and the second prediction value of the optimal candidate marketing strategy, wherein the second difference is used to characterize the degree of influence of the historical marketing strategy on the marketing effect.

[0019] In an embodiment of the present application, the determination module is used to: if the maximum value exceeds a first preset threshold and the second difference does not exceed the second preset threshold, determine the target marketing strategy according to the optimal candidate marketing strategy; if the maximum value does not exceed the first preset threshold and the second difference exceeds the second preset threshold, determine the target marketing strategy according to the historical marketing strategy; if the maximum value exceeds the first preset threshold and the second difference exceeds the second preset threshold, determine the target marketing strategy according to a combination of the historical marketing strategy and the optimal candidate marketing strategy.

[0020] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method for acquiring marketing strategies as described in the above embodiment are implemented.

[0021] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the method for acquiring a marketing strategy as described in the above embodiment.

[0022] In summary, this application proposes a method, device, electronic device and storage medium for obtaining a marketing strategy. This method compares the first predicted values ​​of conversion indicators of different candidate marketing strategies when the influence of historical marketing strategies is excluded to obtain the candidate marketing strategy with the best single effect, that is, the optimal candidate marketing strategy. Then, the predicted values ​​of conversion indicators under the two conditions of considering the influence of historical marketing strategies and excluding the influence of historical marketing strategies are used to determine the degree of influence of historical marketing strategies on the marketing effect of the optimal candidate marketing strategy, thereby attributing the marketing effect that may eventually occur, that is, determining the contribution of historical marketing strategies and the optimal candidate influence strategy to the final marketing effect. The final marketing strategy is formulated with reference to the attribution results to ensure that the marketing strategy to be executed has a good marketing effect and improves marketing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0024] Figure 1 is a flow chart of a method for acquiring a marketing strategy according to an exemplary embodiment of the present application; Figure 2 is a flow chart of a method for acquiring a marketing strategy according to an exemplary embodiment of the present application; Figure 3 is a flow chart of a method for acquiring a marketing strategy according to another exemplary embodiment of the present application; Figure 4 is a flow chart of a method for acquiring a marketing strategy according to another exemplary embodiment of the present application; Figure 5 is a schematic block diagram of a device for acquiring a marketing strategy according to another exemplary embodiment of the present application; Figure 6 It is a schematic block diagram of an electronic device according to an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. The embodiments described by reference to the drawings are exemplary and are intended to be used to explain the present application, and cannot be understood as limiting the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0026] Figure 1 is a flow chart of a method for obtaining a marketing strategy according to an exemplary embodiment of the present application. Figure 1 As shown, the method for obtaining the marketing strategy includes the following steps: S101, obtaining first user information and second user information of each target user in the target customer group, wherein the first user information is user information not affected by historical marketing strategies, and the second user information is user information affected by the historical marketing strategies, and the first user information and the second user information respectively include behavioral parameters of the target user.

[0027] In a specified marketing scenario, the target customer group for the marketing scenario is defined. This application obtains two types of user information for each target user in the target customer group. One is the user information that is not affected by historical marketing strategies, that is, the first user information. The second is the user information under the influence of historical marketing strategies, that is, the second user information. Among them, the first user information and the second user information may include but are not limited to the behavior information of the target user, the basic information of the target user, and the marketing information associated with the target user. The same moment can be used as a reference point, and the first user information or the second user information can be obtained within a statistical period before the moment, or the first user information or the second user information can be obtained within a statistical period after the moment.

[0028] The above-mentioned behavior information of the target user may include at least one behavior parameter.

[0029] For example, behavioral information may refer to the target user's registration of an account, login to the platform, use of rights, entry into designated business operations, purchase of services, and other behavioral sequence data and statistics within a period of time before a specific marketing moment. One of the above-mentioned behavioral parameters may be the sequence data and statistics of login behavior, such as constructing sequence data based on when the user logs in, and constructing statistics based on the total number of times the user logs in within a period of time. Another behavioral parameter may be the sequence data and statistics of the user entering the execution of business operations. Yet another behavioral parameter may be the sequence data and statistics of the user's use of rights, for example, rights may be coupons, discount / activity reminders, special user rights, etc., which are delivered to users through different contact channels.

[0030] The above basic information may include but is not limited to attribute data such as gender, age, education level, occupation type, etc. that the user authorizes and agrees to obtain, as well as business data such as the user's registration account date, customer acquisition channel, customer classification label, specific business identification, etc.

[0031] The marketing information associated with the target user may include information about the marketing strategies that the target user has received, such as the sequence data and statistical data of the marketing reach channels, the sequence data and statistical data of the marketing rights, etc., where the sequence data may be time series data or instant series data. If the user has not received any marketing strategies within a certain period of time, the various types of information in the marketing information associated with the user may be empty or zero to indicate that there is no associated marketing record for the user. For example, the marketing reach channels may be phone calls, text messages, APP push messages, etc. The marketing rights conveyed to customers through different marketing reach channels are the same as the above rights, which may be coupons, discount / activity reminders, user special rights, etc.

[0032] The above-mentioned “user information without the influence of historical marketing strategies” may be the behavior information of the target user without receiving any marketing strategies, the basic information submitted by the user, the marketing information associated with the user, etc. The corresponding “user information under the influence of historical marketing strategies” may be the behavior information, the basic information submitted, the marketing information associated with the target user, etc. when the target user has received marketing strategies.

[0033] In this way, we can obtain user information of target users after they have experienced marketing, and user information of target users before they have experienced marketing. Based on the two types of user information as basic data, we can analyze the impact of a pre-set marketing strategy on user behavior. For example, we can predict changes in user behavior under the influence of the marketing strategy.

[0034] S102, inputting the first user information and the information of the candidate marketing strategy into a prediction model corresponding to a conversion indicator to obtain a first prediction value of the conversion indicator, wherein the conversion indicator is used to characterize the degree of change in the value of the behavior parameter after the execution of the candidate marketing strategy.

[0035] The information of the candidate marketing strategy may include the marketing rights involved in the candidate marketing strategy, how to convey the rights to users, that is, the marketing reach channels, etc. That is, the information of the marketing strategy can be understood as the relevant information on how to implement marketing.

[0036] After implementing user marketing, a visitor user may be converted into a registered user, a new user may be further converted into an old user, the user's activity level and order transaction volume may change, and the magnitude of each change may be characterized by a conversion indicator. That is, in some embodiments, any of the above-mentioned behavior parameters in the user's behavior information may correspond to a conversion indicator, and the conversion indicator may be used to characterize the degree of change in the value of the behavior parameter. Therefore, the conversion indicator may also be able to characterize the marketing effect to a certain extent.

[0037] This application inputs user information that has not been affected by historical marketing strategies and information on a certain candidate marketing strategy into a prediction model corresponding to a certain conversion indicator, and predicts the performance of users who have not experienced marketing after experiencing the candidate marketing strategy, thereby achieving the prediction of the marketing effect of the candidate marketing strategy.

[0038] It should be noted that, when predicting a certain conversion indicator, the first user information input into the prediction model corresponding to the conversion indicator includes at least the behavior parameter corresponding to the conversion indicator.

[0039] This application can select different candidate marketing strategies and prediction models of different conversion indicators, and can use the first user information as basic data to predict the performance of different conversion indicators under the influence of different candidate marketing strategies.

[0040] S103: Filter out the best candidate marketing strategy from the plurality of candidate marketing strategies based on the first predicted value of the conversion indicator corresponding to the candidate marketing strategy.

[0041] For each candidate marketing strategy, the indicator value of one or more conversion indicators can be predicted; If only one conversion indicator value prediction is performed, the optimal candidate marketing strategy is screened out from multiple candidate marketing strategies based on the prediction result of the conversion indicator, for example, a candidate marketing strategy with the largest first prediction indicator value of the conversion indicator is selected as the optimal candidate marketing strategy.

[0042] If indicator value predictions are made for multiple conversion indicators, filtering conditions can be set for each conversion indicator to select the optimal candidate marketing strategy from the candidate marketing strategies that meet the filtering conditions for each conversion indicator; or the average of the first predicted values ​​of all conversion indicators can be calculated to select the candidate marketing strategy with the largest average value from multiple candidate marketing strategies as the optimal candidate marketing strategy.

[0043] Optionally, in some embodiments, the optimal candidate marketing strategy can be screened out from multiple candidate marketing strategies in the following manner: input the first prediction value into a marketing revenue prediction model pre-configured for the conversion indicator to obtain the predicted marketing revenue corresponding to the conversion indicator; and screen out the optimal candidate marketing strategy from the multiple candidate marketing strategies according to the predicted marketing revenues corresponding to different candidate marketing strategies.

[0044] For example, the marketing benefits of different candidate marketing strategies can be calculated by combining factors such as marketing costs, and the optimal candidate marketing strategy can be obtained through comparison.

[0045] Among them, marketing benefits may include direct business benefits and indirect benefits brought by conversion indicators. For example, the method of calculating direct benefits in combination with marketing costs may be: pre-setting the corresponding relationship between the values ​​of different conversion indicators and the newly purchased service volume. In the application, the first predicted value of a certain conversion indicator is used to determine the newly purchased service volume brought by the conversion indicator, and the newly purchased service volume is converted into direct business income according to the business rules of a specific business operation scenario, and then the marketing costs such as marketing manpower and resource usage fees are deducted to obtain the above direct benefits.

[0046] For example, the method of calculating indirect benefits in combination with marketing costs may be: determining the indicators that can characterize the incremental changes in users among the conversion indicators corresponding to different candidate marketing strategies, such as the conversion indicators corresponding to user registration behavior, predicting the number of new users brought by different candidate marketing strategies based on the first predicted value of the conversion indicator, converting the number of new users into possible business income that may be generated in the future according to the business rules of the specific business operation scenario, and then subtracting costs such as marketing manpower and resource usage fees to obtain the above-mentioned indirect benefits.

[0047] S104, inputting the second user information and the information of the optimal candidate marketing strategy into the prediction model to obtain a second prediction value of the conversion indicator.

[0048] In addition to determining the marketing effect of the optimal candidate marketing strategy alone, the present application will utilize the user information under the influence of the marketing strategy and the information of the optimal candidate marketing strategy, and input them together into the prediction model selected in step S102 to predict the marketing effect of the conversion indicator under the joint action of the historical marketing strategy and the optimal candidate marketing strategy.

[0049] S105: Based on the first prediction value and the second prediction value, predict the marketing effect of the optimal candidate marketing strategy and the degree of influence of the historical marketing strategy on the marketing effect.

[0050] The first predicted value of the conversion index is an estimate of the degree of change in the numerical value of the user behavior parameter when the optimal candidate marketing strategy is applied to users who have not received other marketing strategies. The marketing effect A when the optimal candidate marketing strategy is applied alone can be predicted with reference to the first predicted value.

[0051] The second predicted value based on the conversion indicator can predict the marketing effect achieved after superimposing the optimal candidate marketing strategy on the basis of other historical marketing strategies, that is, the marketing effect B when the optimal candidate marketing strategy is superimposed on other marketing strategies.

[0052] In some embodiments, based on the numerical relationship between the first prediction value and the second prediction value, the contribution of the historical marketing strategy to the marketing effect B can be determined as the degree of influence of the historical marketing strategy on the marketing effect A.

[0053] S106, determining a target marketing strategy to be executed with reference to the marketing effect and the degree of influence.

[0054] With reference to the marketing effect when the optimal candidate marketing strategy acts alone, the marketing effect when the optimal candidate marketing strategy and the historical candidate marketing strategies act in combination, and the degree of influence of the historical candidate marketing strategies on the marketing effect when the optimal candidate marketing strategy acts alone, a marketing strategy setting plan with the best marketing effect is determined, and the marketing strategies contained in the marketing strategy setting plan are determined as the target marketing strategies to be executed.

[0055] In summary, this application proposes a method for obtaining a marketing strategy. This application compares the first predicted values ​​of conversion indicators of different candidate marketing strategies when other recent marketing strategies (i.e., historical marketing strategies) are excluded, and obtains the candidate marketing strategy with the best single effect, i.e., the optimal candidate marketing strategy. Then, the predicted values ​​of conversion indicators under two conditions, considering or excluding the influence of other recent marketing strategies, can be used to determine whether the conversion effect is mainly due to the current marketing strategy, other recent marketing strategies, or the combined effect of the two, i.e., to achieve the attribution of marketing conversion. The final marketing strategy is formulated with reference to the attribution results.

[0056] Furthermore, the above prediction model can be trained in the following ways: Obtaining historical marketing records in a target marketing scenario, wherein the historical marketing records include an executed marketing strategy, a marketing moment when the marketing strategy begins to be executed, a user corresponding to the marketing strategy, user information of the user within a preset time after the marketing moment, and user information of the user within a preset time before the marketing moment; The historical marketing records are used to construct a training sample set of the conversion indicators, and the prediction model is trained based on the training sample set of the conversion indicators.

[0057] For example, you can use gradient boosting algorithm frameworks such as XGBoost and LightGBM, as well as machine learning algorithms such as artificial neural network algorithms to build prediction models for different conversion indicators. Each model can be used to predict the value of the corresponding conversion indicator.

[0058] The marketing moment can be understood as the moment when a marketing strategy begins to be implemented. The marketing moment can be set according to a certain day of a certain month, or a certain time of a certain day. Or it can be set according to a pre-defined node, for example, the moment when the user logs in to the platform is a marketing moment, the moment when the user enters the platform interface and browses after a predetermined time is a marketing moment, etc. This application does not limit this.

[0059] Among them, the user information in the training sample may include the user's behavior parameters, marketing information, and may also include the user's basic information, etc. For the description of the behavior parameters, marketing information, and basic information, please refer to the description of the above embodiment, which will not be repeated here.

[0060] This application constructs training samples based on user information before the marketing strategy is executed, when the user has not experienced marketing, and user information after the marketing strategy is executed, when the user has experienced marketing, so that the prediction model learns the user behavior performance in both cases, and trains the user-associated marketing information into a key feature that affects the conversion indicator value, so that in actual applications, the prediction of the conversion indicator value can be guided by the marketing information.

[0061] Furthermore, the information of the above-mentioned candidate marketing strategies may include but is not limited to marketing moments, marketing reach channels and marketing rights and interests.

[0062] In some embodiments, before executing the step S102 of "inputting the first user information and the information of the candidate marketing strategy into the prediction model corresponding to the conversion indicator", different candidate marketing strategies may be constructed in the following manner: Adjust the parameter configuration of at least one of the marketing moment, the marketing reach channel and the marketing rights to obtain different candidate marketing strategies.

[0063] For example, by adjusting one or more configurations of a preset candidate marketing strategy plan, such as the marketing moment, reach channels, marketing rights, etc., to form a new marketing strategy, the newly obtained marketing strategy is also used as a candidate marketing strategy to predict the conversion index. The method of the present invention can be used to estimate the execution effects of as many marketing strategies as possible. Compared with the existing technology, the present application can quickly screen multiple marketing strategies to obtain the marketing strategy with the best effect.

[0064] As a feasible implementation method, Figure 2As shown, in the above step S101, "obtaining the first user information and the second user information of each target user in the target customer group" may include the following steps: S201, obtaining user information of the target user within a preset time period before the marketing moment of the historical marketing strategy as the first user information; S202, obtaining user information of the target user within a preset time period after the marketing moment of the historical marketing strategy as the second user information.

[0065] For example, for users with recent marketing records, the latest basic information and behavior information before the marketing moment when the marketing strategy was implemented in the recent marketing records are obtained. Since the user's basic information, behavior information and other data will be affected by recent marketing, for users with recent marketing records, the latest basic information, behavior information and other data before the recent marketing moment are taken to replace the current basic information, behavior information and other data, thereby eliminating the influence of recent marketing; and for users without recent marketing records, their current basic information, behavior information and other data are obtained.

[0066] This application obtains the second user information of the user under the influence of the historical marketing strategy during a period of time when there are marketing records, and obtains the first user information of the user when there is no marketing during a period of time when there are no marketing records. The first user information and the second user information obtained can clearly show the information changes of the user under the influence of the marketing strategy, provide accurate basic data for predicting the effect of the candidate marketing strategy alone and the effect of the superposition with the historical marketing strategy, and ensure the accuracy of effect prediction and marketing conversion attribution.

[0067] As another feasible implementation, the first user information and the second user information may also include marketing information, such as Figure 3 As shown, the above step S101 of "obtaining the first user information and the second user information of each target user in the target customer group" may include the following steps: S301, obtaining user information of the target user within a preset time period after the marketing moment of the historical marketing strategy as the first user information; S302, setting the time series data set of each parameter under the marketing information in the first user information to an empty set, setting the statistical data of each parameter to zero, and obtaining the second user information.

[0068] Obtain the user's behavior parameters and marketing information after experiencing historical marketing. The marketing information records the historical marketing that the user has experienced. The marketing message may include marketing access channels and marketing rights, such as recording the channels through which the user has received coupons in the marketing message. When determining the second user information, set the time series data set / sequence data set of the marketing access channels under the marketing information in the first user information to an empty set, and set the statistical data of the marketing access channels to zero, so as to indicate to the prediction model that the user has not received historical marketing. The behavior parameters in the second user information can be regarded as the original behavior parameters of the user before experiencing marketing. When predicting conversion indicators, the prediction model only refers to the candidate marketing strategy to predict changes in user behavior, and no longer considers the impact of historical marketing strategies.

[0069] The user information selected in this application includes not only the user's behavior parameters, but also the marketing information associated with the user. The behavior parameters and marketing information are input into the model as input parameters. The model takes into account the characteristics of the marketing information when predicting the conversion index value. The marketing information is used as a factor affecting the conversion index value. This application removes the impact of historical marketing strategies on the conversion index value in the model prediction by setting the marketing information in the second user information to zero. Therefore, the second user information after the marketing information is set to zero is regarded as the first user information that is not affected by the historical marketing strategy. In this way, the efficiency of obtaining basic data is improved and the storage and computing pressures are reduced.

[0070] Based on the above embodiments, Figure 4 As shown, in the above step S105, "predicting the marketing effect of the optimal candidate marketing strategy and the degree of influence of the historical marketing strategy on the marketing effect based on the first prediction value and the second prediction value" may include the following steps: S401, calculating a first difference between a first prediction value of the optimal candidate marketing strategy and first prediction values ​​of other candidate marketing strategies, and determining a maximum value of the first difference, wherein the maximum value is used to characterize the marketing effect; S402, calculating a second difference between a first prediction value of the optimal candidate marketing strategy and a second prediction value of the optimal candidate marketing strategy, wherein the second difference is used to characterize the degree of influence of the historical marketing strategy on the marketing effect.

[0071] In some embodiments, the first prediction value may also be directly used to characterize the marketing effect of the corresponding candidate marketing strategy.

[0072] In the embodiment of the present application, the maximum difference between the first prediction value of the optimal candidate marketing strategy and the first prediction values ​​of other candidate marketing strategies is used to characterize the marketing effect of the optimal candidate marketing strategy. On the basis of observing the actual marketing effect of the optimal candidate marketing strategy when it acts alone by using the above first difference, the leading margin of the effect of the optimal candidate marketing strategy compared with other candidate marketing strategies can be further observed. If the leading margin is high, it can be considered that the marketing effect of the optimal candidate marketing strategy is good, and if the leading margin is low, it can be considered that the marketing effect of the optimal candidate marketing strategy is average.

[0073] For example, by comparing the first prediction values ​​of each candidate marketing strategy, the lead margin of the candidate marketing strategy with the largest first prediction value over other candidate marketing strategies (hereinafter referred to as "magnitude one") can be calculated, and the leading margin can be used as a representation of the marketing effect of the optimal candidate marketing strategy. By comparing the first prediction value and the second prediction value of the candidate marketing strategy with the largest first prediction value, the improvement in the conversion index value by other recent marketing strategies (hereinafter referred to as "magnitude two") can be calculated, and the improvement magnitude two can be used as a representation of the marketing degree of historical marketing strategies on the marketing effect of the optimal candidate marketing strategy.

[0074] By specifically characterizing the effects of the optimal candidate marketing strategy and the degree of influence of historical marketing strategies, a reference is provided for the selection and adjustment of the target marketing strategy and the determination of the marketing strategy setting plan.

[0075] On the basis of the above embodiment, after obtaining the maximum value and the second value of the above difference, the "determining the target marketing strategy to be executed by referring to the marketing effect and the influence degree" in the above step S106 can be divided into the following situations: First, if the maximum value exceeds a first preset threshold and the second difference does not exceed a second preset threshold, determining the target marketing strategy according to the optimal candidate marketing strategy; Second, if the maximum value does not exceed the first preset threshold, and the second difference value exceeds the second preset threshold, determining the target marketing strategy according to the historical marketing strategy; Third, if the maximum value exceeds the first preset threshold and the second difference exceeds the second preset threshold, the target marketing strategy is determined according to the combination of the historical marketing strategy and the optimal candidate marketing strategy.

[0076] For example, when "Amplitude One" is large and "Amplitude Two" is small, it can be considered that the increase in the conversion index value of the marketing strategy with the largest first prediction value is mainly due to the marketing effect of the marketing strategy; the optimal candidate marketing strategy can be directly determined as the target marketing strategy, or the marketing moments, marketing reach channels and marketing rights in the target marketing strategy can be set with reference to the optimal candidate marketing strategy.

[0077] When "Amplitude One" is large and "Amplitude Two" is large, it can be considered that the improvement in the conversion index value of the marketing strategy with the largest first prediction value is mainly due to the marketing effect of the combination of historical marketing strategies and this marketing strategy; the historical marketing strategy and the optimal candidate marketing strategy can be directly determined as the target marketing strategy, such as determining the marketing strategy setting plan as a combination of the historical marketing strategy and the optimal candidate marketing strategy.

[0078] When "Amplitude 1" is small and "Amplitude 2" is large, it can be considered that the conversion index value of the marketing strategy with the largest first prediction value is mainly due to the marketing effect of the historical marketing strategy. The historical marketing strategy can be directly determined as the target marketing strategy, or the marketing moments, marketing reach channels and marketing rights in the target candidate marketing strategy can be set with reference to the historical marketing strategy.

[0079] Through the comprehensive analysis of the actual effectiveness, the target marketing strategy to be implemented is determined.

[0080] It should be noted that the specific values ​​of the first preset threshold and the second preset threshold can be set as needed, and this application does not limit them.

[0081] In summary, this application predicts the indicator effects of marketing strategies by building a machine learning model, replacing the testing method based on a priori combined with comparative experiments. By flexibly modifying the input feature data during model prediction and comparing the results, the best marketing strategy for the target customer group under the current situation and excluding the influence of other marketing strategies is obtained, thereby providing more comprehensive decision-making reference information. By comparing the predicted values ​​of conversion indicators of different strategies excluding the influence of other recent marketing strategies, and comparing the predicted values ​​of conversion indicators of the same strategy plan under two conditions of considering or excluding the influence of other recent marketing strategies, it can be determined whether the conversion effect is mainly due to the current marketing strategy, or due to other recent marketing strategies, or the superposition of the two, that is, to achieve the attribution of marketing conversion. This can effectively shorten the marketing strategy plan formulation cycle and reduce unnecessary resource investment and time waste. Through marketing conversion attribution, the marketing strategy plan is guided to be accurately optimized and marketing efficiency is improved.

[0082] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0083] Figure 5 is a block diagram of a device for acquiring a marketing strategy according to an exemplary embodiment of the present application, such as Figure 5 As shown, the apparatus 500 includes: an acquisition module 501 , a first prediction module 502 , a screening module 503 , a second prediction module 504 , a third prediction module 505 and a determination module 506 .

[0084] An acquisition module 501 is used to acquire first user information and second user information of each target user in the target customer group, wherein the first user information is user information without the influence of the historical marketing strategy, and the second user information is user information under the influence of the historical marketing strategy, and the first user information and the second user information respectively include the behavior parameters of the target user; A first prediction module 502, for inputting the first user information and the candidate marketing strategy information into a prediction model corresponding to a conversion indicator to obtain a first prediction value of the conversion indicator, wherein the conversion indicator is used to characterize the degree of change in the value of the behavior parameter after the candidate marketing strategy is executed; A screening module 503, configured to screen out an optimal candidate marketing strategy from a plurality of the candidate marketing strategies based on the first predicted value of the conversion indicator corresponding to the candidate marketing strategy; A second prediction module 504 is used to input the second user information and the information of the optimal candidate marketing strategy into the prediction model to obtain a second prediction value of the conversion indicator; A third prediction module 505, configured to predict the marketing effect of the optimal candidate marketing strategy and the degree of influence of the historical marketing strategy on the marketing effect based on the first prediction value and the second prediction value; The determination module 506 is used to determine the target marketing strategy to be executed with reference to the marketing effect and the influence degree.

[0085] In an embodiment of the present application, the information of the candidate marketing strategies includes marketing moments, marketing reach channels and marketing rights, and the device is also used to construct different candidate marketing strategies in the following manner: adjust the parameter configuration of at least one of the marketing moments, the marketing reach channels and the marketing rights to obtain different candidate marketing strategies.

[0086] In an embodiment of the present application, the acquisition module is used to: obtain user information of the target user within a preset time period before the marketing moment of the historical marketing strategy as the first user information; and obtain user information of the target user within a preset time period after the marketing moment of the historical marketing strategy as the second user information.

[0087] In an embodiment of the present application, the first user information and the second user information further include marketing information, respectively, and the acquisition module is used to: obtain the user information of the target user within a preset time period after the marketing moment of the historical marketing strategy as the second user information; set the time series data set of each parameter under the marketing information in the second user information to an empty set, and set the statistical data of each parameter to zero to obtain the first user information.

[0088] In an embodiment of the present application, the device is also used to train the prediction model in the following manner: obtaining historical marketing records under the target marketing scenario, the historical marketing records including the executed marketing strategies, the marketing moments when the marketing strategies were started, the users corresponding to the marketing strategies, the user information of the users within a preset time after the marketing moments, and the user information of the users within a preset time before the marketing moments; using the historical marketing records to construct a training sample set of the conversion indicators, and training the prediction model based on the training sample set of the conversion indicators.

[0089] In an embodiment of the present application, the third prediction module is used to: calculate a first difference between a first prediction value of the optimal candidate marketing strategy and first prediction values ​​of other candidate marketing strategies, and determine a maximum value of the first difference, wherein the maximum value is used to characterize the marketing effect; calculate a second difference between the first prediction value of the optimal candidate marketing strategy and the second prediction value of the optimal candidate marketing strategy, wherein the second difference is used to characterize the degree of influence of the historical marketing strategy on the marketing effect.

[0090] In an embodiment of the present application, the determination module is used to: if the maximum value exceeds a first preset threshold and the second difference does not exceed the second preset threshold, determine the target marketing strategy according to the optimal candidate marketing strategy; if the maximum value does not exceed the first preset threshold and the second difference exceeds the second preset threshold, determine the target marketing strategy according to the historical marketing strategy; if the maximum value exceeds the first preset threshold and the second difference exceeds the second preset threshold, determine the target marketing strategy according to a combination of the historical marketing strategy and the optimal candidate marketing strategy.

[0091] In summary, the device compares the first predicted values ​​of conversion indicators of different candidate marketing strategies when excluding the influence of other recent marketing strategies (i.e., historical marketing strategies) to obtain the candidate marketing strategy with the best single effect, i.e., the optimal candidate marketing strategy. Then, the predicted values ​​of conversion indicators under the two conditions of considering or excluding the influence of other recent marketing strategies are used to determine whether the conversion effect is mainly due to the current marketing strategy, other recent marketing strategies, or the combined effect of the two, i.e., to achieve the attribution of marketing conversion. The final marketing strategy is formulated with reference to the attribution results to ensure that the marketing strategy to be executed has a good marketing effect and improves marketing efficiency.

[0092] To implement the above embodiments, the embodiments of the present application also propose an electronic device 600, including a memory 601, a processor 602, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method for obtaining the marketing strategy described in the above embodiments are implemented.

[0093] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0094] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0095] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for acquiring a marketing strategy, characterized in that: include: Acquire first user information and second user information of each target user in the target customer group, wherein the first user information is user information without the influence of the historical marketing strategy, and the second user information is user information under the influence of the historical marketing strategy, and the first user information and the second user information respectively include the behavior parameters of the target user; Inputting the first user information and the information of the candidate marketing strategy into a prediction model corresponding to a conversion indicator to obtain a first prediction value of the conversion indicator, wherein the conversion indicator is used to characterize the degree of change in the value of the behavior parameter after the candidate marketing strategy is executed; Based on the first predicted value of the conversion indicator corresponding to the candidate marketing strategy, selecting the best candidate marketing strategy from the plurality of candidate marketing strategies; Inputting the second user information and the information of the optimal candidate marketing strategy into the prediction model to obtain a second prediction value of the conversion indicator; Based on the first prediction value and the second prediction value, predict the marketing effect of the optimal candidate marketing strategy and the degree of influence of the historical marketing strategy on the marketing effect; The target marketing strategy to be executed is determined with reference to the marketing effect and the degree of influence.

2. The method according to claim 1, characterized in that The information of the candidate marketing strategy includes marketing moments, marketing reach channels and marketing rights. Before inputting the first user information and the information of the candidate marketing strategy into the prediction model corresponding to the conversion indicator, the method further includes: constructing different candidate marketing strategies in the following manner: Adjust the parameter configuration of at least one of the marketing moment, the marketing reach channel and the marketing rights to obtain different candidate marketing strategies.

3. The method according to claim 1, characterized in that The step of obtaining the first user information and the second user information of each target user in the target customer group includes: Acquire user information of the target user within a preset time period before the marketing moment of the historical marketing strategy as the first user information; The user information of the target user within a preset time period after the marketing moment of the historical marketing strategy is obtained as the second user information.

4. The method according to claim 1, characterized in that The first user information and the second user information further include marketing information respectively, and the obtaining of the first user information and the second user information of each target user in the target customer group includes: Acquire user information of the target user within a preset time period after the marketing moment of the historical marketing strategy as the second user information; The time series data set of each parameter under the marketing information in the second user information is set to an empty set, and the statistical data of each parameter is set to zero to obtain the first user information.

5. The method according to claim 1, characterized in that The prediction model is trained by: Obtaining historical marketing records in a target marketing scenario, wherein the historical marketing records include an executed marketing strategy, a marketing moment when the marketing strategy begins to be executed, a user corresponding to the marketing strategy, user information of the user within a preset time after the marketing moment, and user information of the user within a preset time before the marketing moment; The historical marketing records are used to construct a training sample set of the conversion indicators, and the prediction model is trained based on the training sample set of the conversion indicators.

6. The method according to claim 1, characterized in that The predicting, based on the first prediction value and the second prediction value, the marketing effect of the optimal candidate marketing strategy and the degree of influence of the historical marketing strategy on the marketing effect includes: Calculating a first difference between the first prediction value of the optimal candidate marketing strategy and the first prediction values ​​of other candidate marketing strategies, and determining a maximum value of the first difference, wherein the maximum value is used to characterize the marketing effect; A second difference between the first prediction value of the optimal candidate marketing strategy and the second prediction value of the optimal candidate marketing strategy is calculated, and the second difference is used to characterize the degree of influence of the historical marketing strategy on the marketing effect.

7. The method according to claim 6, characterized in that The determining of the target marketing strategy to be executed with reference to the marketing effect and the degree of influence includes: If the maximum value exceeds a first preset threshold value, and the second difference value does not exceed a second preset threshold value, determining the target marketing strategy according to the optimal candidate marketing strategy; If the maximum value does not exceed the first preset threshold, and the second difference value exceeds the second preset threshold, determining the target marketing strategy according to the historical marketing strategy; If the maximum value exceeds the first preset threshold and the second difference exceeds the second preset threshold, the target marketing strategy is determined according to a combination of the historical marketing strategy and the optimal candidate marketing strategy.

8. A device for acquiring marketing strategies, characterized in that: include: An acquisition module, used to acquire first user information and second user information of each target user in the target customer group, wherein the first user information is user information without the influence of the historical marketing strategy, and the second user information is user information under the influence of the historical marketing strategy, and the first user information and the second user information respectively include the behavior parameters of the target user; A first prediction module, used to input the first user information and the candidate marketing strategy information into a prediction model corresponding to a conversion indicator to obtain a first prediction value of the conversion indicator, wherein the conversion indicator is used to characterize the degree of change in the value of the behavior parameter after the candidate marketing strategy is executed; A screening module, configured to screen out an optimal candidate marketing strategy from a plurality of the candidate marketing strategies based on the first predicted value of the conversion indicator corresponding to the candidate marketing strategy; A second prediction module, used for inputting the second user information and the information of the optimal candidate marketing strategy into the prediction model to obtain a second prediction value of the conversion indicator; a third prediction module, configured to predict the marketing effect of the optimal candidate marketing strategy and the degree of influence of the historical marketing strategy on the marketing effect based on the first prediction value and the second prediction value; A determination module is used to determine a target marketing strategy to be executed with reference to the marketing effect and the degree of influence.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method for acquiring a marketing strategy as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method for acquiring a marketing strategy as described in any one of claims 1 to 7 are implemented.

Images