Advertisement delivery adjustment method and device, and storage medium

Abstract

The application discloses an advertisement delivery adjustment method and device and a storage medium. The method comprises the following steps: obtaining a plurality of target objects corresponding to an advertisement activity to be adjusted and associated advertisement cost budget intervals; determining a predicted advertisement cost and a predicted benefit index corresponding to the target objects based on a prediction neural network model, the target objects and the corresponding advertisement cost budget intervals; determining a predicted sales cost corresponding to the target objects according to the predicted advertisement cost and the predicted benefit index corresponding to the target objects; determining a budget change condition corresponding to the advertisement activity to be adjusted according to a current budget and a total historical advertisement cost; and adjusting the delivery value of the plurality of target objects according to the budget change condition and the predicted sales cost corresponding to the target objects. In the case that the current budget is unchanged, the delivery value of the target objects with poor advertisement effects is appropriately reduced, and the delivery value of the target objects with good advertisement effects is increased, so that the budget amount distribution is more reasonable, and the advertisement delivery effect is improved.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and more specifically, to an advertising placement adjustment method, apparatus, and storage medium. Background Technology

[0002] E-commerce platforms facilitate online transactions between users and merchants. Users can enter relevant keywords into the platform, and the page will display products related to those keywords. Merchants typically identify at least one keyword based on their advertising campaigns; when a keyword matches a user's input, the page will display the corresponding product from that campaign.

[0003] Typically, merchants set their own rules for adjusting advertising spending for each keyword and use scheduled tasks to verify the rationality of these rules. However, this relies heavily on the user's extensive operational experience and is prone to inappropriate adjustments to advertising spending, resulting in poor advertising performance. Summary of the Invention

[0004] In view of the above problems, the present invention proposes an advertising placement adjustment method, apparatus and storage medium to improve the above technical problems.

[0005] In a first aspect, embodiments of this application provide an advertising placement adjustment method, which includes: obtaining multiple target objects corresponding to an advertising campaign to be adjusted, and an advertising spending budget range corresponding to each target object; determining the predicted advertising spending and predicted benefit indicators for each target object based on a predictive neural network model, the multiple target objects, and the advertising spending budget range corresponding to each target object; determining the predicted sales cost for each target object based on the predicted advertising spending, predicted benefit indicators, historical advertising spending, historical benefit indicators, and historical sales costs corresponding to each target object; determining the budget change for the advertising campaign to be adjusted based on the current budget and the historical total advertising spending of the advertising campaign to be adjusted; and adjusting the placement value of the multiple target objects based on the budget change and the predicted sales cost of the multiple target objects.

[0006] Secondly, embodiments of this application also provide an advertising placement adjustment device, which includes: an acquisition module, used to acquire multiple target objects corresponding to the advertising campaign to be adjusted, and an advertising spending budget range corresponding to each target object; a first determination module, used to determine the predicted advertising spending and predicted benefit indicators corresponding to each target object based on a predictive neural network model, multiple target objects, and the advertising spending budget range corresponding to each target object; a second determination module, used to determine the predicted sales cost corresponding to each target object based on the predicted advertising spending, predicted benefit indicators, historical advertising spending, historical benefit indicators, and historical sales costs corresponding to each target object; a third determination module, used to determine the budget change corresponding to the advertising campaign to be adjusted based on the current budget and the historical total advertising spending corresponding to the advertising campaign to be adjusted; and an adjustment module, used to adjust the placement value of multiple target objects based on the budget change and the predicted sales cost corresponding to multiple target objects.

[0007] Thirdly, embodiments of this application also provide a computer-readable storage medium storing program code, wherein the above-mentioned advertising placement adjustment method is executed when the program code is run by a processor.

[0008] The technical solution provided by this invention specifically includes: obtaining multiple target objects corresponding to the advertising campaign to be adjusted, and the advertising spending budget range corresponding to each target object; determining the predicted advertising spending and predicted benefit indicators for each target object based on a predictive neural network model, the multiple target objects, and the advertising spending budget range corresponding to each target object; determining the predicted sales cost for each target object based on the predicted advertising spending, predicted benefit indicators, historical advertising spending, historical benefit indicators, and historical sales costs; determining the budget change for the advertising campaign to be adjusted based on the current budget and the historical total advertising spending of the advertising campaign to be adjusted; and adjusting the placement values ​​for the multiple target objects based on the budget change and the predicted sales costs for the multiple target objects. Therefore, by combining the prediction of key parameters reflecting the advertising effect of the advertising campaign to be adjusted using a predictive neural network model, and by optimizing the placement values ​​for the target objects in the advertising campaign to be adjusted using these predicted key parameters, advertising investment can be reasonably adjusted to improve advertising effectiveness. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments and drawings obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0010] Figure 1 The illustration shows a flowchart of an advertising placement adjustment method provided in an embodiment of this application.

[0011] Figure 2 A schematic diagram of the adjustment interface provided in an embodiment of this application is shown.

[0012] Figure 3 A schematic diagram of an advertising placement adjustment device provided in an embodiment of this application is shown.

[0013] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of this application is shown.

[0014] Figure 5 This illustration shows a schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of this application. Detailed Implementation

[0015] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0016] E-commerce platforms facilitate online transactions between users and merchants. Users can enter relevant keywords into the platform, and the page will display products related to those keywords. Merchants typically identify at least one keyword based on their advertising campaigns; when a keyword matches a user's input, the page will display the corresponding product from that campaign.

[0017] Typically, merchants set their own rules for adjusting advertising spending for each keyword and use scheduled tasks to verify the rationality of these rules. However, this relies heavily on the user's extensive operational experience and is prone to inappropriate adjustments to advertising spending, resulting in poor advertising performance.

[0018] To address the aforementioned issues, the inventors have proposed an advertising placement adjustment method, apparatus, and storage medium as provided in this application. The method specifically includes: acquiring multiple target objects corresponding to the advertising campaign to be adjusted, and the advertising spending budget range for each target object; determining the predicted advertising spending and predicted benefit indicators for each target object based on a predictive neural network model, the multiple target objects, and the advertising spending budget range for each target object; determining the predicted sales cost for each target object based on the predicted advertising spending, predicted benefit indicators, historical advertising spending, historical benefit indicators, and historical sales costs; determining the budget change for the advertising campaign to be adjusted based on the current budget and the historical total advertising spending for the advertising campaign to be adjusted; and adjusting the placement values ​​for the multiple target objects based on the budget change and the predicted sales costs for the multiple target objects.

[0019] Therefore, by combining the predictive neural network model to predict the advertising effectiveness of the target audience in the advertising campaign to be adjusted, and to determine the corresponding predicted advertising expenditure, further, based on the predicted effectiveness indicators and predicted advertising expenditure, the predicted sales cost, which reflects the advertising efficiency of the target audience in the advertising campaign to be adjusted, is obtained. Further still, based on the current budget of the advertising campaign to be adjusted, the corresponding budget changes, and the predicted sales cost, the investment value for the target audience in the advertising campaign to be adjusted is optimized to achieve reasonable adjustment of advertising investment and improve the effectiveness of advertising investment. Please refer to the following steps for specific implementation details.

[0020] Please see Figure 1 , Figure 1 The illustration shows a flowchart of an advertising placement adjustment method provided in an embodiment of this application, which may include steps 110 to 150.

[0021] In step 110, multiple target objects corresponding to the advertising campaign to be adjusted, and the advertising spending budget range corresponding to each target object are obtained.

[0022] Merchants can create advertising campaigns on e-commerce platforms. These campaigns can be SP (Sponsored Products) ads, SB (Sponsored Brands) ads, SD (Sponsored Display) ads, etc. In this embodiment, SP ads are used as an example to illustrate the advertising campaign.

[0023] The target object is the object associated with the products in the advertising campaign. Users can search for the products in the advertising campaign corresponding to the target object by searching for the target object on the e-commerce platform.

[0024] In some implementations, the target audience includes key objects. Merchants can select multiple key objects when creating an advertising campaign. The key objects can be determined by the merchant based on the characteristics of the product and user habits. For example, a merchant can set "shirt" and "top" as the two key objects associated with a garment. Merchants can allocate corresponding spending values ​​to key objects based on their budget and importance, thereby adjusting the effectiveness of the advertising campaign. Taking an advertising campaign as an example of a SP (Service Provider) campaign, the key objects are the keywords set when creating the campaign.

[0025] In some implementations, the target object includes a search object. When searching for products on an e-commerce platform, users can input a search object based on their needs. The search object is randomly entered by the user according to their requirements. The search object corresponding to the products in the advertising campaign can be retrieved through the report data provided by the e-commerce platform. For example, if a user enters "printed shirt" into the e-commerce platform and finds products in an advertising campaign, but "printed shirt" is not among the keywords in the advertising campaign, then "printed shirt" can be included in the search object corresponding to the advertising campaign. Taking an advertising campaign as an example of a SP (Special Offer) advertisement, the search object is the search term corresponding to the advertising campaign.

[0026] It is understood that this application is not limited to this, and the target audience may be adjusted accordingly based on the type of advertising campaign to be adjusted.

[0027] In the embodiments of this application, the advertising campaign to be adjusted is an advertising campaign that needs to be optimized and adjusted in its delivery.

[0028] In the embodiments of this application, the advertising spending budget range corresponding to the target object is used to characterize the budget range of the advertising spending corresponding to each target object.

[0029] Optionally, the advertising budget range for the target audience can be customized by the merchant based on the importance of the target audience.

[0030] Optionally, the advertising spending budget range for the target audience can be determined based on the target audience's historical advertising spending, thereby reasonably determining the preset advertising spending range for the target audience according to the target audience's historical performance.

[0031] In some implementations, step 110 may include the following steps.

[0032] (1) Identify multiple key targets for the advertising campaign to be adjusted.

[0033] (2) Obtain multiple search objects corresponding to the advertising campaign to be adjusted.

[0034] (3) The set of multiple key objects and multiple search objects is deduplicated to obtain multiple target objects.

[0035] (4) Obtain the historical advertising expenditure for each target object.

[0036] (5) Determine the historical advertising expenditure for each target object based on the historical advertising expenditure and budget range for each target object.

[0037] In the embodiments of this application, the key objects are set by the merchant based on product characteristics and user habits, while the search objects are input by the user based on product characteristics and actual needs. This may result in duplicate objects between the key objects and search objects corresponding to the advertising campaign to be adjusted. Therefore, duplicate objects in the set of multiple key objects and multiple search objects can be deduplicated first to determine the multiple target objects to be processed later.

[0038] In embodiments of this application, historical advertising spending is used to characterize the advertising spending actually used by the target object.

[0039] Optionally, historical advertising spending can be the daily average of advertising spending for the target audience within a preset time period. For example, the report data of the advertising campaign to be adjusted can be retrieved from the report data corresponding to the advertising campaign provided by the e-commerce platform, the historical total spending of the target audience in the preset time period (e.g., the previous 3 days to the previous 18 days) can be retrieved, and the daily average can be determined based on the historical total spending and the number of days corresponding to the preset time period, and used as historical advertising spending.

[0040] In order to determine the optimal predicted advertising expenditure, in the embodiments of this application, the advertising expenditure budget range corresponding to each target object is further determined based on the historical advertising expenditure and budget range corresponding to each target object.

[0041] The budget range can be set by the merchant based on the actual situation. The larger the budget range, the greater the amount of calculation required to determine the optimal predicted advertising expenditure in subsequent steps, and the higher the accuracy. The specific range can be selected according to the merchant's needs for calculation efficiency and accuracy.

[0042] For example, if the historical ad spend associated with the target object is 100 and the budget range is [80%, 120%], then the ad spend budget range for the target object is [80, 120]; or if the budget range is [60%, 140%], then the ad spend budget range for the target object is [60, 140].

[0043] In step 120, the predicted advertising cost and predicted benefit index for each target object are determined based on the predictive neural network model, multiple target objects, and the advertising cost budget range corresponding to each target object.

[0044] In embodiments of this application, a predictive neural network model that can be pre-trained to determine predictive advertising spending and predictive benefit indicators can be used.

[0045] In some implementations, the predictive neural network model may employ the XGBoost (eXtreme GradientBoosting) model. It is understood that in other implementations, other types of models may also be used, and no limitation is made here.

[0046] In some implementations, report data (such as advertising spending, order conversion rate, click-through rate, etc.) corresponding to completed or ongoing advertising campaigns can be used as the training set for the predictive neural network model.

[0047] To improve the accuracy of predictive neural network models, the model can be trained by incorporating various influencing factors related to the effectiveness of advertising campaigns. These influencing factors can include, for example, discount data corresponding to the advertising campaign, product rating data, and advertising spending. The specific factors can be set according to actual usage needs, and this application does not impose any restrictions on them.

[0048] In the embodiments of this application, the predicted benefit index can reflect the advertising effect when the target object promotes based on predicted advertising expenditure.

[0049] In the embodiments of this application, the predicted benefit indicators may include indicators such as order conversion rate and click-through rate.

[0050] The order conversion rate (Cr) is used to characterize the probability that a user will choose to purchase the product after clicking on the product link to view it. It can be determined based on the ratio between the total number of valid orders for the product and the total number of visits to the product link.

[0051] Click-Thru Rate (Ctr) is used to characterize the probability that a user will click on a product link to view the product when it is displayed on a page. It can be determined by the ratio of the number of clicks on the product link to the number of impressions on the page where the product is displayed.

[0052] Understandably, the higher the order conversion rate and click-through rate, the better the product promotion effect and the better the promotion effect on the target audience.

[0053] In this embodiment of the application, step 120 may include the following steps.

[0054] (1) Determine multiple advertising expenditures to be tested from the advertising expenditure budget range corresponding to each target object.

[0055] (2) Input each advertising expenditure to be tested into the preset neural network model to obtain the expected benefit index corresponding to each advertising expenditure to be tested.

[0056] (3) Determine the predicted benefit indicators from the multiple benefit indicators to be measured corresponding to each target object.

[0057] (4) The advertising expenditure budget corresponding to the predicted benefit index of each target object is used as the predicted advertising expenditure of each target object.

[0058] Among these, the ad spending to be tested falls under the ad spending budget range, and multiple ad spending items to be tested can be determined from the ad spending budget range. The number of ad spending items to be tested and the selection method can be chosen according to actual needs.

[0059] For example, the advertising budget range can be [80, 120]. The values ​​of the advertising expenditure to be tested can be obtained at equal intervals in the preset advertising expenditure range with the same preset difference. For example, with 1 as the preset difference, the advertising expenditure to be tested in the advertising budget range [80, 120] can be determined as: 80, 81, 82...118, 119, 120.

[0060] Each advertising expenditure corresponding to the target object is input into a preset neural network model to obtain the expected benefit index corresponding to each advertising expenditure. The predicted benefit index is then determined from the multiple expected benefit indices corresponding to each target object.

[0061] Among them, the predicted benefit index can indicate the optimal benefit index among multiple benefit indices to be measured for the target object.

[0062] Alternatively, the highest performance indicator to be measured can be used as the performance indicator for prediction.

[0063] Optionally, the expected benefit index can be determined by the ratio of the advertising expenditure to the predicted benefit index, which is the set of data with the lowest advertising expenditure and the highest benefit index, so as to determine the predicted benefit index with the highest cost performance.

[0064] Furthermore, the budget for advertising expenditure to be predicted corresponding to the predicted benefit indicators is used as the predicted advertising expenditure for each target object.

[0065] For example, the target object is "shirt", and the advertising budget range for "shirt" is [80, 82]. The test advertising cost for "shirt" can be 80 yuan, 81 yuan, and 82 yuan. Inputting "shirt" and 80 yuan into a preset neural network model yields a set of test performance indicators, where the order conversion rate is 0.2 and the click-through rate is 0.2. Inputting "shirt" and 81 yuan into the preset neural network model yields another set of test performance indicators, where the order conversion rate is 0.4 and the click-through rate is 0.3. Inputting "shirt" and 82 yuan into the preset neural network model yields yet another set of test performance indicators, where the order conversion rate is 0.3 and the click-through rate is 0.2. Based on these three sets of test performance indicators, the predicted performance indicators are determined, namely, the predicted performance indicators are an order conversion rate of 0.4 and a click-through rate of 0.3, and the predicted advertising cost is 81 yuan.

[0066] In step 130, the predicted sales cost for each target object is determined based on the predicted advertising expenditure, predicted benefit index, historical advertising expenditure, historical benefit index, and historical sales cost for each target object.

[0067] In the embodiments of this application, historical performance indicators can reflect the performance of the target object in an advertising campaign within a preset time period. The value of historical performance indicators can be the average of the performance indicators of the target object within the preset time period. For example, historical performance indicators can be the average of the performance indicators of the target object for each day from the 3 days to the 18 days before the advertising campaign to be adjusted. That is, historical performance indicators can be the average of the total actual order conversion rate of the target object for each day from the 3 days to the 18 days before the advertising campaign to be adjusted. Historical performance indicators can also be the average of the total actual click-through rate of the target object for each day from the 3 days to the 18 days before the advertising campaign to be adjusted.

[0068] Historical cost of sales can measure the cost of promoting a target audience. Historical cost of sales is related to the ratio of historical advertising expenditure to sales revenue for the target audience. Understandably, the lower the historical cost of sales, the lower the promotion cost for the target audience. Historical cost of sales can be determined based on the average actual advertising sales cost of the target audience within a preset period. For example, historical cost of sales can be the average of the actual sales costs incurred by the target audience each day from the 3 to the 18 days prior to the advertising campaign to be adjusted.

[0069] Forecasted cost of sales measures the cost of promoting a target audience based on forecasted advertising spending. Forecasted cost of sales is related to the ratio of forecasted advertising spending to sales revenue for the target audience. Understandably, the lower the forecasted cost of sales, the lower the forecasted promotion cost for the target audience.

[0070] In the embodiments of this application, step 130, which determines the predicted sales cost for each target object based on the predicted advertising expenditure, predicted benefit indicators, historical advertising expenditure, historical benefit, and historical sales cost for each target object, may include the following steps.

[0071] (1) Determine the first predicted value based on the predicted advertising cost and historical advertising cost for each target object;

[0072] (2) Determine the second predicted value based on the predicted benefit indicators and historical benefit indicators corresponding to each target object;

[0073] (3) Determine the predicted sales cost for each target object based on the historical sales cost, the first predicted value, and the second predicted value.

[0074] The first predicted value can be determined based on the ratio between the predicted advertising expenditure and the historical advertising expenditure for the target object.

[0075] The second predicted value can be determined based on the ratio between the predicted benefit index and the historical benefit index corresponding to the target object.

[0076] Furthermore, the predicted cost of sales can be obtained using the following formula:

[0077] acos2 = acos1 * T * C;

[0078] Where acos2 represents the projected cost of sales; acos1 represents the historical cost of sales; T represents the first projected value; and C represents the second projected value.

[0079] Where T = x1 / x2; C = c1 / c2.

[0080] x1 represents the predicted advertising expenditure; x2 represents the historical advertising expenditure; c1 represents the historical benefit index; c2 represents the predicted benefit index.

[0081] In some implementations, the predicted performance metrics include predicted order conversion rate and predicted click-through rate, while the historical performance metrics include historical order conversion rate and historical click-through rate. The predicted cost of sales for each target object is determined based on its predicted advertising expenditure, predicted order conversion rate, predicted click-through rate, historical advertising expenditure, historical order conversion rate, historical click-through rate, and historical cost of sales. Specifically, the predicted cost of sales can be obtained using the following formula:

[0082] acos2=acos1*(x1 / x2)*[(cr1*ctr1) / (cr2*ctr2)];

[0083] Where cr1 represents the historical order conversion rate; ctr1 represents the historical click-through rate; cr2 represents the predicted order conversion rate; and ctr2 represents the predicted click-through rate.

[0084] In step 140, the budget change for the advertising campaign to be adjusted is determined based on the current budget for the campaign and the total historical advertising expenditure for the campaign to be adjusted.

[0085] In some implementations, the current budget for the advertising campaign to be adjusted can be set by the merchant based on the actual situation.

[0086] In other implementations, the current budget for the advertising campaign to be adjusted can also be determined from the historical total advertising expenditure budget for the advertising campaign to be adjusted.

[0087] For example, when creating an advertising campaign, a daily budget is set. The current budget can be calculated using the historical total spending budget for the preset time period of the campaign to be adjusted. The historical total spending budget can be determined based on the daily budget and the number of days in the preset time period. The daily budget is determined by the merchant based on actual needs when creating the advertising campaign. For instance, the current budget for the advertising campaign to be adjusted could be the total daily budget for the previous 15 days of the campaign.

[0088] Budget changes can reflect the relationship between the current budget for the advertising campaign to be adjusted and the historical total advertising expenditure for the advertising campaign to be adjusted.

[0089] The budget change information can include the direction of budget change and the amount of budget change. The direction of budget change can include budget increase and budget decrease, that is, whether the current budget for the advertising campaign to be adjusted is increasing or decreasing relative to the total historical advertising expenditure of the advertising campaign to be adjusted. When the direction of budget change is budget increase, the amount of budget change can indicate the difference between the current budget for the advertising campaign to be adjusted and the total historical advertising expenditure of the advertising campaign to be adjusted.

[0090] In step 150, the delivery values ​​for multiple target objects are adjusted based on budget changes and the projected sales costs for each target object.

[0091] In this embodiment of the application, the delivery value can indicate the promotion fee corresponding to each target object, that is, the merchant will deliver to the target object based on the delivery value corresponding to the target object.

[0092] Optionally, the campaign value can be achieved through keyword bidding (Bid).

[0093] Based on the relationship between the current budget of the advertising campaign to be adjusted and the total historical advertising expenditure of the advertising campaign to be adjusted, we can determine whether the budget change is an increase or a decrease. The following will describe the two cases of budget change being an increase and budget change being a decrease.

[0094] In some implementations, when the direction of budget change is a budget reduction, step 150 may include the following steps.

[0095] (1) Sort multiple target objects in order of predicted sales cost from low to high.

[0096] This allows us to obtain a sequence of target objects. When at least two target objects have the same predicted sales cost, these objects are sorted in descending order of their historical advertising expenditure. In other words, when at least two target objects have the same predicted sales cost, the higher the historical advertising expenditure, the higher the ranking.

[0097] (2) If the budget change direction is budget reduction, the predicted total advertising expenditure corresponding to the fourth-to-last preset percentage of the multiple target objects will be reduced by a preset percentage.

[0098] (3) Allocate the reduction in total advertising forecast spending to the advertising spending of the top five preset percentage of target objects among multiple target objects.

[0099] If the budget change direction is a budget reduction, that is, the current budget corresponding to the advertising campaign to be adjusted is less than the historical total advertising expenditure corresponding to the advertising campaign to be adjusted, then the predicted advertising expenditure corresponding to the target object ranked lower in the target object sequence will be reduced. At the same time, the reduction in the predicted total advertising expenditure of the target object ranked lower will be allocated to the target object ranked higher in the target object sequence, so that the current budget allocation will favor the target object with lower predicted sales cost.

[0100] The values ​​of the fourth and fifth preset percentages can be adjusted according to actual usage needs. It is understood that the smaller the values ​​of the fourth and fifth preset percentages, the fewer target objects need to be adjusted.

[0101] (4) Determine the adjusted ad spend and predicted clicks for each target object.

[0102] In the embodiments of this application, the reduction of the preset percentage can be set based on the actual situation of the merchant. For example, if the predicted sales costs corresponding to the top-ranked target objects differ little (for example, the magnitude of the difference can be determined by calculating the range), it indicates that the campaign performance corresponding to the top-ranked target objects is not significantly different, and the fifth preset percentage can be appropriately increased. If the predicted sales costs corresponding to the top-ranked target objects differ significantly, it indicates that the campaign performance corresponding to the top-ranked target objects differs significantly, and the fifth preset percentage can be appropriately decreased to allocate the current budget to target objects that can perform better.

[0103] For example, the number of target objects is 100, and the fourth preset percentage and the fifth preset percentage are 30%. That is, the predicted advertising cost of the last 30 target objects in the target object sequence is reduced by a preset percentage according to the corresponding predicted total advertising cost, so as to adjust the predicted advertising cost of these 30 target objects. The reduction amount obtained by reducing the last 30 target objects in the target object sequence by a preset percentage is allocated to the predicted advertising cost of the first 30 target objects in the target object sequence.

[0104] More specifically, assuming the total advertising cost for the last 30 target objects in the target object sequence is 30 yuan, and the preset reduction ratio is 10%, then the reduction in the total advertising cost is 30 * 10% = 3 yuan. This 3 yuan will be allocated to the predicted advertising cost for the first 30 target objects in the target object sequence.

[0105] In some implementations, the specific reduction amount for each target object can be determined based on the ratio of the historical advertising expenditure corresponding to each target object in the fourth preset percentage of the target object sequence to the total predicted advertising expenditure corresponding to the fourth preset percentage of the target object sequence.

[0106] For example, the total historical advertising expenditure A1 corresponding to the fourth preset percentage of target objects ranked in the target object sequence can be determined first (determined by the sum of the historical advertising expenditures of each target object). Then, the ratio C1 of the historical advertising expenditure B1 corresponding to each target object to the total historical advertising expenditure A1 can be determined (C1 = B1 / A1). Based on the predicted reduction amount D1 of the total advertising expenditure and the ratio C1 of each target object, the specific reduction amount E1 = D1 * C1 of each target object can be determined.

[0107] For example, if the number of target objects is 100 and the fourth preset percentage is 30%, and the total predicted advertising cost for the last 30 target objects in the target object sequence is 30 yuan, then the preset reduction ratio is 10%, and the reduction in the total predicted advertising cost is 30 * 10% = 3 yuan. If the historical advertising cost for a target object in the last 30 is 4 yuan, then the specific amount of reduction in the predicted advertising cost associated with that target object is (4 / 30) * 3 = 0.4 yuan, and the adjusted predicted advertising cost for that target object is 4 - 0.4 = 3.6 yuan.

[0108] In other implementations, the reduction in the total advertising cost can be evenly distributed among the target objects ranked in the fourth preset percentage from among multiple target objects. The specific distribution method can be selected according to actual needs.

[0109] Furthermore, based on the target objects ranked in the top fifth preset percentage of the target object sequence, the historical advertising expenditure associated with each target object ranked in the top fifth preset percentage is determined; based on the historical advertising expenditure associated with each target object ranked in the top fifth preset percentage, the total historical advertising expenditure associated with each target object ranked in the top fifth preset percentage is determined.

[0110] Based on the ratio between the historical advertising expenditure associated with each target object ranked in the top fifth preset percentage and the total historical advertising expenditure associated with each target object ranked in the top fifth preset percentage, and the amount of reduction in the total predicted advertising expenditure, the specific amount of the increase in predicted advertising expenditure associated with each target object ranked in the top fifth preset percentage is determined.

[0111] For example, the total historical advertising expenditure A2 corresponding to the target object ranked in the top five preset percentages in the target object sequence can be determined first (determined by the sum of the historical advertising expenditures of each target object). Then, the ratio of the historical advertising expenditure B2 corresponding to each target object to the total historical advertising expenditure A2 can be determined (C2 = B2 / A2). Based on the predicted reduction in total advertising expenditure D2 and the ratio of each target object C2, the specific increase amount E2 = D2 * C2 for each target object can be determined.

[0112] For example, the number of target objects is 100, the fifth preset percentage is 30%, assuming that the total historical advertising cost of the first 30 target objects in the target object sequence is 30 yuan, assuming that the reduction in the predicted total advertising cost is 3 yuan, and the historical advertising cost associated with a target object ranked in the first 30 is 4, then the specific amount of increase in the predicted advertising cost associated with the target object is (4 / 30)*3 = 0.4 yuan, that is, the predicted advertising cost associated with the target object after adjustment is 0.4 + 4 = 4.4 yuan.

[0113] It is understood that in other implementations, multiple target objects can be sorted in descending order, and the subsequent calculation steps can be adjusted accordingly. For example, the total advertising prediction cost corresponding to the target object ranked in the "first" fourth preset percentage can be reduced by a preset percentage, and other steps are similar.

[0114] In embodiments of this application, the step of determining the adjusted delivery value for each target object based on the adjusted advertising spending budget and predicted click volume for each target object may include the following steps.

[0115] (1) Determine the adjustment value based on the original delivery value and predicted click volume for each target object.

[0116] (2) Determine the adjusted spending value for each target object based on the adjusted predicted advertising spending and adjustment value for each target object.

[0117] In the embodiments of this application, the delivery value of each target object before adjustment can be obtained by retrieving the report data of the advertising campaign to be adjusted.

[0118] In this embodiment, the predicted click volume for each target object can be determined based on the predicted advertising expenditure, predicted sales cost, unit price of the product, and predicted order conversion rate for each target object. Specifically, this may include the following steps.

[0119] (a) Determine the corresponding sales revenue based on the projected advertising spending and projected sales costs for each target audience.

[0120] (b) Determine the order quantity based on the sales revenue of each target audience and the unit price of the product corresponding to the advertising campaign to be adjusted.

[0121] (c) Determine the predicted click volume based on the order volume and predicted order conversion rate for each target object.

[0122] The predicted click volume can be obtained using the following formula:

[0123] sales = costs / acos;

[0124] orders = sales / price;

[0125] clicks = orders / cr;

[0126] Where sales represents sales revenue; costs represents predicted advertising spending; orders represents order volume; price represents the unit price of the product corresponding to the advertising campaign to be adjusted; clicks represents predicted clicks; and cr represents predicted order conversion rate.

[0127] Sales revenue, order volume, and unit price of goods can be obtained by retrieving report data from the advertising campaign to be adjusted.

[0128] Furthermore, the adjustment value is determined based on the pre-adjustment spending and predicted click volume for each target object; then, the adjusted spending value for each target object is determined based on the post-adjustment advertising cost and the adjustment value; specifically, the adjusted spending value for each target object can be obtained through the following formula.

[0129] y = bid1 / clicks;

[0130]

[0131] Where y represents the adjustment value, bid1 represents the target object's delivery value before adjustment, clicks represents the predicted click volume, bid represents the target object's delivery value after adjustment, and costs represents the target object's predicted advertising cost after adjustment.

[0132] In some other implementations, when the budget change direction is an increase in the budget, step 150 may include the following steps.

[0133] (1) Sort multiple target objects in order of predicted sales cost from low to high.

[0134] This step can be referred to in the aforementioned embodiments, and will not be repeated here.

[0135] (2) If the direction of budget change is budget increase, then the proportion of historical expenditure to total historical expenditure among multiple target objects is the target object with the first preset percentage.

[0136] (3) Determine the adjusted ad spend and predicted clicks for each target object.

[0137] The steps “sorting multiple target objects from low to high according to the predicted sales cost” and “determining the adjusted delivery value for each target object based on the adjusted predicted advertising cost and predicted click volume for each target object” can be described in detail in the foregoing embodiments, and will not be repeated here.

[0138] In the embodiments of this application, if the current budget corresponding to the advertising campaign to be adjusted and the budget corresponding to the total historical advertising expenditure of the advertising campaign to be adjusted change in the direction of budget increase, that is, when the current budget corresponding to the advertising campaign to be adjusted is greater than the total historical advertising expenditure of the advertising campaign to be adjusted, the budget change (the difference between the current budget and the total historical advertising expenditure of the advertising campaign to be adjusted) is allocated to the target objects among multiple target objects whose proportion of historical advertising expenditure to total historical advertising expenditure is a first preset percentage.

[0139] In some implementations, multiple target objects can be sorted according to their historical advertising expenditure (e.g., from high to low or from low to high). Selection can begin with the target object with the highest historical advertising expenditure, choosing those whose historical advertising expenditure represents a percentage of the total historical advertising expenditure. For example, if the total historical advertising expenditure for a target object is 100 yuan, and the first preset percentage is 50%, and the historical advertising expenditures of the multiple target objects, from highest to lowest, are: Target Object 1 (30 yuan), Target Object 2 (20 yuan), Target Object 3 (15 yuan), Target Object 4 (14 yuan), Target Object 5 (10 yuan), Target Object 6 (6 yuan), and Target Object 7 (5 yuan), then the budget change is allocated to Target Object 1 and Target Object 2 (30 yuan + 20 yuan = 50 yuan).

[0140] In other implementations, the historical advertising expenditures corresponding to the target objects are accumulated sequentially according to the target object sequence (multiple target objects are sorted in order of predicted sales cost from low to high) until the accumulated value is equal to the total historical advertising expenditures associated with each target object in the target object sequence, which is a first preset percentage. Then, the target objects whose predicted advertising expenditures need to be adjusted are identified. Based on the historical advertising expenditures associated with the target objects whose predicted advertising expenditures need to be adjusted, the total historical expenditures of the target objects whose predicted advertising expenditures need to be adjusted are determined. Finally, the budget change is allocated to the target objects according to the ratio between the historical advertising expenditures of the target objects whose predicted advertising expenditures need to be adjusted and the total historical expenditures.

[0141] For example, if the historical advertising spending associated with the target objects in the target object sequence is as follows: Target Object 1 (30 yuan), Target Object 2 (20 yuan), Target Object 3 (15 yuan), Target Object 4 (14 yuan), Target Object 5 (10 yuan), Target Object 6 (6 yuan), and Target Object 7 (5 yuan), then the budget change will be allocated to Target Object 1 and Target Object 2 (30 yuan + 20 yuan = 50 yuan).

[0142] For example, the current budget for the advertising campaign to be adjusted is 120 yuan, and the total historical advertising expenditure for the advertising campaign to be adjusted is 100 yuan. Since the current budget of 120 yuan for the advertising campaign to be adjusted is greater than the total historical advertising expenditure for the advertising campaign to be adjusted is 100 yuan, the direction of budget change is determined to be budget increase, and the budget change amount is determined to be 20 yuan (120 yuan - 100 yuan = 20 yuan). The budget change amount of 20 yuan is allocated to the target objects in the target object sequence whose historical expenditure accounts for the proportion of the total historical expenditure is a first preset percentage.

[0143] For example, the number of target objects in the target object sequence is 100, and the first preset percentage is 50%. Based on the historical advertising expenditure corresponding to each target object in the target object sequence, the total historical advertising expenditure w1 is determined. The historical advertising expenditure associated with the target objects is accumulated according to the order of the target object sequence (sorting multiple target objects in order of predicted sales cost from low to high). When the accumulated value accounts for 50% of the total historical advertising expenditure w1, these target objects are determined as objects for which the predicted advertising expenditure needs to be adjusted. Based on the historical advertising expenditure associated with each of these target objects for which the predicted advertising expenditure needs to be adjusted, the total historical advertising expenditure w2 corresponding to the target objects for which the predicted advertising expenditure needs to be adjusted is determined. The budget change is allocated to the target objects for which the predicted advertising expenditure needs to be adjusted based on the ratio of the historical advertising expenditure associated with each target object for which the predicted advertising expenditure needs to be adjusted to the historical advertising expenditure w2.

[0144] For example, if there are 120 target objects in the target object sequence, the first preset percentage is 50%, and the target objects whose predicted advertising spending needs to be adjusted are the first 40 in the target object sequence, and the historical advertising spending associated with a certain target object in these first 40 target objects is 2 yuan, the total historical advertising spending associated with the first 40 target objects is 100 yuan, and the budget change is 20 yuan, then the allocation share of this target object in the budget change is (2 / 100)*20 = 0.4. The adjusted advertising spending budget for this target object is the sum of 0.4 and the predicted advertising spending.

[0145] In some implementations, the budget change can be distributed equally among the target objects in the target object sequence whose predicted advertising spending needs to be adjusted; in other implementations, the budget change can be distributed in other proportions among the target objects in the target object sequence whose predicted advertising spending needs to be adjusted.

[0146] After optimizing the budgeted advertising spending for the target objects as described above, in order to further favor the target objects with better advertising performance, in some embodiments, the predicted total advertising spending of the target objects ranked in the second-to-last preset percentage in the target object sequence is reduced by a preset percentage and the reduction is allocated to the target objects ranked in the first-to-last preset percentage in the target object sequence.

[0147] In the embodiments of this application, the specific description of reducing the preset percentage can be referred to the foregoing embodiments, and will not be repeated here. The values ​​of the second preset percentage and the third preset percentage can be adjusted according to actual usage needs. It is understood that the smaller the values ​​of the second preset percentage and the third preset percentage, the fewer the number of target objects that need to be adjusted. For details, please refer to the foregoing embodiments, and will not be repeated here.

[0148] In some implementations, the specific reduction amount for each target object can be determined based on the ratio of the historical advertising expenditure corresponding to each target object in the second-to-last preset percentage of the target object sequence to the total predicted advertising expenditure corresponding to the target objects in the second-to-last preset percentage of the target object sequence.

[0149] For example, the total historical advertising expenditure A3 corresponding to the target object ranked second to last in the target object sequence can be determined first (determined by the sum of the historical advertising expenditures of each target object). Then, the ratio of the historical advertising expenditure B3 corresponding to each target object to the total historical advertising expenditure A3 can be determined (C3 = B3 / A3). Based on the predicted reduction in total advertising expenditure D3 and the ratio of each target object C3, the specific reduction amount E3 = D3 * C3 for each target object can be determined.

[0150] For example, assuming the target object sequence has 100 target objects, the second preset threshold is 30%, and the last 30 target objects in the target object sequence have a total advertising cost of 30 yuan, with a preset reduction ratio of 10%, then the reduction in total advertising cost is 30 * 10% = 3 yuan. If the historical advertising cost of a certain target object in the last 30 target objects in the target object sequence is 2 yuan and the corresponding predicted advertising cost is 2.1 yuan, then the specific reduction in the predicted advertising cost of this target object is (2 / 30) * 3 = 0.2 yuan. The adjusted predicted advertising cost of this target object is 2.1 - 0.2 = 1.9 yuan.

[0151] In other implementations, the reduction in total advertising cost can be evenly distributed among the target objects ranked last by a second preset percentage. The specific distribution method can be selected according to actual needs.

[0152] Furthermore, based on the target objects ranked in the top third preset percentage of the target object sequence, the historical advertising expenditure associated with each target object ranked in the top third preset percentage is determined; based on the historical advertising expenditure associated with each target object ranked in the top third preset percentage, the total historical advertising expenditure associated with each target object ranked in the top third preset percentage is determined.

[0153] Based on the ratio between the historical advertising expenditure associated with each target object ranked in the top third preset percentage and the total historical advertising expenditure associated with each target object ranked in the top third preset percentage, and the amount of reduction in the total predicted advertising expenditure, the specific amount of the increase in predicted advertising expenditure associated with each target object ranked in the top third preset percentage is determined.

[0154] For example, the total historical advertising expenditure A4 corresponding to the target object ranked in the top third preset percentage in the target object sequence can be determined first (determined by the sum of the historical advertising expenditures of each target object). Then, the ratio of the historical advertising expenditure B4 corresponding to each target object to the total historical advertising expenditure A4 can be determined (C4 = B4 / A4). Based on the predicted reduction in total advertising expenditure D4 and the ratio of each target object C4, the specific increase amount E4 = D4 * C4 for each target object can be determined.

[0155] For example, the number of target objects is 100, the third preset percentage is 30%, assuming that the total historical advertising cost of the first 30 target objects in the target object sequence is 30 yuan, assuming that the reduction in the total predicted advertising cost is 3 yuan, and the historical advertising cost associated with a target object ranked in the first 30 is 4, then the specific amount of increase in the predicted advertising cost associated with the target object is (4 / 30)*3 = 0.4 yuan, that is, the predicted advertising cost associated with the target object after adjustment is 0.4+4 = 4.4 yuan.

[0156] The ad placement values ​​for multiple target objects are optimized and adjusted based on budget changes and the predicted sales costs corresponding to each target object, so that the budget amount is directed towards target objects with better advertising performance, thereby improving advertising effectiveness. To facilitate user adjustments to the placement values, in some embodiments, the ad placement adjustment method provided in this application further includes displaying an adjustment interface based on the adjusted placement values.

[0157] In some implementations, the adjustment interface can display the deployment value before and after the adjustment, making it convenient for users to compare the adjustment of the deployment value.

[0158] In some implementations, the interface may also include adjustment controls, which allow users to determine whether to use the adjusted delivery value.

[0159] In some implementations, the display adjustment interface can also specify the type of target object, such as a keyword object or a search object. Since search objects are not included in the objects set during campaign creation, users can add search objects to the keyword list to expand the keyword list and improve campaign performance. Optionally, the adjustment interface may also include an add control, allowing merchants to determine whether to add search objects as keyword objects.

[0160] Please see Figure 2 , Figure 2 The diagram illustrates an adjustment interface provided in an embodiment of this application, as shown below. Figure 2 As shown, the adjustment interface 200 displays target object 1, target object 2, and target object 3.

[0161] Among them, the type of target object 1 is a key object. The corresponding investment value of target object 1 before adjustment is 100 yuan, and the corresponding investment value after adjustment is 80 yuan, which is a reduction of 20 yuan (100-80=20 yuan). If the user wants to use the adjusted investment value, he / she can update the investment value associated with target object 1 to the corresponding investment value after adjustment through the adjustment control 210A. That is, the user can change the current investment value of target object 1 from 100 yuan before adjustment to 80 yuan after adjustment through the adjustment control 210A.

[0162] The type of target object 2 is search object. The corresponding target value after adjustment is 100 yuan. If the user wants to add target object 2 to the key objects, the type of target object 2 can be changed from search object to key object by adding control 220.

[0163] The target object 3 is a key object. The payout value for target object 3 before adjustment was 100 yuan, and the payout value after adjustment is 120 yuan, an increase of 20 yuan (120-100=20 yuan). If the user wants to use the adjusted payout value, they can use the adjustment control 210B to update the payout value associated with target object 3 to the adjusted payout value. That is, the user can use the adjustment control 210B to change the current payout value of target object 3 from 100 yuan before adjustment to 120 yuan after adjustment.

[0164] It is understood that this application is not limited to this. The adjustment interface may also display other content, such as the adjustment range, historical report data of the target object, etc. The layout of the adjustment interface may also be other forms of layout. The specific layout can be adjusted according to the actual use needs. This application does not limit this.

[0165] Therefore, by combining the prediction neural network model with multiple target audiences and corresponding advertising budget ranges in the advertising campaign to be adjusted, multiple sets of predicted order conversion rates and predicted click-through rates (CTRs) can be predicted, reflecting the advertising effectiveness of the target audiences in the campaign. Based on these multiple sets of CTRs, the optimal advertising expenditure within the budget range is determined, i.e., the predicted advertising expenditure is determined. Furthermore, based on the predicted advertising expenditure and its corresponding CTRs, the predicted sales cost, which reflects the advertising efficiency of the target audiences in the campaign to be adjusted, is obtained. Furthermore, based on predicted sales costs, multiple target audiences are ranked, and the ad spend for each target audience in the campaign to be adjusted is optimized according to the current budget and corresponding budget changes. More specifically, when the budget is increasing, the predicted ad spend for the target audience whose historical spending accounts for a percentage of total historical spending is increased by a first preset percentage. After this first round of adjustment, the predicted ad spend for the second-lowest preset percentage of target audiences is decreased, while the predicted ad spend for the third-highest preset percentage is increased. When the budget is decreasing, the predicted ad spend for the fourth-lowest preset percentage of target audiences is decreased, while the predicted ad spend for the fifth-highest preset percentage is increased. This ensures that the limited budget is allocated towards target audiences with lower promotion costs, achieving a reasonable adjustment of ad spending and maximizing its effectiveness.

[0166] Please see Figure 3 , Figure 3 An advertising placement adjustment device 300 provided in an embodiment of this application is shown. The device includes: an acquisition module 310, a first determination module 320, a second determination module 330, a third determination module 340, and an adjustment module 340. Specifically:

[0167] The acquisition module 310 is used to acquire multiple target objects corresponding to the advertising campaign to be adjusted, as well as the advertising spending budget range corresponding to each target object.

[0168] In some embodiments, the first acquisition module 310 may further include:

[0169] The first acquisition submodule is used to acquire multiple key objects of the advertising campaign to be adjusted;

[0170] The second acquisition submodule is used to acquire multiple search objects corresponding to the advertising campaign to be adjusted;

[0171] The deduplication module is used to deduplicatize the set of multiple key objects and multiple search objects to obtain multiple target objects;

[0172] The third acquisition submodule is used to obtain the historical advertising expenditure corresponding to each target object;

[0173] The budget range determination module is used to determine the advertising spending budget range for each target object based on its historical advertising spending and budget range.

[0174] The first determining module 320 is used to determine the predicted advertising expenditure and predicted benefit index for each target object based on the predictive neural network model, the multiple target objects, and the advertising expenditure budget range corresponding to each target object.

[0175] In some embodiments, the first determining module 320 may further include:

[0176] The first determination submodule is used to determine multiple advertising expenditures to be tested from the advertising expenditure budget range corresponding to each target object;

[0177] The second determining submodule is used to input each advertising expenditure to be tested into the preset neural network model to obtain the expected benefit index corresponding to each advertising expenditure to be tested.

[0178] The third determination submodule is used to determine the predicted benefit indicators from multiple test benefit indicators corresponding to each target object;

[0179] The fourth determination submodule is used to take the advertising expenditure budget corresponding to the predicted benefit index of each target object as the predicted advertising expenditure for each target object.

[0180] The second determining module 330 is used to determine the predicted sales cost corresponding to each target object based on the predicted advertising expenditure, the predicted benefit index, the historical advertising expenditure, the historical benefit index, and the historical sales cost corresponding to each target object.

[0181] In some embodiments, the second determining module 330 may further include:

[0182] The first prediction value determination module is used to determine the first prediction value based on the predicted advertising cost and the historical advertising cost corresponding to each target object.

[0183] The second prediction value determination module is used to determine the second prediction value based on the prediction benefit index corresponding to each target object and the historical benefit index.

[0184] The second determination submodule is used to determine the predicted sales cost corresponding to each target object based on the historical sales cost, the first predicted value, and the second predicted value corresponding to each target object.

[0185] The third determining module 340 is used to determine the budget change of the advertising campaign to be adjusted based on the current budget of the advertising campaign to be adjusted and the historical total advertising expenditure of the advertising campaign to be adjusted.

[0186] The budget change information may include the direction of budget change and a preset change amount. The total historical advertising spending budget corresponding to the advertising method is used as the current budget.

[0187] In some embodiments, the third determining module 340 may further include:

[0188] The budget increase determination module is used to determine that if the current budget corresponding to the advertising campaign to be adjusted is greater than the total historical advertising expenditure corresponding to the advertising campaign to be adjusted, the direction of the budget change for the advertising campaign to be adjusted is budget increase;

[0189] In some embodiments, the budget increase determination module may further include:

[0190] The preset change amount determination module is used to take the difference between the current preset and the total historical advertising expenditure as the budget change amount.

[0191] The budget reduction determination is used to determine the direction of budget change in the advertising placement adjustment method as budget reduction if the current budget corresponding to the advertising campaign to be adjusted is less than the historical total advertising expenditure corresponding to the advertising campaign to be adjusted.

[0192] The adjustment module 350 is used to adjust the delivery value of the multiple target objects based on the budget changes and the predicted sales costs corresponding to the multiple target objects.

[0193] In some embodiments, the adjustment module 350 may further include:

[0194] The first sorting module is used to sort the multiple target objects in order of predicted sales cost from low to high.

[0195] The first allocation module is used to allocate the budget change amount to the target objects among the plurality of target objects whose historical spending accounts for a first preset percentage of the total historical spending if the direction of the budget change is budget increase.

[0196] In some embodiments, the first allocation model may further include:

[0197] The second allocation module is used to allocate the budget change to the target objects whose historical spending accounts for a first preset percentage of the total historical spending among the multiple target objects, and to reduce the predicted total advertising spending of the target objects ranked in the second preset percentage among the multiple target objects by a preset percentage and allocate the reduction amount to the target objects ranked in the first third preset percentage among the multiple target objects.

[0198] The first delivery value determination module is used to determine the adjusted delivery value for each target object based on the adjusted predicted advertising cost and the corresponding predicted click volume for each target object.

[0199] In some embodiments, the adjustment module 350 may further include:

[0200] The second sorting module is used to sort the multiple target objects in order of predicted sales cost from low to high.

[0201] In some embodiments, the second sorting module and the first sorting module may further include:

[0202] The first sorting submodule is used to sort the multiple target objects in order of predicted sales cost from low to high.

[0203] The second sorting submodule is used to sort target objects with the same predicted sales cost in descending order of historical advertising expenditure if at least two target objects have the same predicted sales cost.

[0204] The fourth allocation module is used to reduce the predicted total advertising cost of the target object ranked last by the fourth preset percentage among the multiple target objects by a preset percentage if the budget change direction is budget reduction;

[0205] The fifth allocation module is used to allocate the reduction in the total predicted advertising cost to the target objects ranked in the top fifth preset percentage among the multiple target objects;

[0206] The second delivery value determination module is used to determine the adjusted delivery value for each target object based on the adjusted predicted advertising cost and the corresponding predicted click volume for each target object.

[0207] In some embodiments, the second delivery value determination module and the first delivery value determination module may further include:

[0208] The adjustment value determination submodule is used to determine the adjustment value based on the pre-adjustment delivery value and the predicted click volume for each target object.

[0209] Among them, the predicted benefit indicators include the predicted order conversion rate;

[0210] In some embodiments, the adjustment value determination submodule may include:

[0211] The sales revenue determination module is used to determine the corresponding sales revenue based on the predicted advertising expenditure and the predicted sales cost for each target object.

[0212] The order quantity determination module is used to determine the order quantity based on the sales revenue of each target object and the unit price of the product corresponding to the advertising campaign to be adjusted.

[0213] The predicted click volume determination module is used to determine the predicted click volume based on the order volume and the predicted order conversion rate for each target object.

[0214] The delivery value determination submodule is used to determine the adjusted delivery value for each target object based on the adjusted predicted advertising cost and the adjustment value for each target object.

[0215] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described device and module can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0216] In the several embodiments provided in this application, the coupling or direct coupling or communication connection between the modules shown or discussed may be an indirect coupling or communication connection through some interface, device or module, and may be electrical, mechanical or other forms.

[0217] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0218] Please see Figure 4 Based on the above-described advertising placement adjustment method, this application embodiment also provides an electronic device 400 that can execute the aforementioned advertising placement adjustment method.

[0219] In embodiments of this application, the electronic device 400 includes one or more processors 410, a memory 420, and one or more application programs. The one or more application programs are stored in the memory 420, which stores programs capable of executing the contents of the foregoing embodiments, and the processor 410 can execute the programs stored in the memory.

[0220] The processor 410 may include one or more cores for data processing and message matrix units. The processor 410 connects to various parts of the electronic device using various interfaces and lines, and performs various functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in memory, and by calling data stored in memory. Optionally, the processor 410 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 410 may integrate one or more of a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), and a modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the displayed content; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 410 and may be implemented separately using a communication chip.

[0221] The memory 420 may include random access memory (RAM) or read-only memory (ROM). The memory 420 can be used to store instructions, programs, code, code sets, or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function, instructions for implementing the various method embodiments described below, etc. The data storage area may also store data created by the terminal during use.

[0222] Please see Figure 5 , Figure 5 A schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of this application is shown. The computer-readable storage medium 500 stores program code, which can be called by a processor to execute the advertising placement adjustment method described in the above method embodiment.

[0223] The computer-readable storage medium 500 may be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Optionally, the computer-readable storage medium 500 includes a non-transitory computer-readable storage medium. The computer-readable storage medium 500 has storage space for program code 510 that performs any of the method steps described above. This program code can be read from or written to one or more computer program devices. The program code 510 may, for example, be compressed in a suitable form.

[0224] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for adjusting advertising placement, characterized in that, The method includes: Obtain multiple target objects corresponding to the advertising campaign to be adjusted, and the advertising spending budget range corresponding to each target object; the target object is an object associated with the product in the advertising campaign to be adjusted, and users can search for the product in the advertising campaign corresponding to the target object by searching the target object through the e-commerce platform; Based on the predictive neural network model, the multiple target objects, and the advertising spending budget range corresponding to each target object, the predicted advertising spending and predicted benefit indicators for each target object are determined. The predicted sales cost for each target object is determined based on the predicted advertising expenditure, the predicted benefit index, the historical advertising expenditure, the historical benefit index, and the historical sales cost. The budget change for the advertising campaign to be adjusted is determined based on the current budget for the campaign to be adjusted and the total historical advertising expenditure for the campaign to be adjusted. The allocation values ​​for the multiple target objects are adjusted based on the budget changes and the projected sales costs corresponding to the multiple target objects. The budget changes include both the direction and the amount of the budget changes. The adjustment of the allocation value for the multiple target objects based on the budget changes and the predicted sales costs corresponding to the multiple target objects includes: The multiple target objects are sorted in order of predicted sales cost from low to high; If the direction of the budget change is an increase in the budget, then the amount of the budget change is allocated to the target objects among the plurality of target objects whose historical spending accounts for a first preset percentage of the total historical spending; The adjusted ad spend and the corresponding predicted click volume for each target object are used to determine the ad delivery value for each target object. The step of adjusting the allocation value for the multiple target objects based on the budget changes and the predicted sales costs corresponding to the multiple target objects further includes: The multiple target objects are sorted in order of predicted sales cost from low to high; If the budget change direction is a budget reduction, then the total predicted advertising cost corresponding to the fourth preset percentage of the target objects ranked last among the multiple target objects will be reduced by a preset percentage. The reduction in the predicted total spending of the advertising is allocated to the target objects ranked in the top fifth preset percentage among the multiple target objects; The adjusted ad spend and the corresponding predicted click volume for each target object are used to determine the adjusted ad spend for each target object.

2. The method according to claim 1, characterized in that, The process of obtaining multiple target objects corresponding to the advertising campaign to be adjusted, and the advertising spending budget range corresponding to each target object, includes: Obtain multiple key targets for the advertising campaign to be adjusted; Obtain multiple search objects corresponding to the advertising campaign to be adjusted; The set of multiple key objects and multiple search objects is deduplicated to obtain multiple target objects; Get the historical advertising cost for each target object; Determine the advertising budget range for each target object based on its historical advertising spending and budget range.

3. The method according to claim 1, characterized in that, The determination of the predicted advertising cost and predicted effectiveness indicators for each target object based on the predictive neural network model, the multiple target objects, and the advertising cost budget range corresponding to each target object includes: Multiple ad spends to be tested are determined from the ad spend budget range corresponding to each target object; Each advertising expenditure to be tested is input into the predictive neural network model to obtain the expected benefit index corresponding to each advertising expenditure to be tested. Determine the predicted benefit indicators from multiple measured benefit indicators corresponding to each target object; The predicted advertising expenditure budget corresponding to the predicted benefit index for each target object is taken as the predicted advertising expenditure for each target object.

4. The method according to claim 1, characterized in that, The step of determining the predicted sales cost for each target object based on the predicted advertising expenditure, the predicted effectiveness index, historical advertising expenditure, historical effectiveness index, and historical sales cost includes: A first predicted value is determined based on the predicted advertising cost and the historical advertising cost for each target object; The second predicted value is determined based on the predicted benefit index and the historical benefit index corresponding to each target object; The projected sales cost for each target object is determined based on its historical sales cost, first forecast, and second forecast.

5. The method according to claim 1, characterized in that, The allocation of the budget change to the target objects among the plurality of target objects whose historical spending accounts for a first preset percentage of total historical spending includes: The budget change is allocated to the target objects whose historical spending accounts for a first preset percentage of the total historical spending among the multiple target objects, and the predicted total advertising spending of the target objects ranked in the second preset percentage among the multiple target objects is reduced by a preset percentage and the reduction is allocated to the target objects ranked in the first third preset percentage among the multiple target objects.

6. The method according to claim 1 or 5, characterized in that, The step of sorting the multiple target objects according to the predicted sales cost from low to high includes: The multiple target objects are sorted in order of predicted sales cost from low to high; If at least two target objects have the same predicted sales cost, then the target objects with the same predicted sales cost are sorted in descending order of historical advertising expenditure.

7. The method according to claim 1 or 5, characterized in that, The step of determining the adjusted ad spend and corresponding predicted clicks for each target object includes: The adjustment value is determined based on the pre-adjustment delivery value and the predicted click volume for each target object; The adjusted delivery value for each target object is determined based on the adjusted predicted advertising cost for each target object and the adjusted value.

8. The method according to claim 7, characterized in that, The predicted benefit indicators include the predicted order conversion rate; the method further includes: The corresponding sales revenue is determined based on the predicted advertising expenditure and the predicted sales cost for each of the target objects; The order quantity is determined based on the sales revenue of each target object and the unit price of the product corresponding to the advertising campaign to be adjusted; The predicted click volume is determined based on the order volume and the predicted order conversion rate for each target object.

9. The method according to claim 1, characterized in that, The step of determining the budget change for the advertising campaign to be adjusted based on the current budget and the historical total advertising expenditure of the advertising campaign to be adjusted includes: If the current budget for the advertising campaign to be adjusted is greater than the total historical advertising expenditure for the advertising campaign to be adjusted, then the direction of budget change for the advertising campaign to be adjusted is determined to be budget increase; If the current budget for the advertising campaign to be adjusted is less than the total historical advertising expenditure for the advertising campaign to be adjusted, then the budget change direction of the advertising placement adjustment method is determined to be budget reduction.

10. The method according to claim 9, characterized in that, When the budget change direction is budget increase, if the current budget corresponding to the advertising campaign to be adjusted is greater than the historical total advertising expenditure corresponding to the advertising campaign to be adjusted, then the budget change direction of the advertising campaign to be adjusted is determined to be budget increase; including: If the current budget for the advertising campaign to be adjusted is greater than the total historical advertising expenditure for the advertising campaign to be adjusted, then the direction of budget change for the advertising campaign to be adjusted is determined to be budget increase; The difference between the current budget and the total historical advertising expenditure is used as the budget change.

11. The method according to claim 1, characterized in that, The method further includes using the historical total advertising spending budget corresponding to the advertising campaign to be adjusted as the current budget.

12. An advertising placement adjustment device, characterized in that, The device includes: The acquisition module is used to acquire multiple target objects corresponding to the advertising campaign to be adjusted, and the advertising spending budget range corresponding to each target object; the target object is an object associated with the product in the advertising campaign to be adjusted, and users can search for the product in the advertising campaign corresponding to the target object by searching the target object through the e-commerce platform; The first determining module is used to determine the predicted advertising expenditure and predicted benefit index for each target object based on the predictive neural network model, the multiple target objects, and the advertising expenditure budget range corresponding to each target object; The second determining module is used to determine the predicted sales cost corresponding to each target object based on the predicted advertising expenditure, the predicted benefit index, the historical advertising expenditure, the historical benefit index, and the historical sales cost for each target object. The third determining module is used to determine the budget change of the advertising campaign to be adjusted based on the current budget and the historical total advertising expenditure of the advertising campaign to be adjusted; the budget change includes the direction of budget change and the amount of budget change; adjusting the ad spend of the multiple target objects based on the budget change and the predicted sales cost of the multiple target objects includes: sorting the multiple target objects in order of rising predicted sales cost; if the budget change direction is an increase in budget, allocating the budget change to the target objects whose historical expenditure accounts for a first preset percentage of the total historical expenditure; and adjusting the predicted advertising expenditure of each target object and the budget change based on the current budget and the predicted sales cost of the multiple target objects. The corresponding predicted click volume determines the adjusted ad spend for each target object; adjusting the ad spend for the multiple target objects based on budget changes and the predicted sales costs corresponding to the multiple target objects further includes: sorting the multiple target objects in ascending order of predicted sales costs; if the budget change direction is a budget reduction, reducing the predicted total advertising expenditure of the target objects ranked in the bottom fourth preset percentage among the multiple target objects by a preset percentage; allocating the reduction in the predicted total advertising expenditure to the target objects ranked in the top fifth preset percentage among the multiple target objects; and determining the adjusted ad spend for each target object based on the adjusted predicted advertising expenditure and the corresponding predicted click volume. The adjustment module is used to adjust the allocation value of the multiple target objects based on the budget changes and the predicted sales costs corresponding to the multiple target objects.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains program code, which can be called by a processor to execute the advertising placement adjustment method as described in any one of claims 1-11.

Images