Advertisement putting method and device, equipment and medium

Abstract

The invention relates to an advertisement putting method and device, equipment and a medium, and the method comprises the steps: obtaining the current advertisement putting data of a target advertisement in response to an advertisement putting request of the target advertisement; comparing the advertisement putting data with different advertisement effect stages divided by adopting advertisement effect indexes in advance, and determining a current advertisement effect stage of the target advertisement; wherein different revenue strategies are matched in advance in different advertisement effect stages; according to a target revenue strategy matched with the current advertisement effect stage, determining an estimated revenue value of putting the target advertisement; and determining the putting cost of the target advertisement according to the pre-estimated profit value. According to the invention, the accuracy of pre-estimating the profit value can be improved.

Description

Technical Field

[0001] The present disclosure relates to the field of advertising technology, and in particular to an advertisement delivery method, device, equipment and medium. Background Art

[0002] Performance advertising is a form of advertising based on performance, and is usually priced according to measurable behaviors (such as clicks, downloads, and purchases). When advertisers place performance ads on an advertising platform, they need to go through multiple stages from the start of new ad placement to the maturity of the advertising data. However, in each stage before the advertising data matures, inaccurate click-through rate and conversion rate estimates will occur due to the lack of account characteristics and advertising characteristics. Correspondingly, the ECPM (effective cost per mille) of an ad calculated based on inaccurate click-through rate and conversion rate will also be inaccurate, and ECPM will affect advertisers' control over delivery costs. Therefore, inaccurate ECPM calculations can easily lead to unreasonable control of advertisers' delivery costs, resulting in a mismatch between the actual delivery cost of an ad and the actual conversion effect of the ad. Summary of the invention

[0003] In order to solve the above technical problems, the present disclosure provides an advertisement delivery method, device, equipment and medium to solve the problem in the related art that inaccurate ECPM calculation easily leads to unreasonable control of advertiser delivery costs.

[0004] According to one aspect of the present disclosure, there is provided an advertisement delivery method, comprising:

[0005] In response to an advertisement delivery request of a target advertisement, obtaining current advertisement delivery data of the target advertisement;

[0006] Compare the advertisement delivery data with different advertisement effect stages pre-divided by advertisement effect indicators to determine the current advertisement effect stage of the target advertisement; wherein different advertisement effect stages are pre-matched with different revenue strategies;

[0007] Determine the estimated revenue value of delivering the target advertisement according to the target revenue strategy matched to the current advertisement effect stage;

[0008] The delivery cost of the target advertisement is determined according to the estimated revenue value.

[0009] According to another aspect of the present disclosure, there is also provided an advertisement delivery device, comprising:

[0010] A data acquisition module, configured to respond to an advertisement delivery request of a target advertisement and acquire current advertisement delivery data of the target advertisement;

[0011] A stage determination module, used to compare the advertisement delivery data with different advertisement effect stages pre-divided by advertisement effect indicators, and determine the current advertisement effect stage of the target advertisement; wherein different advertisement effect stages are pre-matched with different revenue strategies;

[0012] A revenue determination module, used to determine the estimated revenue value of delivering the target advertisement according to the target revenue strategy matched to the current advertisement effect stage;

[0013] The cost determination module is used to determine the delivery cost of the target advertisement according to the estimated revenue value.

[0014] According to another aspect of the present disclosure, an electronic device is provided, the electronic device comprising:

[0015] processor;

[0016] a memory for storing instructions executable by the processor;

[0017] The processor is used to read the executable instructions from the memory and execute the instructions to implement the above method.

[0018] According to another aspect of the present disclosure, a computer-readable storage medium is provided, wherein the storage medium stores a computer program, and the computer program is used to execute the above method.

[0019] Compared with the prior art, the technical solution provided by the embodiments of the present disclosure has the following advantages:

[0020] The technical solution provided by the embodiment of the present disclosure includes: first responding to an advertisement delivery request of a target advertisement to obtain current advertisement delivery data of the target advertisement; then comparing the advertisement delivery data with different advertisement effect stages pre-divided by advertisement effect indicators to determine the current advertisement effect stage of the target advertisement; wherein different advertisement effect stages are pre-matched with different revenue strategies; and then determining an estimated revenue value of delivering the target advertisement based on the target revenue strategy matched with the current advertisement effect stage; and determining the delivery cost of the target advertisement based on the estimated revenue value.

[0021] Compared with the prior art that treats the bidding process as a whole and adopts a unified revenue strategy from beginning to end, this technical solution takes into account the flow process of advertising effects in the bidding process, innovatively divides the bidding process into different advertising effect stages, and uses a segmented calculation method to determine the estimated revenue value of the target advertisement, and then uses the estimated revenue value to determine the target advertisement's delivery cost; specifically, different advertising effect stages are matched with different revenue strategies, and the advertising effect stages and revenue strategies match each other, which can ensure the accuracy of the estimated revenue value of each advertising effect stage. The estimated revenue value is a key influencing factor of the advertising delivery cost. The use of accurate estimated revenue values ​​can more reasonably control the delivery cost of the target advertisement. Based on this, a good closed loop is formed among advertising effect, revenue strategy, estimated revenue value and delivery cost. On this basis, the current advertising effect stage of the target advertisement is determined according to the current advertising delivery data. In the current advertising stage, the estimated revenue value of the target advertisement is determined according to the target revenue strategy that matches it, which can improve the accuracy of the estimated revenue value in the current advertising effect stage, and then determine the reasonable delivery cost of the target advertisement based on the accurate estimated revenue value; the above is equivalent to optimizing the revenue strategy of the current advertising effect stage each time, and then merging the estimated revenue value and delivery cost obtained by these local optimal revenue strategies into the overall situation, so that the entire bidding process can reasonably control the delivery cost of the target advertisement on the basis of obtaining the optimal estimated revenue value, so that the actual delivery cost of the advertisement is more matched with the actual conversion effect of the advertisement. This solution can more reasonably determine the delivery cost of the target advertisement, which is conducive to bringing greater advertising delivery value to advertisers, from strong reach to users. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

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

[0024] Figure 1 A flowchart of an advertisement delivery method provided by an embodiment of the present disclosure;

[0025] Figure 2 A schematic diagram of the advertising effect flow provided by an embodiment of the present disclosure;

[0026] Figure 3 A schematic diagram of different advertising effect stages provided by an embodiment of the present disclosure;

[0027] Figure 4 A schematic diagram of the relationship between the advertisement click rate and ECPM provided in an embodiment of the present disclosure;

[0028] Figure 5 A structural block diagram of an advertisement delivery device provided in an embodiment of the present disclosure;

[0029] Figure 6 A schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0030] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0031] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0032] For advertisers who place advertisements on advertising delivery platforms, especially through newly established accounts, in the early stages, it is necessary to consider the inaccurate calculation of ECPM due to inaccurate click-through rate and conversion rate estimates; ECPM will not only affect whether the advertisement can be displayed to users, but also affect the rationality of advertisers' control over advertising delivery costs. Therefore, it is necessary to optimize the advertising revenue strategy to improve the accuracy of ECPM and the rationality of advertising delivery costs, so that the actual delivery cost of the advertisement matches the actual conversion effect of the advertisement, and avoid cost losses to advertisers. Based on this, the embodiments of the present disclosure provide an advertising delivery method, device, equipment and medium, which can adopt different revenue strategies for different advertising effect stages in the bidding stage, and then use the estimated revenue value to improve the rationality of the delivery cost of the target advertisement.

[0033] To facilitate understanding of the solution, an example of the application scenario of the embodiment of the present disclosure is first used here. In the advertising delivery process, advertisers need to first create the advertisements they want to deliver on the delivery end. After the advertiser successfully creates an advertisement, it usually needs to go through the steps of recall, rough ranking, fine ranking, bidding and display in sequence; among them, the bidding step will affect whether an advertising order can be displayed to the user. Therefore, a better revenue strategy should be selected or designed in the bidding step to improve the bidding ability of the advertisement, and on this basis, the cost of advertising delivery should be reasonably controlled.

[0034] Figure 1This is a flowchart of an advertising delivery method provided in an embodiment of the present disclosure. The method can be applied to various scenarios in which a server equipped with an advertising delivery platform delivers advertisements to a client, for example, a scenario in which a video playback APP server delivers advertisements to a video playback APP client, a scenario in which a social Web server delivers advertisements to a social Web client, etc., without limitation herein. Also, the advertising delivery method provided in this embodiment can be executed by an advertising delivery device, which can be implemented in software and / or hardware; specifically, for example, an electronic device or a server. Among them, the electronic device can include devices with communication functions such as tablet computers, desktop computers, laptop computers, and smart phones. The server can be a device with storage and computing functions such as a cloud server or a server cluster.

[0035] Reference Figure 1 The advertising delivery method provided in this embodiment may include the following steps.

[0036] S102, in response to an advertisement delivery request of a target advertisement, obtaining current advertisement delivery data of the target advertisement, wherein the target advertisement is any advertisement among a plurality of advertisements to be bid by the account.

[0037] In some embodiments, the user can use the client by logging into an application, clicking on application content, playing a video, etc. During the use of the client, the client can send an advertisement delivery request for a target advertisement to the advertisement delivery platform of the server. The server can respond to the advertisement delivery request and obtain the information of the account that logged into the client and the current advertisement delivery data of the target advertisement. The account information may include the account code, login address, and first exposure time, etc.

[0038] In most scenarios, considering the problem of lack of historical accumulated data in the early stage of advertising delivery, the embodiment of the present disclosure can be used to determine the advertising delivery cost for any advertising delivery request for a target advertisement. In another possible scenario, for a newly registered account (new account), the problem of lack of historical accumulated data will be more significant. In this case, the present embodiment can determine whether the account is a new account based on the account information. If it is a new account, since the account attributes (such as age, gender, preference information, etc.) and advertising attributes (such as advertising materials, advertising entities, the industry to which the advertisement belongs, etc.) of the new account are relatively few or absent, inaccurate ECPM estimates will occur. Therefore, for advertising delivery requests of new accounts, the embodiment of the present disclosure is used to determine the advertising delivery cost.

[0039] After the target advertisement is delivered to the client, the target advertisement appears on the client interface browsed by the user, which can be called the exposure of the advertisement. In practical applications, the number of advertisement exposures is usually increased by large-scale advertisement delivery, thereby increasing the number of advertisement clicks and conversions. Based on this, the advertisement delivery data in this embodiment includes: the number of exposures, the number of clicks, and the number of conversions, etc. Specifically, it can be counted from the first time the advertisement delivery request is issued, according to the preset time window (such as 10 minutes as a time window), to obtain the advertisement delivery data currently accumulated for each time window of the target advertisement.

[0040] S104, comparing the advertisement delivery data with different advertisement effect stages pre-divided by advertisement effect indicators, and determining the current advertisement effect stage of the target advertisement; wherein different advertisement effect stages are pre-matched with different revenue strategies.

[0041] Reference Figure 2 , the flow of advertising effects in the bidding process is exposure, click and conversion respectively. Therefore, this embodiment can determine the advertising effect index based on exposure, click and conversion, and use the advertising effect index to divide different advertising effect stages, and then set different revenue strategies according to the congestion control algorithm, so as to quickly pass the cold start stage. The so-called congestion control algorithm is to set different revenue strategies according to different advertising effect stages, so as to obtain stable advertising delivery.

[0042] As a possible embodiment, the advertising effectiveness indicators may include: a click count threshold and / or an exposure count threshold for dividing the first advertising effectiveness stage and the second advertising effectiveness stage; a conversion count threshold for dividing the second advertising effectiveness stage and the third advertising effectiveness stage; wherein the first advertising effect stage, the second advertising effect stage and the third advertising effect stage are multiple stages of advertising effectiveness indicators from small to large.

[0043] For example, when the target advertisement is first launched, its exposure and click times are bound to be very low. At this time, the first advertising effect stage is entered with the goal of quickly seeking exposure and clicks. Therefore, the first advertising effect stage can also be called the rapid quantity-seeking stage.

[0044] As the target advertisement is delivered, its exposure times and click times gradually accumulate, and the target advertisement can enter the second advertisement effect stage from the first advertisement effect stage. In this embodiment, the first advertisement effect stage and the second advertisement effect stage are divided based on a preset click count threshold and / or exposure count threshold.

[0045] Specifically, if the total number of clicks is not greater than the click threshold, and the total number of exposures is not greater than the exposure threshold, then the first advertising stage is retained, and clicks and exposures continue to be accumulated. Until the accumulated total number of clicks is greater than the click threshold and / or the total number of exposures is greater than the exposure threshold, it indicates that the exposure and click volume sought in the first advertising effect stage have been reached, and then the second advertising effect stage with the goal of stabilizing the click rate is entered. Therefore, the second advertising effect stage can also be called the click stabilization stage.

[0046] After entering the second advertising effect stage, the most important conversion effect needs to be paid attention to, which is reflected by the user completing conversion behaviors such as purchase, registration, download, etc. In this embodiment, the second advertising effect stage and the third advertising effect stage are divided based on a preset conversion number threshold.

[0047] Specifically, if the total accumulated conversion times are greater than the conversion times threshold, it means that the conversion behavior has occurred enough times, and it is possible to enter the third advertising effect stage of normal delivery. Therefore, the third advertising effect stage can also be called the normal delivery stage.

[0048] This embodiment matches different revenue strategies for the above-mentioned different advertising effect stages, so as to determine the estimated revenue value and delivery cost for each advertising effect stage, thereby realizing the switch from a scheme using a unified revenue strategy to a scheme that divides the advertising effect stages and uses different revenue strategies to determine the estimated revenue value and delivery cost in each advertising effect stage. This can ensure the bid accuracy in each advertising effect stage, which is conducive to bringing greater advertising delivery value to advertisers and satisfying the advertising delivery effect that users hope to achieve. The different revenue strategies pre-matched at different advertising effect stages will be described in detail below.

[0049] Based on the above embodiments, this embodiment can compare the advertising data such as the number of exposures, clicks and conversions currently counted for the target advertisement with the different advertising effect stages divided by advertising effect indicators, so as to determine the current advertising effect stage of the target advertisement (i.e., the current advertising stage).

[0050] S106, determining an estimated revenue value of delivering the target advertisement according to the target revenue strategy matched to the current advertisement effect stage.

[0051] After determining the current advertising effect stage, this embodiment can determine the target revenue strategy that matches the current advertising effect stage according to the matching relationship, and then determine the estimated revenue value of the target advertisement based on the target revenue strategy and the advertisement delivery data.

[0052] S108: Determine the delivery cost of the target advertisement according to the estimated revenue value.

[0053] In this embodiment, the estimated revenue value can be used as a reference indicator for advertisers to adjust the delivery cost of the target advertisement. For example, if the estimated revenue value is low and does not reach the revenue value expected by the advertiser, it means that the bidding ability of the target advertisement is insufficient. In this case, the delivery cost of the target advertisement can be increased to increase the estimated revenue value. Alternatively, when the estimated revenue value reaches the revenue value expected by the advertiser, the current delivery cost can be kept unchanged. Alternatively, when the estimated revenue value exceeds the revenue value expected by the advertiser, the delivery cost of the target advertisement can be appropriately reduced to ensure that the estimated revenue value meets the revenue value expected by the advertiser. Alternatively, considering that some advertising effect stages require saving delivery costs, and some advertising effect stages require further increasing delivery costs to obtain higher revenue values, the needs of advertisers at different advertising effect stages are often different; in this case, the estimated revenue value and the current advertising effect stage can be considered together to determine whether to increase, reduce or keep the delivery cost of the target advertisement unchanged.

[0054] Of course, the above is only an example of determining the target advertising cost based on the estimated revenue value and should not be understood as a limitation. In actual applications, the target advertising cost can be determined more reasonably by comprehensively considering various factors such as the estimated revenue value and advertising demand.

[0055] In the advertising delivery method provided in the above embodiment, the advertising delivery request of the target advertisement is first responded to, and the current advertising delivery data of the target advertisement is obtained; then the advertising delivery data is compared with different advertising effect stages pre-divided by advertising effect indicators to determine the current advertising effect stage of the target advertisement; wherein different advertising effect stages are pre-matched with different revenue strategies; and then, based on the target revenue strategy matched with the current advertising effect stage, the estimated revenue value of delivering the target advertisement is determined; and the delivery cost of the target advertisement is determined based on the estimated revenue value.

[0056] Compared with the prior art that treats the bidding process as a whole and adopts a unified revenue strategy from beginning to end, this technical solution takes into account the flow process of advertising effects in the bidding process, innovatively divides the bidding process into different advertising effect stages, and uses a segmented calculation method to determine the estimated revenue value of the target advertisement, and then uses the estimated revenue value to determine the target advertisement's delivery cost; specifically, different advertising effect stages are matched with different revenue strategies, and the advertising effect stages and revenue strategies match each other, which can ensure the accuracy of the estimated revenue value of each advertising effect stage. The estimated revenue value is a key influencing factor of the advertising delivery cost. The use of accurate estimated revenue values ​​can more reasonably control the delivery cost of the target advertisement. Based on this, a good closed loop is formed among advertising effect, revenue strategy, estimated revenue value and delivery cost. On this basis, the current advertising effect stage of the target advertisement is determined according to the current advertising delivery data. In the current advertising stage, the estimated revenue value of the target advertisement is determined according to the target revenue strategy that matches it, which can improve the accuracy of the estimated revenue value in the current advertising effect stage, and then determine the reasonable delivery cost of the target advertisement based on the accurate estimated revenue value; the above is equivalent to optimizing the revenue strategy of the current advertising effect stage each time, and then merging the estimated revenue value and delivery cost obtained by these local optimal revenue strategies into the overall situation, so that the entire bidding process can reasonably control the delivery cost of the target advertisement on the basis of obtaining the optimal estimated revenue value, so that the actual delivery cost of the advertisement is more matched with the actual conversion effect of the advertisement. This solution can more reasonably determine the delivery cost of the target advertisement, which is conducive to bringing greater advertising delivery value to advertisers, from strong reach to users. In addition, for the scenario of delivering target advertisements to new accounts, the use of this solution can help improve the startup of new accounts and quickly pass the cold start stage.

[0057] Figure 3 A schematic diagram of an advertising bidding process corresponding to the advertising effect stage provided in an embodiment of the present disclosure; based on the above embodiment, reference is made to Figure 3 As shown, comparing the advertising delivery data with different advertising effect stages pre-divided by advertising effect indicators to determine the current advertising effect stage of the target advertisement; and determining the estimated revenue value of delivering the target advertisement according to the target revenue strategy matched with the current advertising effect stage, which may include:

[0058] Determine the total number of clicks, total number of exposures, and total number of conversions accumulated by the target advertisement based on the advertisement delivery data; compare the total number of clicks with a click threshold, compare the total number of exposures with an exposure threshold, and compare the total number of conversions with a conversion threshold.

[0059] According to the above comparison results, the current advertising effect stage of the target advertisement is determined, and according to the target profit strategy matched with the current advertising effect stage, the estimated profit value of the target advertisement is determined. The following describes the steps for each advertising effect stage.

[0060] In one embodiment, if the total number of clicks is not greater than the click threshold and the total number of exposures is not greater than the exposure threshold, it is determined that the current advertisement effect stage of the target advertisement is the first advertisement effect stage.

[0061] In this first advertising effect stage, according to the target revenue strategy matched in the first advertising effect stage, an embodiment of determining the estimated revenue value of delivering the target advertisement may include: determining the larger value of the preset advertising bid price and the traffic price as the reference price; determining the click-through rate according to the total number of clicks and the total number of exposures in the advertising delivery data; and determining the estimated revenue value of delivering the target advertisement based on the reference price and the click-through rate.

[0062] When a new account has just started to deliver target ads, there may be no ad delivery data or only a small amount of ad delivery data; in this case, the current advertising effect stage of the target ad will be determined to be the first advertising effect stage based on the total number of clicks not being greater than the click threshold and the total number of exposures not being greater than the exposure threshold.

[0063] In the first advertising effect stage, the click-through rate and conversion rate obtained by the conventional estimation model are not accurate, and the profit value calculated by this method has a large deviation, making it difficult to bid successfully. Therefore, in this embodiment, when determining that the current advertising effect stage of the target advertisement is the first advertising effect stage, the influence of the click-through rate and conversion rate on the estimated profit value can be weakened, and the total number of clicks and total number of exposures, which are two real and accurate data, are used as the basis, combined with the advertising bid price set by the advertiser and the traffic price set by the advertising delivery platform, so that the estimated profit value of the target advertisement in the first advertising effect stage can be determined more accurately.

[0064] For an advertisement delivery request, it has advertisement attributes and traffic data. Among them, the traffic attribute can be understood as the lowest selling price of the advertisement delivery platform for a whole traffic, which is called the traffic price in this embodiment, and can be specifically expressed as floor_price; the advertisement attribute can be understood as the price that the advertiser describes for an advertisement, which is called the advertisement bidding price in this embodiment, and can be specifically expressed as bidding_price.

[0065] Based on this, when the target revenue strategy of the first advertising effect matching stage is implemented, the larger value of the advertising bidding price bidding_price and the traffic price floor_price is determined as the reference price by first taking the maximum value function max(bidding_price, floor_price).

[0066] Secondly, divide the total number of clicks (sum(click)) in the advertising data by the total number of impressions (sum(impression)) to get the click-through rate, that is: (sum(click) / sum(impression).

[0067] Refer to the following formula (1) to determine the estimated revenue value of the target advertisement based on the reference price and click-through rate:

[0068] ecpm 1 =max(bidding_price,floor_price)*ln(1 / (sum(click) / sum(impression))(1)

[0069] Among them, ecpm 1 It represents the estimated revenue value of the first advertising effect stage, max(bidding_price,floor_price) represents the reference price, and (sum(click) / sum(impression) represents the click-through rate.

[0070] According to the above formula (1), the smaller the click rate, the larger the value of the above logarithmic function ln() will be. Correspondingly, the estimated revenue value ecpm 1 Therefore, this embodiment estimates the disadvantage of the target revenue strategy represented by the above formula in the first advertising effect stage, and determines the target advertisement delivery cost according to the estimated revenue value. The estimated revenue value can improve the bidding ability of the target advertisement, and the actual delivery cost of the target advertisement can be matched with the actual conversion effect of the advertisement by determining the delivery cost, thereby strongly reaching users and quickly accumulating advertisement delivery data in the first advertising effect stage, so as to provide more effective data for the next advertising effect stage, and then correctly estimate the delivery behavior.

[0071] According to the above embodiment, in the first advertising effect stage, determining the target advertising delivery cost according to the estimated revenue value may include: adjusting a preset advertising bid price according to the estimated revenue value, and determining the target advertising delivery cost according to the adjusted price.

[0072] In a specific example, if the estimated revenue value does not reach the revenue value expected by the advertiser, since the first advertising effect stage is the earliest stage of advertising delivery, it is hoped that the target advertisement can bring ideal revenue effects in the future stage; in this case, the preset advertising bid price can be increased, and the estimated revenue value of the target advertisement can be raised by increasing the advertising bid price, and then the new advertising bid price after the increase will be used as the delivery cost of the target advertisement.

[0073] In another specific example, if the estimated revenue value exceeds the revenue value expected by the advertiser, the preset advertisement bidding price may be kept unchanged and the advertisement bidding price may be determined as the delivery cost of the target advertisement.

[0074] Alternatively, the first advertising effect stage is a period of low traffic. In order to save costs and reserve sufficient cost budget for the future peak traffic stage, the advertising bid price can be reduced, and the new advertising bid price after the reduction will be used as the target advertising delivery cost. Of course, it is understandable that the new advertising bid price after the reduction still needs to meet the estimated revenue value to achieve the revenue value expected by the advertiser.

[0075] In this embodiment, if Figure 4 As shown in the above formula (1), as the click rate increases, the estimated revenue value ecpm 1 The above changes indicate that as the target advertisement delivery process develops, its advertisement delivery data and advertisement effect have changed. The target revenue strategy and delivery cost matched in the first advertisement effect stage are no longer suitable for the current advertisement effect stage of the target advertisement. The revenue strategy will gradually withdraw until it is determined to enter the second advertisement effect stage by comparing the advertisement delivery data with the different advertisement effect stages divided by the advertisement effect indicators in advance, referring to the contents described in the following embodiments.

[0076] In one embodiment, if the total number of clicks is greater than the click number threshold and / or the total number of exposures is greater than the exposure number threshold, and the total number of conversions is not greater than the conversion number threshold, the current advertising effect stage of the target advertisement is determined to be the second advertising effect stage.

[0077] In this second advertising effect stage, the number of exposures and clicks has accumulated to a sufficient number, and more feature data of the account and the target advertisement has been accumulated. At this time, the click-through rate estimation result of the click-through rate estimation model is credible. Therefore, in the second advertising effect stage, for a large number of advertisements under the account, you can select high-quality advertisements with good click-through rates to bid, and bid at the reserve price for other advertisements that do not meet the requirements. In other words, as an individual among a large number of advertisements, you can judge whether the target advertisement is a high-quality advertisement with a good click-through rate. If it is, bid for it, otherwise bid at a low price.

[0078] Based on this, according to the target revenue strategy matched in the second advertising effect stage, an embodiment of determining the estimated revenue value of delivering the target advertisement may include the following steps (1)-(5).

[0079] Step (1) obtains the account characteristics of the account and the advertisement characteristics of the target advertisement within the current time window; wherein the current time window includes: a time interval with the sending time of the advertisement delivery request as the end time and the duration is not greater than the preset duration, that is, a time interval before the sending time of the advertisement delivery request.

[0080] This embodiment uses a web crawler or other method to obtain account characteristics of accounts and advertisement characteristics of target advertisements within the current time window, such as gender, registration time, login address, preference information, and user groups; advertisement characteristics include advertisement materials, advertisement entities, and the industry to which the advertisement belongs.

[0081] Step (2) determines the first estimated click rate of the target advertisement according to the account characteristics and advertisement characteristics through a pre-trained click rate prediction model.

[0082] Specifically, the account features and the advertisement features can be input into a pre-trained click-through rate prediction model, and the first estimated click-through rate of the target advertisement output by the click-through rate prediction model can be obtained; wherein the first estimated click-through rate is used to indicate the probability that the user clicks on the target advertisement when the target advertisement is exposed. The above-mentioned click-through rate prediction model can include but is not limited to LR (Logistic Regression) model, FM (Factorization Machine), DeepFM (Deep Factorization Machine), etc. This embodiment does not specifically limit the above-mentioned click-through rate prediction model.

[0083] Step (3) determines whether the first estimated click rate meets the preset click rate requirement.

[0084] The implementation process of this step includes:

[0085] According to the number of clicks and exposures in multiple historical time windows, the average click rate (expressed as pctr_avg) and the standard deviation of the click rate (expressed as pctr_std) corresponding to the multiple historical time windows are determined. Determine the first estimated click rate (expressed as pctr 1 ) and the target ratio between the average click rate pctr_avg and the standard deviation of click rate pctr_std, that is, (pctr 1 -pctr_avg) / pctr_std.

[0086] Determine whether the target ratio is greater than a preset confidence threshold; if so, determine that the first estimated click-through rate meets the preset click-through rate requirement, in which case, execute the following step (4); if not, determine that the first estimated click-through rate does not meet the click-through rate requirement, in which case, execute the following step (5).

[0087] The target ratio in this embodiment can reflect the distribution of clicks. If the target ratio is greater than the preset confidence threshold, it not only indicates that the click rate is a relatively accurate value, but also indicates that the exposure traffic obtained by the target advertisement is high-quality traffic, and users are more willing to click on the target advertisement. For target advertisements with high-quality traffic, the following step (4) can be used to bid. On the contrary, if the target ratio is not greater than the preset confidence threshold, it means that the exposure traffic obtained by the target advertisement is low-quality traffic, which will correspondingly cause the actual conversion cost of the target advertisement to exceed the conversion cost preset by the advertiser, causing certain losses to the advertiser, so the following step (5) is used to bid at a low price.

[0088] If step (4) is satisfied, the larger value of the preset advertisement bid price and the traffic price is determined as the reference price, and the estimated revenue value of delivering the target advertisement is determined based on the reference price and the first estimated click-through rate.

[0089] In this embodiment, the target ratio corresponding to the first estimated click rate (pctr 1 -pctr_avg) / pctr_std is greater than the preset confidence threshold, it is determined that the first estimated click-through rate meets the preset click-through rate requirement, and the exposure traffic obtained by the target advertisement is of higher quality. For the target advertisement with high-quality traffic, refer to the following formula (2) to determine the estimated revenue value of the target advertisement based on the reference price and the target ratio corresponding to the first estimated click-through rate:

[0090] ecpm 2 =max(bidding_price,floor_price)*(pctr-pctr_avg) / pctr_std(2)

[0091] Among them, ecpm 2 Indicates the estimated revenue value of the second advertising effect stage, max(bidding_price,floor_price) indicates the reference price, (pctr 1 -pctr_avg) / pctr_std represents the target ratio corresponding to the first estimated click rate.

[0092] If step (5) is not satisfied, the traffic price is determined as the estimated revenue value of delivering the target advertisement.

[0093] In this embodiment, the target ratio (pctr-pctr_avg) / pctr_std corresponding to the first estimated click rate is not greater than the preset confidence threshold, and it is determined that the first estimated click rate does not meet the preset click rate requirement. For the target advertisement with poor exposure traffic, in order to avoid the actual conversion cost of the target advertisement exceeding the conversion cost preset by the advertiser, causing losses to the advertiser, this embodiment directly determines the traffic price as the estimated revenue value of the target advertisement, that is, ecpm' 2 =floor_price.

[0094] In the second advertising effect stage, this embodiment first determines whether the target advertisement has high-quality exposure traffic through the target ratio corresponding to the first estimated click-through rate. For target advertisements with high-quality exposure traffic, the estimated revenue value is determined based on the target revenue strategy represented by the above formula (2), and the target advertisement delivery cost is determined according to the estimated revenue value. The estimated revenue value can improve the bidding ability of high-quality target advertisements, and the delivery cost can be determined to match the actual delivery cost of the target advertisement with the actual conversion effect of the advertisement, thereby strongly reaching users and continuing to accumulate advertising delivery data in the second advertising effect stage, so as to provide more effective data for the next advertising effect stage, and thus correctly estimate delivery behavior. For target advertisements with poor exposure traffic, the traffic price is used to provide low-price discounts for the target advertisements, saving advertising delivery costs for advertisers and advertising delivery platforms.

[0095] According to the above embodiment, in the second advertising effect stage, determining the target advertising delivery cost according to the estimated revenue value may include: adjusting a preset advertising bid price according to the estimated revenue value, and determining the target advertising delivery cost according to the adjusted price.

[0096] Specifically, for the case where the target advertisement has high-quality exposure traffic, since the second advertising effect stage belongs to the peak traffic period, and the second advertising effect stage can predict that the target advertisement will bring higher returns in the future. Based on this, the advertising bid can be appropriately increased during the peak traffic period to obtain more exposure opportunities. In this case, this embodiment can obtain the delivery cost of the target advertisement determined in the first advertising effect stage, and use the delivery cost as the preset advertising bid price; then, the preset advertising bid price is increased, and the estimated revenue value of the target advertisement is further increased by increasing the advertising bid price, and then, the increased new advertising bid price is used as the delivery cost of the target advertisement in the second advertising effect stage.

[0097] In this embodiment, after entering the second advertising effect stage, the click-through rate tends to be stable, and the impact of increasing or decreasing the delivery cost on the click-through rate is no longer obvious; at the same time, the accumulated advertising delivery data increases further, so we can start to pay attention to the accumulated number of conversions; the number of conversions is an important indicator for evaluating advertising delivery.

[0098] In one embodiment, if the total number of conversions is greater than the conversion number threshold, it is determined that the current advertisement effect stage of the target advertisement is the third advertisement effect stage.

[0099] In this third advertising effect stage, the total number of conversions has accumulated to the specified conversion threshold, and it is believed that the estimation model can correctly estimate the bid for the target advertisement. At this time, it is necessary to consider the account's delivery cost and maximize the revenue of the advertising delivery platform.

[0100] Based on this, according to the target revenue strategy matched in the third advertisement effect stage, an embodiment of determining the estimated revenue value of delivering the target advertisement may include:

[0101] First, a second estimated click rate of the target advertisement is determined based on the pre-acquired account characteristics and advertisement characteristics through a pre-trained click rate prediction model; and an estimated conversion rate of the target advertisement is determined based on the account characteristics and advertisement characteristics through a pre-trained conversion rate prediction model.

[0102] Then, the estimated revenue value of the target advertisement is determined according to the preset advertisement bidding price, the second estimated click rate, the estimated conversion rate and the preset cost control factor; specifically, the following formula (3) can be referred to:

[0103] ecpm 3 =bidding_price*pctr 2 *pcvr*α(3)

[0104] Among them, ecpm 3 Indicates the estimated revenue value of the third advertising effect stage, bidding_price indicates the advertising bidding price, pctr 2 represents the second estimated click-through rate, pcvr represents the estimated conversion rate, and α represents the cost control factor.

[0105] In this embodiment, when determining the delivery cost of the target advertisement according to the estimated revenue value in the third advertisement effect stage, the advertisement effect in this stage is basically stable. Therefore, the delivery cost of the target advertisement determined in the second advertisement effect stage is obtained based on the estimated revenue value in the third advertisement effect stage having reached the expected revenue value, and the delivery cost is used as the preset advertisement bid price. Then, the preset advertisement bid price is kept unchanged, and the delivery cost of the target advertisement in the third advertisement effect stage is determined to be the preset advertisement bid price. Alternatively, in the case of ensuring that the estimated revenue value reaches the revenue value expected by the advertiser, the preset advertisement bid price is appropriately reduced, and the reduced advertisement bid price is determined as the delivery cost of the target advertisement in the third advertisement effect stage.

[0106] In summary, the advertising delivery method provided in the above embodiments takes into account the flow process of advertising effects in the bidding link, innovatively uses advertising effect indicators to divide different advertising effect stages, and uses the advertising effect stage to calculate the estimated revenue value, and then uses the estimated revenue value to determine the delivery cost of the target advertisement. Specifically, by comparing the advertising delivery data with different advertising effect stages, the current advertising effect stage of the target advertisement is determined; because different advertising effect stages match different revenue strategies, the estimated revenue value of the target advertisement is determined according to the target revenue strategy that matches it in the current advertising stage, which can ensure the accuracy of the estimated revenue value of the current advertising effect stage. The estimated revenue value is a key influencing factor of the advertising delivery cost. Using an accurate estimated revenue value can more reasonably control the delivery cost of the target advertisement. By analogy, as the target advertisement delivery process proceeds, the estimated revenue value and delivery cost will be accurately calculated according to the matching revenue strategy at each advertising effect stage in the bidding link, and then the estimated revenue value and delivery cost obtained by these local optimal revenue strategies at each advertising effect stage are merged into the global state, so as to achieve the optimization of the estimated revenue value and delivery cost of the entire bidding link.

[0107] In addition, in each advertising effect stage, more advertising data will be accumulated to provide a data basis for the next advertising effect stage and its revenue strategy. The previous advertising effect stage will have a positive impact on the next advertising effect stage, so that the delivery behavior can be correctly estimated in the next advertising effect stage, which will help to further improve the accuracy of the estimated revenue value and delivery cost.

[0108] Corresponding to the advertisement delivery method provided by the above disclosed embodiment, the disclosed embodiment also provides an advertisement delivery device. Figure 5 A schematic diagram of the structure of an advertisement delivery device provided by an embodiment of the present disclosure is shown in FIG. Figure 5 As shown, the advertisement delivery device may include the following modules:

[0109] The data acquisition module 210 is used to respond to an advertisement delivery request of a target advertisement and acquire current advertisement delivery data of the target advertisement;

[0110] The stage determination module 220 is used to compare the advertisement delivery data with different advertisement effect stages pre-divided by advertisement effect indicators to determine the current advertisement effect stage of the target advertisement; wherein different advertisement effect stages are pre-matched with different revenue strategies;

[0111] The revenue determination module 230 is used to determine the estimated revenue value of placing the target advertisement according to the target revenue strategy matched with the current advertisement effect stage;

[0112] The cost determination module 240 is used to determine the delivery cost of the target advertisement according to the estimated revenue value.

[0113] The implementation principle and technical effects of the device provided in this embodiment are the same as those of the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the device embodiment, reference may be made to the corresponding contents in the aforementioned method embodiment.

[0114] Figure 6 The structure diagram of an electronic device provided by the embodiment of the present disclosure is shown in FIG. Figure 6 As shown, the electronic device 300 includes one or more processors 301 and a memory 302 .

[0115] The processor 301 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 300 to perform desired functions.

[0116] The memory 302 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 301 may execute the program instructions to implement the advertising delivery method of the embodiment of the present disclosure described above and / or other desired functions. Various contents such as input signals, signal components, noise components, etc. may also be stored in the computer-readable storage medium.

[0117] In one example, the electronic device 300 may further include: an input device 303 and an output device 304 , and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0118] In addition, the input device 303 may also include, for example, a keyboard, a mouse, and the like.

[0119] The output device 304 can output various information to the outside, including the determined distance information, direction information, etc. The output device 304 can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and the like.

[0120] Of course, to simplify, Figure 6 Only some of the components related to the present disclosure in the electronic device 300 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application situations, the electronic device 300 may also include any other appropriate components.

[0121] Furthermore, this embodiment also provides a computer-readable storage medium, wherein the storage medium stores a computer program, and the computer program is used to execute the above-mentioned advertisement delivery method.

[0122] The computer program product of an advertisement delivery method, device, electronic device and medium provided in the embodiments of the present disclosure includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the previous method embodiments. The specific implementation can be found in the method embodiments and will not be repeated here.

[0123] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0124] The above description is only a specific embodiment of the present disclosure, so that those skilled in the art can understand or implement the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to the embodiments described herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An advertisement delivery method, characterized in that: include: In response to an advertisement delivery request of a target advertisement, obtaining current advertisement delivery data of the target advertisement; Compare the advertisement delivery data with different advertisement effect stages pre-divided by advertisement effect indicators to determine the current advertisement effect stage of the target advertisement; wherein different advertisement effect stages are pre-matched with different revenue strategies; Determine the estimated revenue value of delivering the target advertisement according to the target revenue strategy matched to the current advertisement effect stage; The delivery cost of the target advertisement is determined according to the estimated revenue value.

2. The method according to claim 1, characterized in that The advertising effectiveness indicators include: A click count threshold and / or exposure count threshold for dividing the first advertising effect stage and the second advertising effect stage; and a conversion count threshold for dividing the second advertising effect stage and the third advertising effect stage; The first advertising effect stage, the second advertising effect stage and the third advertising effect stage are multiple stages in which the advertising effect indicator increases from small to large.

3. The method according to claim 2, characterized in that The step of comparing the advertisement delivery data with different advertisement effect stages pre-divided by using advertisement effect indicators to determine the current advertisement effect stage of the target advertisement includes: Determine the total number of clicks, total number of exposures, and total number of conversions accumulated by the target advertisement based on the advertisement delivery data; comparing the total number of clicks with the click number threshold, comparing the total number of exposures with the exposure number threshold, and comparing the total number of conversions with the conversion number threshold; If the total number of clicks is not greater than the click threshold, and the total number of exposures is not greater than the exposure threshold, determining that the current advertisement effect stage of the target advertisement is the first advertisement effect stage; If the total number of clicks is greater than the click number threshold and / or the total number of exposures is greater than the exposure number threshold, and the total number of conversions is not greater than the conversion number threshold, determining that the current advertising effect stage of the target advertisement is the second advertising effect stage; If the total number of conversions is greater than the conversion number threshold, it is determined that the current advertising effect stage of the target advertisement is the third advertising effect stage.

4. The method according to claim 1, characterized in that: The current advertising effect stage is a first advertising effect stage; and determining the estimated revenue value of placing the target advertisement according to the target revenue strategy matched to the current advertising effect stage includes: The larger value of the preset advertisement bidding price and the traffic price is determined as the reference price; Determine the click-through rate based on the total number of clicks and the total number of exposures in the advertisement delivery data; An estimated revenue value of delivering the target advertisement is determined based on the reference price and the click rate.

5. The method according to claim 1, characterized in that The current advertising effect stage is a second advertising effect stage; and determining the estimated revenue value of placing the target advertisement according to the target revenue strategy matched to the current advertising effect stage includes: Acquire account characteristics of the account and advertisement characteristics of the target advertisement within a current time window; Determining a first estimated click rate of the target advertisement according to the account characteristics and the advertisement characteristics by using a pre-trained click rate estimation model; Determining whether the first estimated click rate meets a preset click rate requirement; If the conditions are met, the larger value of the preset advertisement bidding price and the traffic price is determined as the reference price, and the estimated revenue value of placing the target advertisement is determined based on the reference price and the first estimated click rate; If not, the traffic price is determined as the estimated revenue value of delivering the target advertisement.

6. The method according to claim 5, characterized in that The determining whether the first estimated click rate meets a preset click rate requirement includes: Determine, according to the number of clicks and exposures in the multiple historical time windows, an average click rate and a standard deviation of the click rate corresponding to the multiple historical time windows; Determine a target ratio between a difference between the first estimated click rate and the average click rate and a standard deviation of the click rate; Determining whether the target ratio is greater than a preset confidence threshold; If it is greater than, it is determined that the first estimated click rate meets the preset click rate requirement; If not, it is determined that the first estimated click rate does not meet the click rate requirement.

7. The method according to claim 1, characterized in that The current advertising effect stage is a third advertising effect stage; and determining the estimated revenue value of placing the target advertisement according to the target revenue strategy matched to the current advertising effect stage includes: Determining a second estimated click rate of the target advertisement by using a pre-trained click rate prediction model and according to pre-acquired account features and advertisement features; Determine the estimated conversion rate of the target advertisement according to the account characteristics and the advertisement characteristics by using a pre-trained conversion rate prediction model; The estimated revenue value of placing the target advertisement is determined according to the preset advertisement bidding price, the second estimated click rate, the estimated conversion rate and a preset cost control factor.

8. The method according to any one of claims 1 to 7, characterized in that: The step of determining the delivery cost of the target advertisement according to the estimated revenue value includes: The preset advertisement bidding price is adjusted according to the estimated revenue value, and the delivery cost of the target advertisement is determined according to the adjusted price.

9. An advertisement delivery device, characterized in that: include: A data acquisition module, configured to respond to an advertisement delivery request of a target advertisement and acquire current advertisement delivery data of the target advertisement; A stage determination module, used to compare the advertisement delivery data with different advertisement effect stages pre-divided by advertisement effect indicators, and determine the current advertisement effect stage of the target advertisement; wherein different advertisement effect stages are pre-matched with different revenue strategies; A revenue determination module, used to determine the estimated revenue value of delivering the target advertisement according to the target revenue strategy matched to the current advertisement effect stage; The cost determination module is used to determine the delivery cost of the target advertisement according to the estimated revenue value.

10. An electronic device, characterized in that: The electronic device comprises: processor; a memory for storing instructions executable by the processor; The processor is used to read the executable instructions from the memory and execute the instructions to implement the method described in any one of claims 1-8.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device implements the method according to any one of claims 1 to 8.

Images