Title: System and Method for Calculating Powers of Interactions

Field of the Invention

[0001] The invention relates to a method and system for calculating the powers of interactions, particularly of interactions between a consumer and a brand. The powers of interaction can be used to identify those consumers that are deeply interacting with the brand.

Background to the Invention

[0002] The method and system of the invention enables the collection and calculation of the power of online interactions and offline interactions (generally termed "interactions") between consumers and brands. The power of the interactions includes, but is not limited to, the intensity, the motivation, as well as the timing of the interactions. The collected interactions are processed and analysed to determine which of the consumers might be categorised as being "genuine brand ambassadors". The term "genuine brand ambassadors" is used for those consumers belonging to a smaller subset of a brand's consumers that are most vocal about the brand and ultimately more valuable to the brand because word-of-mouth recommendations are a primary factor for twenty to fifty percent of all purchasing decisions. Brand ambassadors can also be called "brand advocates" or "brand recommenders".

[0003] There is no single solution to define which ones of the online interactions or the offline interactions between a consumer and brand indicate whether the consumer can be categorized as a genuine brand ambassador for the brand. Some brands thrive on social media, such as Facebook or Twitter, while other brands are more focused on a web store or even a physical store. There is a need for a system that allows the brands to define which types of the online interaction or the offline interaction should be analyzed and the degree as well as a motivation of interaction between the consumer and the brand. [0004] The degree or power of the interaction is important because not every interaction is equally important to the brand. The brand might find that the consumer submitting a video review of one of the brand's products is infinitely more valuable than tweeting about the brand, thus the brand would assign the video review a greater power of interaction than the tweet. Furthermore the motivation for the interaction can be equally important, the consumer could be writing a blog post informing other ones of the consumers about the brand. The motivation for this interaction could be to educate the other consumers and indicate that this interaction does not only signify the interaction between the brand and the consumer but also knowledgeability about the brand. These are all factors in the process of identifying the "genuine brand ambassadors".

[0005] Finally, not every one of the brands has the same definition of what a concept of the "genuine ambassador" might be to the brand concerned. The system provides access to a few types of the ambassadors that the brand might find interesting. These types are not set in stone and evolve with time as marketing theories evolve. A few of the current types are included as examples.

Summary of the Invention [0006] A method of calculation of powers of interactions between a brand and a consumer of the brand to determine an ambassador score is disclosed. The method comprises recording a plurality of interactions between the consumer and the brand, calculating a set of metric values relating to the interaction by applying a context matrix to one or more of the recorded interactions, whereby the context matrix is representative of the context of the interaction and subsequently analyzing the stored metric values to determine an ambassador score. The ambassador score is used to determine the value of the consumer to the brand as an ambassador for the brands.

[0007] In one aspect of the invention, an algorithm is used to determine the ambassador score and this algorithm weights the store metric values against a population of consumers to "normalize" the store metric values. [0008] The method of this disclosure defines one type of the interactions to be a referral interaction in which a first one of the consumers refers a second one of the consumers to the brand and one of the metric values is representative of the influence of consumers on the brand. These referrals are particularly useful in determining the consumers who are ambassadors of the brand and thus increase the ambassador score. This increase can be done during the analysis of the stored metric values. The metric value representing the influence of the first one of the consumers is boosted by other metric values of the second one of the consumers. [0009] In one aspect of the disclosure there is no single type of ambassador score, but a plurality of ambassador scores that are calculated by using a plurality of different types of scoring algorithms. The calculated plurality of ambassador scores can be compared in one aspect with pre-stored ranges to determine an ambassador type. [0010] The system for calculation of powers of interactions between a brand and a consumer of the brand comprises a collection service for receiving data relating to the interactions from a plurality of interaction sources, a database for storing the data relating to the interactions, a context matrix for storing algorithms relating to contexts of the interactions and an analysis engine for calculating a set of metric values relating to the interaction by applying the context matrix to one or more of the data of the recorded interactions, and analyzing the stored metric values to determine an ambassador score.

Description of the Drawings [0011] Fig. 1 shows a non- limiting example of the system for calculating the intensity of interactions between consumers and brands.

[0012] Figs. 2a and 2b show the structure of the database. [0013] Fig. 3 shows an outline of the method of calculating the degree of the interactions. Detailed Description of the Invention

[0014] Fig. 1 shows a system 10 used in the collection of interactions for analysis and processing according to the method of this disclosure. The system 10 comprises a collection service 20 that is connected to the Internet 30 via a web connection 40 and accepts web requests 50 from a plurality of interaction sources 60. These interaction sources 60 include but are not limited to interaction with a website or a web shop, visits to a physical store, viewing a relevant video, liking a product, etc.

[0015] These web requests 50 will typically contain relevant data about the interaction between a consumer and a brand. Such relevant data includes a consumer identification 52 and a brand identification 54 as well as an interaction type 58. The web requests 50 may also include a referrer identification 56. The web requests 50 will be stored by the collection service 20 in a database 25 for further analysis by an analysis engine 70.

[0016] The referrer indication 56 is indicative of the person that referred the consumer (identified by the consumer identification 52) to the brand (identified by the brand identification 54) in question. The referrer indication 56 may be identical with another consumer identification 52 as generally most referring persons will also have been identified as one of the consumers interacting with the brand, or the referrer indication 56 may be independent of the consumer identification 52 or have another method of linking the referrer indication 56 to the consumer identification 52.

[0017] The interaction type 58 comprises information detailing the nature of the interaction. This interaction type will include a meaning identifier 58a that describes the meaning of the interaction, a time stamp 58b indicating the time and date at which the interaction took place, a context indication 58c to identify the motivation behind the interaction and, optionally, some data 58d that is specific to the interaction itself. The concept of contextualization will be explained later.

[0018] The system 10 also includes an analysis engine 70 to determine the genuine ambassadors for a brand as will be explained in connection with Fig. 3. The analysis engine 70 can access the data base 25 and process the stored web requests 50 to determine numerical ambassador scores 75, which indicates the level of genuineness indicative of whether the consumer (identified by the consumer identification 52) is one of the ambassador types for the brand. These ambassador scores 75 are calculated by taking into account the power of interaction the brand has assigned to one of the interactions as well as a set of filters and multipliers that leverage the times and the contexts of these interactions to better determine the ultimate value of the consumer to the brand.

Interaction sources [0019] As mentioned, the collection service 20 accepts the web requests 50 detailing the interactions between the consumers and the brands from a variety of interaction sources 60. Examples of the interaction sources 60 will now be described.

Website activity tracking

[0020] One example of the plurality of interaction sources 60 is the tracking of website activity. This tracking is done by an administrator (or programmer) including a script 84 one or more of the web pages 82 of a website 60. The website 60 could represent an online shop, a magazine, a blog, a crowd-sourced review service, such as TripAdvisor, or similar, but this is not limiting of the invention. The script 84 is similar to the analytics scripts such as Google Analytics or Kissmetrics. The script 84 is included on those web pages 82 of the website 80 that need tracking and will be processed by a browser 86 running on a device used by the consumer. The device can be a computer or a smartphone, but this is not limiting of the invention. The script 84 is able to detect the interactions between one of the consumers and the website 80. For example, the script 84 is able track visits by individual ones of the consumers to the webpage 82 on which the script 84 is placed. The script 84 can also be configured to track more advanced interactions between the consumer and the web page 82, such as clicking on a specific button displayed on the web page 82, such as a "like" button.

[0021] Once the script 84 has detected the interaction, such as the visit by the consumer, the script 84 will generate an appropriate one of the web request 50 and will communicate this web request 50 to the collection service 20. These generated web requests 50 are built using JavaScript, with the URI pointing to the web address of the collection service 20 and containing all of the interaction data about the interaction as GET-parameters. Once the URI is constructed, the script 84 will trick the browser 86 into thinking that the script 84 wants to request an image resource from that URI. The browser 86 will try to fetch the image from the URI. On receipt of this collection request, the collection service 20 parses the interaction data from the URI and returns a (empty, lxl pixel) GIF file to the browser 86. This technique is more commonly known as pixel tracking and broadly used by advertisers.

[0022] An example of the website tracking will now be described. Let us suppose that the administrator of the web page 82 in question includes the script 84 onto the web page 82 and configures the script 84 to detect that someone is viewing a video on that web page 82. The administrator specifies an identifier and the context, identified by the interaction type 58, for this interaction (e.g. 'video.watched' as the identifier and 'LOVE' as the context). Every time one of the visitors (consumers) to the web page 82 watches that video, the script 84 will know of this visit and respond by generating the URI containing the interaction data and requesting this URI from the collection service 82. One non-limiting example of the URI would look something like this:

http://cs.bubobox.com/beacon?n=video.watched&c=LOVE. The web request 50 can also indicate the length of time for which the video is viewed by the consumer, as this can indicate the degree of interest in the video by the consumer. The script 84 will also include an identifier of the consumer, the brand and optionally the consumer who referred this consumer. A complete URI would therefore read: http://cs.bubobox.com/beacon?aid=F43AEB4E323&rid=FFAAFFA AFFAA&uid=435 &c=LOVE&n=video.watched

[0023] Other examples of interactions on the website 80 are the number of visits to the same web page 52, clicking on a sign-up button, playing a game, etc.

[0024] Building on this general script 84, it would also be possible to create extended scripts 84 that are pre-configured to handle certain types of the interactions. Taking the video-viewing example from above, the script 84 could be provided on the web page 82 that detects any embedded YouTube videos on the web page 82 and by way of the exposed JavaScript- API of these embedded videos, ask to be notified in the event of one or more of the embedded YouTube videos being played. On notification of the playing of the YouTube video, the script 84 would react by sending the web request 50 to the collection service 20 with the interaction type 58 'y° ^uturj e.video.wachted". There is no need the brand to configure this behavior on the web page 82.

[0025] Other examples of these extensions are:

· Pre-configured tracking scripts for various digital marketing suites like the

BuboBox challenges package, Shortstack, Woobox, Offerpop, ...

• Wordpress plugins

• Forum- software integration

Non-limiting examples of these extensions are outlined below.

Offline Interactions

[0026] A further source for the interactions to take into account, which is often overlooked because of the focus on the internet, are offline interactions between the consumer and the brand. The brand might, for example, be running a coupon promotion which lets their customers receive a bigger discount by getting their friends to come into the (physical) store with them. These kinds of offline interactions are not inherently digital, so the interactions cannot be sent to the collection service 20 automatically. To combat this issue, the brands are provided with an interface 27 to the collection service 20, which the brand can use to turn any data collected during an offline campaign into types of the interactions that can be processed by the collection service 20.

[0027] An example will serve to illustrate this offline interaction. This relates to the filling of personal information on the coupon to get a 20% discount on one of the brand's products. The brand creates some plain coupons that the brand prints in a magazine (or indeed downloadable from the Internet). The customers can cut these printed coupons out and fill in their name and email address on the cutout coupons to make them valid in the brand's store. Every one of the coupons that is used in the stores would now be typed into a list or scanned using image-to-text software to generate a file containing the names and email addresses of the consumers that used the coupon. [0028] The interface 27 will now allow the brand to import the name and address file into the collection service 20. The brand can manually define which types of the interactions took place between the brand and the consumer in the imported list and all of these offline interactions will be sent to the collection service 20. [0029] The interface 27 is programmed as an API and the collection service 20 is publicly available. The brand can develop an in-house application to better automate these kinds of interactions. For example, the brand could develop an application that monitors a folder for new ones of the name and address files and put every scanned coupon into this folder, to which the application would respond by applying some image-to-text processing on the scanned coupon and automatically send the resulting data as the interaction to the collection service 20.

Email tracking [0030] Tracking the opening of emails 90 is a further example of the interactions. This tracking is done by way of including a tracking pixel 92 that is actually an implementation of the web request 50 to the collection service 20 inside the email 90. These tracking pixels 92 are essentially a static version of the same web requests 50 that are dynamically generated by the website activity tracking script 84 described above. Generally it is not possible to run scripts within the emails 90, as many email readers will not process the scripts 84. In this case, the URI's for these tracking pixels 92 need to be constructed prior to sending the email 90.

[0031] The tracking pixel 92 can then be included as an (empty) image and (most) email clients running on a computer or smartphone will then try to load the image when the email 90 is opened, which results in the web request 50 being sent to the collection service 20. The collection service 20 is thus aware of the interaction that is opening the email 90. The brand can customize these tracking pixels 92 to alter the interaction and the brand will record the type of interaction according to their needs. Non-limiting examples of the interaction types include a 'mail.opened.newsletter' interaction type that the brand can include on their monthly newsletter, and a different 'mail.opened.promotion' interaction type for a batch of emails 90 sent out for a specific promotion.

Link tracking

[0032] Some webpages 82 or websites 80 will not allow the use of the dynamic tracking script 84 disclosed above. This issue can be resolved by a "tracking- link" creator 100 that brands can use to turn regular links on the websites 80 into tracking links. The programmer of the website 80 would input the original link into the tracking link creator 100 and choose the interaction type (e.g. 'link.clicked.facebook-page') to attach to the clicking of this link. The creator 100 will now store the original link 103 in a link data store 105 and this interaction in the link data store 105 and return a tracking link that contains a link identifier 107 that can be used to find the original data again. This tracking link is shown to the brand.

[0033] Suppose the consumer clicks this update link. One of the web requests 50 will first be sent to the collection service 20. The collection service 20 will use the identifier in the updated link to look up the original link 103 and attached interaction type. The interaction type will be recorded in the database 25 and the consumer will be redirected to the original link 103. [0034] This concept of link tracking can also be used in the emails 90. The brand can not include the tracking script 84 in the email 90, as mentioned before, but the brand can still track the interaction of clicking one of the email links 94 in the email 90. The brand can replace the original email link with a tracking link as the email link 94 to achieve this. The tracking link is generated by the creator 100 as described above. Collection service

[0035] The collection service 20 is the endpoint to record the interactions between the consumers and the brands. The web requests 50 received by the collection service 20 have a uniformly structured set of data detailing the interaction.

[0036] The web requests 50 will contain the following parameters which are required to record a valid interaction:

[0037] Let us take an example: one of the consumers, ambassador F43AEB4E323, who viewed a YouTube video in context 'LOVE' of the brand 435. The ambassador was referred by ambassador FFAAFFAAFFAA. The consumer viewed the video (http://www.youtube.com/Ge45Rt) for 45 seconds.

[0038] The interaction data will be as follows: ambassador: 'F43AEB4E323',

referrer: TFAAFFAAFFAA',

brand: 435,

context: 'LOVE',

name: 'video .view',

meta: [

{ name: 'youtube_url', value: 'http://www.youtube.com/Ge45Rt' },

{ name: 'time', value: 45 }

]

[0039] The URI for this web request 50 would look like this:

http://cs.bubobox.com/beacon?aid=F43AEB4E323&rid=FFAA FFAAFFAA&uid=435 &c=LOVE&n=video.view&m=~(~(name~'youtube_uri^

utube.com*2fGe45Rt)~(name~'time~value~45))

[0040] The metadata 59 is contained in an array of key- value pairs. This way the web request 50 can contain as much of the metadata 59 as the web request 50 needs without the collection service 20 being aware of what the metadata 59 actually means, but still is able to process the interactions.

[0041] The collection service 20 will now store details of these interactions received via the web requests 50 into the data store 25. The data store 25 has a plurality of records. Each of the records has a combination consumer-brand-context-timeframe. In other words, each of the records done by the consumer X for the brand Y in a context Z during a time frame T (which is in one aspect of the disclosure "the previous minute", but could be set to be other time frames) is contained in a single record recording the interactions. In a typical SQL structure this would have the form: CONSUMER X BRAND Y SEGMENT Z TIMEFRAME T INTERACTIONS

Gert Cola Promotion 1 Minute 1 - Visit

- view video

- share

Gert Cola Promotion 1 Minute2 - share

Gert Cola Promotion2 Minute2 -share

-visit

Rob Cola Promotion2 Minute2 -share

-visit

Gert Cola Promotion3 Minute2 -view

Gert Pepsi Promotion 1 Minute 1 -visit

-view

[0042] This concept is coded as shown in Figs. 2a (a general version) and 2b (a specific version) in which one of the records 210 has a metadata including the consumer 215, the brand 220, the context 225 and the timeframe 230. The record 210 contain details of multiple interactions 240-1, 240-2.

[0043] The record 210 also includes details of the interactions for which the consumer has acted as a referrer by referring another consumer (who has his or her own ambassador record 210) to the brand. The interaction 240-2 leading to the referral will be persisted in two entries in the data store 25. One entry will be included in the referring consumer record 210 and the other entry will be included in the referred ambassador record 210. The referring ambassador record 210 will indicate that the interaction is a 'referred- interaction' instead of a regular one of the interactions. [0044] The MongoDB database structure is used as a non- limiting example to store the multiple interactions per record. This is a document-oriented, schema less data store 25. In this type of data store 25, every one of the records 210 has its own ID. The interaction 240 recorded in the records 210 can otherwise be freely filled in with key- value pairs. The program supports a number of different types of data structure, including but not limited to numbers, strings, dates and Booleans as well as arrays and collections. Collections are sets of key- value pairs that have the same possibilities.

[0045] The ID of a record is constructed by appending the aforementioned identifiers: "[ambassador_id]/[user_id]/[context]/[timestamp]". Besides the ID, it will also have a collection called "metadata" that contains all these identifiers again as separate values. [0046] Finally, there are two more collections, "interactions" and "referred-interactions" that will contain the interaction data for that record. The keys for the key- value pairs within these collections are the names of the interactions.

[0047] The interaction will be represented by a collection that has two properties associated with the collection: "count" 250 and "meta" 255. The property count 250 is a simple number that contains the amount of times this type of interaction was recorded for the same consumer 210, brand 220, and context 225 during a specified timespan 230. The property meta 255 consists of a collection that contains all the metadata for these recorded interactions.

Analysis process [0048] The analysis process will be run at a set interval (i.e. once every day) to update every consumer's ambassador score based on the interactions that were stored for the consumers since the last update. There are, in addition to the time span, two other components that can be used by the analysis process: the brand's interaction library and the previously mentioned contexts.

Interaction library

[0049] Every one of the interactions of the consumer with the brand is interpreted by the system 10 and assigned an interaction power that represents how important the brand values the interaction, thus the degree of interaction between brand and consumer.

[0050] Every brand can configure their relevant interactions and powers of interaction to allow a flexible system that can accommodate different kinds of brand. This configuration is done using an interaction library 28. The interaction library 28 is a key- value store in which the brand can define the name of every one of the interactions that should be taken into account during the analysis process as well as the interaction power belonging to the interaction. The interaction power can also be calculated by an arithmetic formula that leverages values from the metadata of the interaction to calculate the actual interaction power for the interaction. One example would be to use various values obtained from watching the video mentioned previously, such as the number of seconds watched as well as links to and from the video, and then calculating this interaction power from these values.

[0051] A few entries in the interaction library 28 might look like this:

Contexts

[0052] The brands can define a plurality of different contexts for their interactions with the consumers. The context 58c of an interaction signifies the motivation behind the interaction and will direct the system's analysis process (shown in Fig. 3) when the system 10 is turning data about the interactions into metrics that can be used to calculate a final score, which represents a level of genuineness indicative of whether the consumer can be considered one of the genuine ambassadors for the brand.

[0053] The brand will have its own context- library 29 in which the brand can configure the behavior of different ones of the contexts and, optionally, create their own contexts. The context-library 29 can be visualized as a table or context metric as follows:

Engagement Knowledge Influence Greed

LOVE Multiply by 3 Multiply by 0 Multiply by 0 Multiply by 0

EDUCATE Divide by 2 Multiply by 3 Multiply by 0 Multiply by 0

WIN Multiply by 1 Multiply by 0 Multiply by 0 Multiply by 3

INFLUENCE Multiply by 0 Multiply by 0 Multiply by 3 Multiply by 0 The columns represent one of the metrics provided by the system 10, while the rows represent the different contexts that the brand can use when recording the interactions. The cells that are not headers contain instructions processed by the system needs to take the interaction power and turn the interaction power into the different metrics for each column. The headers are non-limiting and can be changed as required.

Example:

[0054] An interaction with power 10 and the context EDUCATE would result in the following set of metrics:

[0055] An example of the use of contexts relates to the use of an interactive FAQ through which consumers of the brand can help each other by responding to their questions. A user providing correct answers should be detected by the system as being knowledgeable as well as engaging with the brand, while the users that ask questions are only engaging and sharing answers on social media might indicate influence in the user.

[0056] Each one of the interactions could have its own context coupled to the interaction. In the example above, this would mean the answering a question would be tagged with the "EDUCATE" context, while asking a question could be "ENGAGEMENT" context or "CURIOSITY" context.

[0057] So, as mentioned above, every interaction that is sent to the collection service will have a context attached to the interaction. Method of Analyzing the Interactions

[0058] The method of analyzing the interactions to become an ambassador score for every consumer consists of several steps and is shown in Fig. 3. The method is carried out in the analyzer 70.

[0059] The first part of the analysis method starts at a set time in step 300. The new interactions are accessed from the data store 25 in step 310 and in step 320 the new interactions are compared with the interactions in the interaction library 28. If one or more of the new interactions are not in the interaction library 28, then the new interaction can be dropped as being irrelevant in step 330. If the new interaction is in the interaction library in step 330, the interaction power of the new interaction can be read in step 340 from the interaction library 28. The interaction power may also be calculated in step 340 by using the appropriate formula from the interaction library 28 and the metadata of the interaction. One example of the calculation in step 340 is for the watching of a video which is calculated using the formula i "interaction power = 2 * [seconds watched]".

[0060] After calculation or reading of the interaction power for the interaction in step 340, the second part of the analysis process will leverage the brand's context library 29 to output the set of metrics in step 350. As explained above, the analysis process will take the interaction power and, depending on the context, apply the rules in the context library 29 (like multiply by 2) to generate an individual metric value for the metrics in step 345.

[0061] The method will check if the interaction is a regular interaction or a referred- interaction in step 350. In the case of the referred interaction, it will generate a sum of all the values generated by the previous step and assign this metric value to the influence metric in step 355 while setting all other metric values back to zero.

[0062] Once all the interactions have been processed, the method has a set of set of metric values for the metrics engagement, knowledge and influence. These metric values can be leveraged in the third and final part of the analysis process. [0063] The third part of the analysis method starts in step 360. The third part of the analysis method will weigh the metrics values calculated above against the population. Here, population means the collection of all of the consumers for a specific brand for whom data has been calculated. The weighing process turns the regular, numeric metric values into percentage values. For example, if the consumer A has a metric value of 50 for engagement and the sum of all of the metric values for engagement across the population is 200, the consumer A's weighted engagement value becomes 25%.

[0064] Once these weighted values have been generated, the analysis process will run in step 380 through every type of ambassador that is defined in the system 10 and apply the rules to the consumer's metric values to generate a final ambassador score 75 that indicates how closely this consumer matches the recorded types of ambassadors. The rules of the ambassador type are algorithms that take as inputs the metrics (both the weighted values and the original metric values) and output the final score 75 for that type of ambassador. Non-limiting examples of the ambassador types and the way of calculating the algorithms are:

The real deal [0065] These are the ambassadors that possess all the "best" characteristics and none of the "negative" characteristics. The real deal ambassadors represent "genuine brand ambassadors". The metrics used for the calculation of this type of ambassador and whether the metrics are positive or negative depends on the type. In the context of this disclosure, the positive metrics are engagement, knowledge and influence, while greed is considered negative.

[0066] The algorithm that measures scores for these metric types would take the unweighted (original) scores for the metrics engagement, knowledge and influence and calculate similarity of each of the metrics. Since this type of ambassador requires the consumer to exhibit all of the positive characteristics equally, the closer the positive characteristics are aligned, the greater the final score. Furthermore, the greater the value of the greed metric, the more this consumer will be penalized, thus lowering the final score. Finally, an average of all of the positive weighted scores would be taken and used to multiply the score that was calculated by the previous steps. The system has now calculated the final score that is indicative of how closely this consumer matches the "real deal" ambassador type as well as how important this consumer is in relation to other ones of the consumers.

Price hunter

[0067] The price hunter is a special kind of ambassador. These consumers may drive a lot of engagement, but their motivation is primarily personal gain instead of brand love. They might be the people that only participate in the brand's contests where there are prizes to be won. However, the fact that the consumers do participate can still be important from a marketing perspective and so the consumers should not be ignored by the system. [0068] The algorithm that identifies consumers that match this type is heavily geared towards the greed metric, which is a positive metric for this algorithm. There are no negative metrics here, so the higher a consumer's greed score, the higher this consumer will rank for the price hunter type. Expert

[0069] Experts are the consumers who exhibit a lot of knowledge about the brand. The algorithm works similarly to the price hunter algorithm but focuses on the knowledge metric instead.

Influencer

[0070] Influencers are the consumers who cause the most interactions from other ones of the consumers. The algorithm works similar to the price hunter algorithm but focuses on the influence metric instead. Applications

[0071] Content challenges [0072] One method of engaging with consumers is the so-called brand challenge in which consumers are encourage through rewards to generate content for the brand. One example of this is produced by BuboBox and is called the challenges package. The brands use this package to create marketing campaigns that are geared towards user-generated content. This package generates the interactions between the consumer and the brand. Examples of the interaction include, but are not limited to, uploading a picture, creating a video, viewing content, sharing content, voting on content, etc. An extension can be built into this package to record the interactions and send the recorded interactions as web requests to the store. Forum integration

[0073] Forums are a further source of interactions between the brand and the consumer. The consumers use the forums to generate events, such as ask questions, express opinions, share knowledge, etc. that the consumers wish to share with their peers and the brand. The forum is provided a pre-built extension into the forum software that hooks into these events and responds by sending the events as interactions to the collection service.