"We want to increase the quality of advertising"

The future will bring us less advertising—but better and more targeted advertising. Specifically, this refers to the next stage of development in mobile advertising targeting. Adello is introducing the Swiss Mobile Tool/AdCTRL for this purpose.

Unbenannt-2_5
   
Mark Forster (left) is the founder and Chairman of the Board of Directors of Adello. Sandor Laczko works in Zurich as Managing Director.
 
WW: What is Swiss Mobile Tool/AdCTRL? Mark Forster: In this IT project, an analytics team evaluates success factors from historical data sets and in real time, enabling a statistically relevant prediction of how recipients will react to future advertising campaigns. For the first time, user interests can be compared with the targeted advertising. From the user's point of view, the number of unwanted advertisements can be significantly reduced. "AdCTRL" is being developed by a leading international team of experts in Switzerland, Germany and the USA with considerable financial and human resources and is unique to date. We will be ready to launch it at Adello in February. Tests are currently underway on hundreds of millions of data records. In technological terms, it's all about "big data". Companies such as Facebook rely on similar architecture.
 
What idea is this tool based on? AdCTRL originated from our "Swiss Mobile Tool". We developed this in 2011 and started using it internally in 2012. The idea was that we wanted to have a way to measure: How do users react to different campaigns? We recorded and evaluated a few hundred campaigns until we could say: This particular campaign was better than the average or performed better in a particular week than the rest of the time. This allowed us to optimize campaigns manually. Then we saw: We could actually do a lot more with the whole topic. This is how AdCTRL was born.
 
So how can success be measured? There are different types of measurement. There is the hard and fact-based measurement. A click is a fact that the consumer has done something. It is also a fact that we can see which device a user is using, what screen size they have viewed something on and for how long. These are all hard facts that we can measure. The complexity lies in the fact that we move around digitally and leave behind different traces. And change certain parameters in the process. With AdCTRL, we are constantly developing our algorithms in order to predict where we can pick up which target group with which reaction.
 
What added value does the further development offer? We asked ourselves: What is the future of mobile advertising? Is it the case that the traditional standard advertising formats - as we know them in print and online - will be reproduced on mobile? Or simply more advertising faster, but cheaper? In our view, quality is the key to the customer. On the one hand, this is defined by the advertising environment, relevance and an understanding of the situation and needs. With AdCTRL, we know what the consumer is doing and why. We also see certain trends in their usage behavior. This allows us to try to involve the consumer in the advertising. On the one hand, we give them a tool that allows them to say what kind of advertising they are interested in and what advertising they are not interested in at all. In other words, for the first time, consumers have the opportunity to say what they want. Until now, they have seen advertising everywhere, regardless of whether they wanted to or not. As a result, certain companies launched ad blockers. This made it possible to turn off advertising completely. All advertising, in fact. However, we believe that this is not the ideal way to deal with advertising, that you either have to choose everything or nothing. I may not be interested in detergent advertising, for example. But I am interested in advertising for cars, investments or a new cell phone. Advertising can certainly be desirable. That's why we said: At this point, it would make sense for us to give consumers a tool that allows them to have a say and decide for themselves which advertising they want to receive in the future.
 
How do I benefit from AdCTRL?
There are two elements to this. As a consumer, I can exert influence via an AdCTRL app. I can use it to set which advertising I want to see and which I don't. We will first test the offer abroad with our partners before launching it in Switzerland. Independently of this, we have been running our analytics for some time now. This already allows us to make certain statements and optimizations. If someone also happens to download the application, then you can match it perfectly.
 
If I never click on a certain product group,
then the advertising no longer appears?
In principle, yes, that's what our models are designed for. AdCTRL represents a departure from the watering can principle. We can also largely rule out misclicks. In return, we try to display advertising that is carefully optimized to the interests and situation of the user. Just as Google optimizes the search results displayed to the individual person, we do the same for advertising.
 
How can this work?
This is a very complex project. We are working with the latest technologies such as Hadoop, Cassandra, Pig. These allow us to run analyses in real time in order to understand: Are there certain patterns of behavior? Can we form segmentations and clusters? With such analyses, we can ultimately not only say what a consumer has done in the past, but we can also observe their current interests and even predict the likelihood of a consumer doing this or that while browsing. In order to be able to calculate this in real time, we are moving into the area of "big data". This requires not only a very high performance infrastructure, which we had to build up, but also the appropriate specialists. At the moment, we have designed AdCTRL for a performance of around one billion unique clients. We don't need that in Switzerland. We have developed a technology that we say must be able to hold its own internationally. In other words, we have to be able to compete with international players.
 
What are the chances here?
We were surprised to discover that even the large, internationally active companies do not see mobile analytics as a priority and mostly outsource it. Presumably because it is enormously cost-intensive and only brings returns in the medium term. We are really breaking new ground here. To be able to do this, we have had to put together a multinational team.
 
What is Adello's job?
We had the idea in Switzerland, developed the concept and secured the funding. However, we quickly realized that we couldn't find the people here who could actually implement our project. AdCTRL is complex. It involves a lot of hypotheses, mathematical illustrations and modeling. For example, we have three specialists who have doctorates in statistics, machine learning and physics. A typical data set that we analyze comprises 10 GB. In addition to specialists who can handle this, we also need the right machines. This cannot be processed on a notebook with Excel. That means investments on a completely different scale.
 
Which partners has Adello found?
We have searched accordingly, especially in the USA. They are much further ahead on the subject there. I was in Silicon Valley and spoke to various specialized companies. We then put together a project team consisting of Americans with developers in Germany and here in Zurich. We are now working on this international basis. We decided against completely outsourcing the development. And we didn't want to have it done in Eastern Europe or Asia either. Due to the complexity, it has to be absolutely top people, otherwise we won't achieve the quality we want.
 
But the lead is with Adello in Zurich?
Absolutely. We have full control, not only financially, but also in terms of the structure.
 
How big is the budget?
We're talking about several man-years. AdCTRL is never completely finished. We have already invested five man-years in the initial phase, and we will have two people working on the project internally each year, as well as at least one permanent external employee. We are investing a good seven-figure sum in this project on our own initiative. Outsourced, it would be around ten million francs. These are dimensions that a single publisher or agency cannot afford. Let alone bring together the necessary expertise.
 
Will it be worth it?
Most publishers simply try to display as much advertising as possible, even if this is not in the customer's interest. We want to prevent that. Just as Google limits searches to the essential minimum, our aim is to achieve the same on mobile. This is in the interest of the customer and therefore also the publisher. The higher relevance and the lower number of ads mean that we can sell individual ads at a higher price. In our view, increasing quality for consumers, publishers and us is the best way forward for mobile advertising in the long term.
 
How to proceed?
We have developed the basic features of AdCTRL, conducted analyses and we know that we can deliver. Now we have to ramp it up with individual partners. We use it in all countries in which Adello is already active today and wants to be active in the future. The major providers in the mobile sector have nothing comparable at the moment. We are trying to extend our lead.
 
Not afraid of a copy?
We have talked to many companies and most of them are optimizing by increasing the number of ad impressions and simple retargeting. A two-man team can build such a technology in a month. A banner that changes three times within ten seconds in order to generate more advertising is nice, but it doesn't interest the customer at all. With AdCTRL, we are moving in the direction of quality. The development costs are so high that there is little chance of anyone catching up with us nationally. We also work on an international level.
 
A Swiss invention will also influence advertising in the USA?
That would be our goal. We are completely self-financed. Every franc that goes into development is no longer available to me privately. I wouldn't be doing this if I didn't believe that we could significantly improve advertising. We are currently focusing on Europe, but in the medium term we will go global. There is nothing comparable in this area so far.
 
Interview: Andreas Panzeri

Unbenannt-1_3

More articles on the topic