Evaluation of the effectiveness of an advertising clip

Abstract

The invention relates to a method and to a client-server system for evaluating the effectiveness of an advertising clip that can be transmitted via a communication network. At least one item of reference information is allocated to the advertising clip. The following is proposed: - ascertaining (103) a first value, which realises a user interest prior to the transmission of the advertising clip (5), - ascertaining (104) a second value, which realises a user interest after the transmission of the advertising clip (5), - ascertaining (105) a change in user interest by comparing the first value and the second value; - determining (105) a value for the effectiveness of the advertising clip (5) depending on the change in user interest.

Description

[0001] Assessing the effectiveness of an advertisement

[0002] The invention relates to a method and a client-server system for evaluating the effectiveness of an advertising spot transmitted via a communication network, wherein at least one piece of reference information is assigned to the advertising spot.

[0003] A client-server system comprises clients and servers, which can be implemented in software or hardware and exchange information over one or more communications networks. A server, for example, is configured to provide specific information or resources, so-called services. These services are typically used by one or more clients by sending requests to the server over the communications network. The server evaluates the incoming request, checks whether the client is authorized to make the request if necessary, and transmits the relevant information or makes the requested resources available. The server can be configured to, in turn, send requests to other servers, for example to obtain information, which it then processes if necessary and makes available to the requesting clients.In this function, the server also acts as a client. Likewise, a client can provide information to other clients and thus acts as a server.

[0004] Commercials can be transmitted in various data formats via the internet and / or broadcast-capable communication networks and received by suitable receiving devices, such as computers, smartphones, smart TVs, and / or HbbTV-enabled devices. Commercials typically refer to at least one product, at least one service, and / or at least one company. The information to which a commercial refers is hereinafter referred to as the reference information associated with the commercial.

[0005] Commercials can be optimized based on a variety of criteria, such as broadcast time, target audience, content, or length. Optimization can also be achieved by placing the commercial at a specific position within a commercial break in a TV program. Optimization can also be achieved by selecting the TV program itself during which the commercial is shown or by selecting other information presented to a user during which a commercial is broadcast. Optimization of a commercial can therefore fundamentally affect its presentation (time, duration, context, etc.) and its content.

[0006] To decide whether and how a commercial should be optimized, various criteria can be used individually or in any combination. These criteria include, for example, the costs of production and presentation as well as the conversion rate, which indicates the relationship between the number of users reached by a commercial and the resulting user response. For example, if a commercial advertises a product in an online shop, the relationship between the number of users shown the advertisement on their device and the resulting sales can be determined and used to decide whether this advertisement should be optimized. The conversion rate is therefore one way to measure the effectiveness of a commercial.

[0007] However, commercials often cover multiple products or services, refer to an entire product line, or concern a company as such, which means that all products and services are at least indirectly included. A statement about the effectiveness of the commercial based on a conversion rate attributed to a single product cannot usually be made with sufficient accuracy. Furthermore, many products or services are sold through a variety of online or offline stores, making it almost impossible to attribute individual sales to the impact of a specific commercial.

[0008] Methods are known for recording the behavior of users after an advertisement has been displayed ("tracking"), but this requires monitoring the user's end devices, which is often not permitted, not desirable or not possible for data protection reasons, for example because tracking data does not exist, access to possibly existing tracking data is too costly and / or the user has taken precautions to prevent tracking.

[0009] The object of the invention is to create a method and a client-server system that enable the effectiveness of an advertisement to be determined regardless of whether tracking data is available.

[0010] The problem is solved by a method of the type mentioned above in that the method comprises the following steps:

[0011] - determining at least a first value that reflects a user interest before the broadcast of the commercial, the first value being related to a relative or absolute frequency with which search queries relating to the reference information were requested before the broadcast of the commercial;

[0012] - determining at least a second value that reflects a user interest after the broadcast of the commercial, the second value being related to a relative or absolute frequency with which search queries relating to the reference information are requested after the broadcast of the commercial;

[0013] - Determining a change in user interest by comparing at least the first value and at least the second value;

[0014] - Determining a value for the effectiveness of the commercial depending on the change in user interest. According to the invention, user interest is determined by evaluating search queries, particularly comparing the change in the frequency of these search queries before and after the broadcast of the commercial. The search queries considered here relate to one or more pieces of reference information associated with the commercial, for example, an advertised product, an advertised service, a company, and / or a brand under which a company offers goods or services.

[0015] The communications network over which the commercial is transmitted is preferably a communications network suitable for transmitting TV programs. For this purpose, familiar technologies such as cable, satellite, and radio transmissions can be used. The commercial is therefore preferably a TV commercial. Searches for the relevant information before or after the broadcast of the commercial are preferably conducted via another communications network, particularly preferably the Internet.

[0016] Preferably, the data relating to the search queries to be evaluated are provided by an Internet search engine, an Internet service provider, or by the Internet browsers used by their users to communicate with the search engine, by means of additional software (a plug-in). According to this possible embodiment, the at least one first value that represents a user interest before the broadcast of the commercial and the at least one second value that represents a user interest after the broadcast of the commercial are thus

[0017] - directly from the search queries to at least one search engine,

[0018] - from the data of at least one Internet service provider by evaluating the search queries entered by its users to one or more search engines in the corresponding periods (ie before or after the broadcast of the commercial) and / or

[0019] - generated by collecting and evaluating search queries submitted by users via their internet browsers during the relevant time periods. Both search engine operators and internet service providers can make such data available for retrieval via a portal, whereby this data is preferably anonymized.

[0020] If several reference information items are assigned to an advertising spot, then according to a possible embodiment of the invention, a respective partial effectiveness of the advertising spot is determined for at least two reference information items, and an overall effectiveness is inferred by selecting the highest of the at least two partial effectivenesses, forming an average from the partial effectivenesses and / or weighting each of the partial effectivenesses with regard to a relevance of its underlying reference information for the advertising spot and determining an average from the weighted partial effectivenesses.

[0021] If a commercial is associated with multiple pieces of reference information, such as the name of a product, the name of a product line, and the name of the company that manufactures and / or distributes the product, the accuracy of the determined effectiveness of the commercial can be increased by considering search queries for several of these pieces of reference information and determining a partial effectiveness for each of the considered pieces of reference information according to the method according to the invention. An overall effectiveness is then calculated from the partial effectiveness values ​​thus determined. It can be advantageous to weight the partial effectiveness values ​​beforehand.For example, the partial effectiveness relating to a product can be weighted with a factor of 1, the partial effectiveness determined from search queries relating to a product line with a factor of 0.5 and the partial effectiveness determined for the search term relating to a company's name with a factor of 0.3 in order to calculate the overall effectiveness for the commercial.

[0022] According to a particularly advantageous embodiment, a temporal progression of search queries relating to at least one piece of reference information is recorded, and the effectiveness of the commercial is determined depending on a change in the temporal progression before and after the broadcast of the commercial. Temporal search histories can be continuously determined by search engines and / or generated by the search engine from the available information, depending on a specific requirement regarding the temporal progression of changes in search queries relating to a specific piece of reference information.

[0023] Time series can be requested, for example, from the operators of the search engine "Google" using the service known as "Google Trends." Because Google Trends accesses the search queries of the world's most popular search engine, it can be assumed that a large portion of the commercial's recipients will direct a possible search query for one or more related pieces of information to this very search engine after viewing the commercial. This data is therefore particularly well-suited for determining the effectiveness of the commercial, as it is representative of a large portion of the commercial's recipients and a large portion of the search queries.

[0024] It is possible that the temporal progression of user interest and / or the at least first value representing user interest before the commercial is broadcast and the at least second value representing user interest after the commercial is broadcast are provided in the form of relative values, so that the absolute effectiveness of the commercial in question cannot be directly determined from them. An example of this is the temporal progression of search terms that can be requested from Google Trends.

[0025] However, it may be desirable to obtain an increase in profit generated by an advertisement in absolute terms, for example to see whether the absolute costs of an advertisement are economically justifiable in relation to the increase in profit generated.

[0026] According to another possible embodiment of the invention, the absolute effectiveness of the commercial can be determined by comparing the relative effectiveness of the commercial and / or the relative values ​​that realize a user interest with the corresponding values ​​of another commercial, wherein absolute values ​​for the effectiveness and / or for the values ​​that realize a user interest are known for the other commercial. Preferably, another commercial is selected that advertised a comparable product or service, and from whose profit figures the absolute effectiveness to be expected with the currently viewed commercial can be deduced. Comparable commercials for which absolute effectiveness is present can originate from the same company or from different companies.

[0027] In particular, this allows inferences to be drawn from relative first and second values ​​to corresponding absolute values ​​that represent user interest before and after the broadcast of the commercial. The first and second values ​​preferably correspond to search queries regarding the at least one piece of reference information to a search engine during the corresponding period.

[0028] Preferably, if relative values ​​are available that represent user interest before and after the broadcast of the commercial, an absolute effectiveness is determined using artificial intelligence, particularly applying techniques from the field of machine learning. The relative values ​​of the currently viewed commercial are fed into an AI system, as are data from at least one commercial that was transmitted or broadcast in the past and in which a different company and / or at least one different product was advertised. Such "past performance data" are particularly well suited to generating sufficiently accurate absolute values ​​for evaluating the effectiveness of a commercial using a trained AI system.

[0029] According to a preferred embodiment of the invention, demographic data is taken into account when determining the effectiveness of the commercial. Demographic data can, for example, describe a specific target group defined by age, occupation, gender, and / or the region in which the commercial can be received. After the commercial has been transmitted, information can then be retrieved from a suitable database or online service that provides information about what proportion of users who received this commercial belong to this target group. Data such as that provided by AGF Videoforschung GmbH could be used here, for example.In a further step, the at least one first value that realizes a user interest before the broadcast of the commercial and was determined by a search engine by reducing the search queries that were requested by the search engine with regard to the at least one piece of reference information of the commercial before the broadcast of the commercial to search queries from those users that correspond to the demographic data of those users who received the commercial.Likewise, the at least one second value that a user interest realizes after the broadcast of the commercial and was determined by a search engine by reducing the search queries requested by the search engine regarding the at least one piece of reference information of the commercial after the broadcast of the commercial to search queries from those users who match the demographic data of those users who received the commercial. This allows the effectiveness of the commercial in increasing user interest to be determined and evaluated even more precisely.

[0030] Commercials typically include video data that provides visually representable information and / or audio data that provides acoustically perceptible information, such as speech and music. According to a further development of the method according to the invention, at least one piece of text-based content information relating to at least one piece of content in the video data is generated from the video data, and / or at least one piece of content information relating to at least one piece of content in the audio data is generated from the audio data.

[0031] Content information extracted from video data can be an automatically generated description of the scenes depicted in the video. The automatic generation of scene descriptions from video data is referred to as "action recognition." AI-driven software products are available for this purpose. Furthermore, AI-driven online services can be used to create and / or train specialized action recognition systems, such as Vertex AI, Video Intelligence API, or AWS or Amazon Rekognition.

[0032] Content information obtained from audio data can, for example, be a textual representation of the audio data generated by automatic transcription. Specialized online services and / or commercially available software packages can be used to automatically generate a transcription.

[0033] According to this embodiment, keywords are automatically extracted from the generated content information. Such keywords are also referred to as "spot content information." These can be, for example, verbs and nouns, which can be filtered according to their particular significance with regard to the product or general trends. One or more locations mentioned in the commercial or in which a scene takes place can also be a keyword, as can technical objects, the color of a particularly relevant object, the manner of an actor's appearance (proud, aggressive, reserved, etc.), or a basic style that accompanies the commercial (glamour, relaxed, etc.). A multitude of different categories and individual words are conceivable here, which are preferably generated from the content information using a specially trained, AI-controlled system.

[0034] After the effectiveness of the commercial has been determined or evaluated, the keywords generated in this way (spot content information) are in turn evaluated in this embodiment depending on the determined effectiveness of the entire commercial. Preferably, additional knowledge is applied here, which has been gathered, for example, from the evaluation of the effectiveness of other commercials. This knowledge can either be taken from a database provided for this purpose, or an AI-controlled system is used which has been trained using the content information and, in particular, the keywords generated from this from other commercials and the associated evaluations of their respective effectiveness. For example, it could be found that commercials set in a private kitchen achieve a higher average effectiveness than commercials set in a living room.Or it might turn out that commercials in which actors drive a purple vehicle are less effective than commercials in which red or black vehicles are used.

[0035] The object underlying the invention is also achieved by a client-server system of the type mentioned at the outset in that the client-server system is set up to carry out the method according to the invention.

[0036] Further features and advantages of the present invention are explained in more detail below with reference to the figures. It is emphasized that individual features shown in the figures may also be essential to the invention on their own, even if this is not shown in the figures and not explicitly mentioned in the description. Furthermore, the individual features shown in the figures may be combined with one another in any desired manner, even if this combination is not shown in the figures and not explicitly mentioned in the description. They show:

[0037] Figure 1 shows a client-server system according to the invention according to a first embodiment;

[0038] Figure 2 shows a flow diagram of a method according to a possible embodiment; and

[0039] Figure 3 is a flowchart of a method according to another possible embodiment.

[0040] Figure 1 shows a client-server system 1 which is connected, for example, by means of an Ethernet cable or a WiFi connection to a first communications network 2, which is embodied, for example, as the Internet. Figure 1 also schematically shows a company 3 and a broadcasting company 4, wherein the company 3 has deposited an advertisement 5 for a product or service with the broadcasting company 4, so that the broadcasting company 4 broadcasts the advertisement 5 at a fixed time and / or within a predetermined TV broadcast and / or with a predetermined regional coverage via a second communications network 6 which is suitable for transmitting TV broadcasts. The advertisement 5 is preferably a TV commercial. Known technologies such as cable, satellite, and radio transmissions can be provided for broadcasting the advertisement 5.In particular, the second communication network 6 can also be connected to the first communication network 2, and it is further possible that the first communication network 2 and the second communication network 6 are realized as a connected network, in particular as parts of the Internet.

[0041] According to the embodiment shown in Figure 1, a plurality of clients 7 configured to receive TV broadcasts and / or radio broadcasts can be connected to the second communications network 6. The clients 7 are assigned to users 8 and can be configured, for example, as set-top boxes, TV sets, smart TVs, HbbTVs, PCs, laptops, tablets, smartphones, and other known or future-developed terminal devices suitable for receiving TV broadcasts via one or more networks 2, 6.

[0042] The users 8 are assigned additional end devices 9, which may be configured, for example, as PCs, laptops, tablets, or smartphones. However, the end devices 9 can also be configured as set-top boxes, smart TVs, or HbbTVs. The end devices 9 are connectable to the internet 2, so that, in particular, a search query can be directed to a search engine 30.

[0043] A copy of the commercial 5, which was broadcast or transmitted, for example, via the communications network 6 and received by the client-server system 1, is stored in the client-server system 1. Several functional groups are implemented in the client-server system 1, which can be implemented in software, hardware, or a combination thereof. A functional unit 10 is configured to extract textual scene descriptions from the video data of the commercial 5. The functional unit 10 can also be implemented by a service from a provider specializing in this field, whose services are requested via the Internet 2. AI-controlled software products are available for this purpose. In addition, AI-controlled online services can be used to create and / or train specialized action recognition systems, such as "Vertex AI," "Video Intelligence API," or "AWS or Amazon Rekognition."The extracted content information can then be stored in a database 12.

[0044] In a functional unit 14, the audio data is extracted, and the resulting content information in the form of a textual transcription is stored in a database 16. For the automated generation of a transcription from the audio data, specialized online services and / or commercially available software packages can be used, possibly even AI-controlled services or software products. Of course, databases 12 and 16 can also be implemented as a single functional unit.

[0045] In a functional unit 18, a significant set of terms, referred to as 'spot content information', is extracted from the textual descriptions extracted from the video and audio data in functional units 10 and 14. Functional unit 18 uses suitable heuristics and / or one or more AI systems trained on data extracted from other commercials and assigned a sufficiently high relevance.

[0046] The client-server system 1 includes a functional unit 22 that includes a database in which, among other things, absolute conversion rates from previous commercials are stored. Furthermore, demographic data that includes information on previous commercials, and in particular information on the commercial 5, can be stored in the functional unit 22. This demographic information can be retrieved, for example, via the Internet 2 from a service 32 set up for this purpose. Such a service 32 is provided, for example, by 'AGF Videoforschung GmbH'. In a functional unit 24, the effectiveness of the commercial 5 can be determined depending on the content information generated in the functional unit 18 and selected information provided by the functional unit 22.For this purpose, in particular, the changes in user interests before and after the broadcast of the commercial 5 are used. The client-server system 1 can retrieve suitable information from the search engine 30 for this purpose. This information can be retrieved as absolute, but in particular also as relative values. In order to form absolute values ​​with sufficient accuracy from the relative values, a KL system can be implemented in the functional unit 24, which generates a meaningful absolute evaluation of the effectiveness from information on previous commercials, for example regarding conversion rates achieved there or absolute profit increases from the relative values ​​available for the current commercial 5 and, if necessary, taking demographic data into account.

[0047] In particular, the functional unit 24 identifies at least one commercial that has a particularly high level of effectiveness, ie, has been particularly well-received by users 8. The textual scene descriptions of the video data associated with this particularly effective at least one commercial, as well as the corresponding textual transcription of the audio data, are stored in the databases 12 and 16 and are available for retrieval there.

[0048] Finally, a functional unit 26 can contain a Kl system that is designed and trained to generate a synthetic TV commercial from a plurality of keywords. The keywords used are preferably textual scene descriptions generated from the video data from the database 12 and / or textual transcriptions of the audio data generated from the audio data from the database 16. Particularly preferably, the keywords for the at least one particularly effective commercial identified in the functional unit 24 are used. As a result, a particularly effective synthetic commercial can be generated fully automatically in the functional unit 26 using Kl within a very short time, which is particularly well received by users 8, possibly taking into account certain demographic data (e.g., by users 8 of a specific target group).

[0049] Figure 2 shows a flowchart showing individual steps of a possible embodiment of the method according to the invention (so-called “processing pipeline”). The method begins in step 100. In step 101, an advertisement 5 is received via a communications network 6 from a client 7 of a user 8. In step 102, a time profile of the user interest is requested from a service provider, for example an Internet search engine 30. The search term used here is the at least one piece of reference information associated with the received advertisement 5. The reference information can be, for example, the name of a product, a product line, a service, a product category, a brand and / or the name of a company. A time profile can be requested, for example, from the “Google Trends” service, which accesses data from the currently most frequently used search engine, “Google.”

[0050] The temporal progression of user interest includes, in particular, the user interest before and after receipt of the commercial 5 by the client 7 of the user 8. The user interest is represented, in particular, by an absolute or relative number of search queries related to the first reference information from the terminal device 9 of the user 8 via the first communication network 2. It is assumed that the user 8 activates a terminal device 9 (e.g., smartphone or laptop computer) when they find the commercial 5 interesting in order to search for the first reference information as a search term using a search engine on the Internet 2.

[0051] In a step 103, at least one first value for the user interest before the transmission of the commercial 5 is determined from the user interest history requested and received in step 102, and in a step 104, at least one second value for the user interest after the transmission of the commercial 5 is determined from the user interest history. In a step 105, a value is formed from the at least one first value representing the user interest before the transmission of the commercial 5 and the at least one second value representing the user interest after the transmission of the commercial 5, which value can have significance for the effect of the broadcast or transmitted commercial 5.If the values ​​are simply one value representing user interest before the commercial is transmitted to users 8, and a second value representing user interest after the commercial is transmitted, the difference between these values ​​can be calculated. If multiple values ​​determined at different times before the broadcast of the commercial 5 and multiple values ​​representing current user interest at different times after the broadcast of the commercial 5 are used, a value can be calculated from these values ​​using known statistical methods, representing a change in user interest caused by the broadcast of the commercial 5.

[0052] In a step 106, the effectiveness of the commercial 5 is inferred from this, or the effectiveness of the commercial 5 is evaluated. If the change in user interest is only available through relative values, absolute values ​​can be formed with sufficient accuracy by forming a meaningful absolute evaluation of the effectiveness of the commercial 5 in step 106 from information on previous commercials, for example regarding conversion rates or absolute profit increases achieved there, by comparing the relative values ​​available for the current commercial 5 and, if appropriate, taking demographic data into account.

[0053] The process (“processing pipeline”) comprising steps 101 to 106 can be run multiple times, each for different commercials 5, in order to identify at least one commercial 5 that has a particularly high level of effectiveness. The textual scene descriptions of the video data associated with this particularly effective at least one commercial 5, as well as the corresponding textual transcription of the audio data, are stored in databases 12 and 16 and are available for retrieval there. After identifying the particularly effective commercial(s) 5, in a further step, the textual scene descriptions of the video data, as well as the corresponding textual transcription of the audio data for the particularly effective commercial(s) 5, can be fed as keywords to a CL system that is designed and trained to generate a synthetic TV commercial from several keywords.This means that a synthetic advertising campaign can be generated fully automatically within a very short time using Kl, which is particularly well received by users.

[0054] Figure 3 shows another possible embodiment of the method according to the invention. This method begins in step 200. In step 201, the commercial 5 is transmitted via the communications network 6 to the clients 7 of a specific target group of users 8. For the commercial 5, for example, it can be specified within which TV program it is displayed to users 8. It can also be specified at which point within a commercial break the commercial 5 is shown.

[0055] In the embodiment illustrated in Figure 3, it is assumed that three pieces of reference information are available for the commercial 5. These can be, for example, the name of the company 3, the name of a product advertised by the commercial 5, and a brand associated with the product. The reference information is preferably provided by the company 3.

[0056] In a step 202-1, depending on a first piece of reference information, a request is sent to the search engine 30 or a service provider associated with the search engine 30 to obtain information about the frequency of search queries directed to the search engine 30 that are associated with the first piece of reference information. This can, for example, be a minute-by-minute chronological history of the search queries to the search engine that concern the first piece of reference information. If these results are provided, for example, by the service known as 'Google Trends', they can be relative values ​​designed such that the highest value within the queried period is assigned the value 100, and the other values ​​are then given in relation to this highest value.

[0057] In a step 203-1, the information describing the frequency of search queries depending on the first reference information is further processed to obtain a statement about the effectiveness of the commercial 5. For this purpose, demographic data can be used to determine how many of the search queries most likely originate from the users 8 reached by the commercial 5. Furthermore, at least sufficiently accurate statements can be generated about how many of the users 8 reached by the commercial 5 belong to a target group reached by the commercial 5. Statistical methods or specially trained AI systems can be used for this purpose.

[0058] Steps 202-2 and 203-2 are performed analogously to steps 202-1 and 203-1, but here the effectiveness of commercial 5 is determined with respect to a change in those search queries that relate to the second piece of reference information. Steps 202-3 and 203-3 are performed analogously for the values ​​that relate to the third piece of reference information.

[0059] In a step 204, the values ​​determined in steps 203-1, 203-2, and 203-3 are combined to obtain an overall statement about the effectiveness of the commercial 5. For example, an average can be calculated here, whereby the individual values ​​can additionally be weighted according to their likely significance for assessing the effectiveness of the commercial 5.

[0060] In the embodiment shown in Figure 3, it is assumed that the information received from the service provider regarding the frequency of search queries concerning the reference information contains relative data. In a step 205, absolute values ​​are determined from the relative data so that the effect of the commercial can be absolutely assessed. In particular, information from commercials that were broadcast at an earlier time and whose effectiveness is already known can be used for this purpose. Such information is stored, for example, in the functional unit 22 in Figure 2. For example, these commercials relate to products and / or companies that are sufficiently similar to those of commercial 5. Preferably, a KI system is also used for this purpose, which has been trained with the information from a database in which the relevant data on previous advertising campaigns is stored.

[0061] In step 206, a decision is made as to whether the process should be terminated here or whether suggestions for optimizing commercial 5 should be generated. If no optimization suggestions are provided, the process ends in step 207.

[0062] Otherwise, in a step 208, content information in the form of textual scene descriptions is generated from the video data of the commercial 5. AI systems specifically trained for this task can be used for this purpose, or a corresponding service from an external provider can be accessed via the Internet 2. In a step 209, the extracted content information is stored in a database 12.

[0063] In a step 210, content information is extracted from the audio data and in a step 211 the resulting textual transcription is stored in a database 16.

[0064] In a step 212, a significant set of keywords ('spot content information') is extracted from the textual descriptions extracted from the video data and the audio data in functional units 10 and 14. For this purpose, suitable heuristics and / or one or more AI systems are used, which have been trained using data extracted from other commercials and assigned a sufficiently high relevance.

[0065] In a step 213, the extracted keywords are compared with keywords extracted from other commercials, with preference given to commercials that have achieved greater effectiveness than the commercial 5 currently being evaluated. Possible results are then displayed or output in a step 214. If, for example, one of the relevant scenes in commercial 5 takes place in a garage, whereas scenes from comparable commercials take place in the driveway of a villa, this is detected, and in step 214, it is suggested to replace the scene in the garage with a scene that takes place in the driveway of a villa. The method ends in a step 215.

[0066] Of course, the suggested optimizations can affect not only one location, but many other areas, such as the time of day, the gender of important actors, the color of vehicles used, the interior design of rooms, the expression of the actors, the style and / or color of clothing, etc.

[0067] To categorize the relevant terms, identify similar commercials, and suggest potentially optimizable features of the commercial 5, a client-server system that has been specifically trained for this purpose using historical data from previously analyzed commercials can again be advantageously used. Thus, with the method according to the invention and the client-server system 1 according to the invention, it is not only possible to evaluate the effectiveness of the commercial 5, but also to propose suggestions for optimizing its effectiveness.

Claims

AMENDED CLAIMS received by the International Bureau on 18 November 2024 (18.11.2024) 1. A method for evaluating the effectiveness of an advertising spot (5) transmittable via a communications network (6), wherein the advertising spot (5) is assigned multiple reference information items, characterized in that for at least two pieces of reference information, a respective partial effectiveness of the advertising spot (5) dependent thereon is determined (202-1, 202-2, 202-3), and an overall effectiveness is inferred (203-1, 203-2, 203-3) by selecting the highest of the at least two partial effectivenesses; forming an average of the partial effectivenesses; and / or weighting each of the partial effectivenesses with respect to the respective relevance of the underlying reference information for the advertising spot (5), and determining an average from the weighted values; and in that determining the partial effectivenesses for each of the at least two pieces of reference information comprises the following steps: Determining (103) at least one first value that realizes a user interest before the transmission of the commercial (5), wherein the first value is related to a relative or absolute frequency with which search queries regarding the respective reference information were requested to a search engine (30) before the transmission of the commercial (5); Determining (104) at least one second value that realizes a user interest after the transmission of the commercial (5), wherein the second value is related to a relative or absolute frequency with which search queries regarding the respective reference information are requested to a search engine (30) after the transmission of the commercial (5); Determining (105) a change in user interest by comparing the at least first value and the at least second value; AMENDED SHEET (ARTICLE 19) Determining (105) a value for the partial effectiveness of the commercial (5) depending on the determined change in user interest.

2. Method according to claim 1, characterized in that the at least one first value which realizes a user interest before the transmission of the commercial (5) and the at least one second value which realizes a user interest after the transmission of the commercial (5), - from a service connected to the search engine (30), - using the data of an Internet service provider by evaluating the search queries entered by its users and / or - is generated by collecting and evaluating search queries sent by users (8) via their Internet browsers.

3. Method according to one of the preceding claims, characterized in that a temporal progression of the search queries with regard to the respective reference information is recorded and, depending on a change in the temporal progression before and after the transmission of the commercial (5), its partial effectiveness is determined.

4. Method according to one of the preceding claims, wherein - the temporal course of user interest and / or - the at least first value, which represents a user interest before the transmission of the commercial (5), and the at least second value, which represents a user interest after the transmission of the commercial (5), are relative values, characterized in that an absolute effectiveness of the commercial (5) is determined (24; 106; 205) by comparing the relative effectiveness of the commercial 4 and / or the relative values ​​realizing a user interest with the corresponding values ​​of another commercial, wherein absolute values ​​for the effectiveness and / or values ​​realizing a user interest are known for the other commercial. AMENDED SHEET (ARTICLE 19) 5. Method according to claim 4, characterized in that the absolute effectiveness is determined using artificial intelligence, in particular techniques from the field of 'machine learning' are applied.

6. Method according to one of the preceding claims, characterized in that demographic data are taken into account when determining the effectiveness of the commercial (5) (22; 106; 203-1, 203-2, 203-3).

7. Method according to claim 6, characterized in that in particular a regional broadcast and / or a target group are taken into account when determining the effectiveness.

8. Method according to one of the preceding claims, wherein the commercial (5) comprises video data and / or audio data, characterized in that the method comprises the following steps: - automatically generating (208) at least one text-based content information item relating to at least one content of the video data and / or - automatically generating (210) at least one text-based content information item relating to at least one content of the audio data, and the following steps: - automatically extracting (212) keywords from the content information; - Evaluating (213) the content information depending on the determined change in user interest and / or the determined effectiveness of the commercial (5).

9. Method according to one of the preceding claims, characterized in that an optimization of the effectiveness of the advertising spot (5) is proposed (214), wherein the optimization - the selection of a different time for the transmission of the commercial (5) via the communication network (6), and / or AMENDED SHEET (ARTICLE 19) - concerns the modification of one or more pieces of content, whereby content whose content information does not contribute or only contributes little to the effectiveness of the commercial (5) is removed and / or content whose content information contributes to greater effectiveness in comparable commercials is added.

10. Method for determining a particularly effective commercial (5) that can be transmitted via a communication network (6), characterized in that the method according to one of the preceding claims is run through several times in order to determine at least one commercial (5) that has a particularly high effectiveness, and to determine text-based content information associated with the at least one commercial (5) with regard to at least one content of the video data and / or at least one content of the audio data, wherein the text-based content information is preferably stored in at least one database (12; 16).

11. The method according to claim 10, characterized in that the text-based content information relating to at least one content of the video data and / or at least one content of the audio data is supplied as keywords to a Kl system which is designed and trained to generate a synthetic TV commercial from a plurality of keywords.

12. Method according to claim 11, characterized in that the text-based content information of the at least one commercial (5) for which a particularly high effectiveness has been determined is supplied to the Kl system.

13. Client-server system (1), wherein the client-server system (1) is connectable to a communication network (2, 6), characterized in that a method according to one of the preceding claims can be executed on the client-server system (1). AMENDED SHEET (ARTICLE 19)

Images