Real uses of Big Data in Marketing and Advertising
The most important thing about Big Data is knowing how to make an intelligent use of the data, as we saw in our post about analytics and metrics. The first guarantee of success will be, on one hand, knowing how to identify the questions that are of interest to your business, and on the other hand, having the necessary tools and technology (whatever they are) to collect and process the data that will help answer these question.
If you are part of a large company with a digital presence, you probably have read a lot of posts talking about Big Data and all the benefits that the use of data can have in your business. However, it is true that these posts are usually too generic and offer only bird's eye view of the real applications you have in your company's day to day.
Before starting, there are two notes to address. When we talk about Big Data, we talk about a huge amount of information both structured and unstructured (really, you start talking about Big Data from several tens of terabytes). This information would not be possible to treat it manually and of course it is not something manageable since the classic excel, so the benefits obtained from it are not comparable to other techniques of data analysis that we usually perform.
Two uses of big data in marketing
In this post we will try to explain two practical cases about how data can be a generator of change in marketing and advertising. In addition, we have tried cases to be replicated immediately and not necessarily require a large investment in servers or high-cost technologies.
Reputational assessment of your audience through data
The techniques of text analysis allow us to know the polarity of comments or mentions they make about our brand. The habitual feeling analysis is an automatic process assigned to an opinion expressed to be positive, negative or neutral value. Normally, the analysis of feelings works with the analysis of a text as an inseparable block.
Taking a step further in the technological, we can apply the same technique separating the components of these texts, since in the same news, in a Twitter publication or in a message sent from a forum, users often talk about more than one thing, with a different feeling for each of those themes. The analysis of sentiment by entities allows to analyze within the same text chain the polarity assigned to different entities that are mentioned in it, such as products, brands or people. In this post we can find a very interesting case study.
A few months ago, Google celebrated its annual I/O, an event where the innovations that are going to implement the already developed products as well as the next expected releases of the company are presented.
Analyzing more than 8,500 news stories that spoke about this event, a sample of the 20 key words that appear in publications that mention Google´s new technology, Duplex, an artificial intelligence that, for example, will be able to call by phone to make a reservation in the restaurant without a human interlocutor can distinguish talking to a robot.
In the following table, we show 20 keywords mentioned in these articles and related to artificial intelligence, like other companies that use it in their technology.
The author draws attention to two points:
- Ethics is mentioned in a more negative tone than any other word. It seems that the ability of the machine to interact with the human being generates some doubts about how this can be applied in an uncontrolled way.
- Both Amazon and Alexa are mentioned with a much more positive polarity than Google.
Seeing these results we can recommend Google some actions for its next annual conference since it seems that this time the coverage to this conference has been more to favor Amazon than themselves.
Data at the service of creativity in advertising campaigns
Most of us who are dedicated to marketing believe that we have reliable insights about our audience and what kind of creativity can find attractive.
Normally, we look at design which is something that all (or most of the audience) have in common, and use it as a basis for a single campaign.
We can also create multiple creatives to attract different segments of our audience, with the temporary and monetary cost that this requires.
However, there are already some cases of brands that have created custom designs to get attention thanks to all the data.
In a marketing campaign at Caesars Palace, a video campaign was conceived where the ads showed all the activities that guests of this famous hotel in Las Vegas can do, using #LikeACaesar as a meeting hashtag for potential customers.
Prior to the release of the video, ideas were collected from his audience through Google Consumer Surveys. When asking questions like ¨Which Las Vegas resort would you describe as the most modern?¨, they discovered that the younger audience did not perceive Caesars Palace as the brand wanted to show itself, and using all this quantitative and qualitative information, they could understand more deep to its demographic core.
Turning now to creativity, and using Youtube Director Mix, a tool that allows mass customization of the creatives, for this campaign a 30-second video was created with a brand message that could be customized to attract different interest of the audience, a simple change. The result was more than 150 different variations of that message, which for example used trends and events relevant to the chosen audience.
With the hundreds of creative variations, a media plan was created to ensure that each message reached the closest audience. For example, people who had shown interest in nightlife or were food lovers. They also used custom affinity audiences to find people who recently searched for relevant keywords related to events, such as the Country Music Awards, Coachella (a music and art festival) or the finals of a hockey championship.
The results of this action were a visualization rate of 60& when the reference index in the travel category is 42%. The ads generated 11 million minutes of viewing time for their users.
In addition, they managed to get real people to answer critical questions about their brand, diagnosing a perception problem and, of course, correcting with hyper-relevant video ads for the passion points of the audience.
For a long time, INNN has tried to create instruments that manage this flow of information. Applying data analytics to strategic brand plans we create scorecards with the main indicators that interest our clients. In our constant effort for innovation we are working on applying the data to business decisions in areas of communication such as politics, sports, and, of course, online products.