Adboost: TV/search synchronisation reinvented 

Smart and scalable TV/search synchronisation, based on machine learning 

Goodbye old synchronisation… Hello new and improved TV search synchronisation!
 Your TV ads as well as those of your competitors generate search peaks on Google. Boosting your Google Adwords campaign during these TV appearances will put you at the top of searches and will allow you to retrieve a maximum of TV-engaged visitors. However, not all ads should be synchronised. How should you determine which moment is best to activate a synchronised TV/search? 
The answer: thanks to machine learning. The arbitration of TV/search synchronisation should be based on past performances, which is why Admo.tv launched Adboost, the new generation of TV/search synchronisation. Adboost’s machine learning algorithms automatically choose the best channels and time slots to boost your Adword ads. 
  • Text Hover
  • Text Hover

With TV/search synchronisation comes learning

  • Text Hover
Adboost is a unique tool because its technology understands, compares and retains the performances of millions of synchronised combinations. Here’s why. 
First, a quick zoom in on AdWords and how it works!
AdWords gives access to the hourly performance of ads (impressions, clicks and CTR). It is not possible to go deeper into details and to obtain information second by second or even minute by minute. Yet, it is extremely likely that multiple TV ads are broadcasted during one hour. 
Adboost can model the impact of all of the possible groups of channels on the CTR. In order to do so, it simulates the average CTR that the groups studied would have obtained if only the channels of these groups were broadcasted. Adboost then obtains millions of performances of different groups of channels which can be compared with one another.
This learning phase requires at least 48 hours of analysis and is then pursued continuously throughout the whole TV AdWords synchronisation campaign.

With advanced TV/search synchronisation comes optimisation

After its learning phase, Adboost chooses the best sets of channels for TV/search synchronisation.
In order to not choose groups that contain only one or two channels, the only ones that are retained are those that have at least 4 channels. But ultimately, which groups of channels should really be retained
Adboost’s algorithm starts by categorizing the combinations of the average CTR in ascending order. This creates the green curve shown in the diagram on the right. The average CTR rises abruptly at one point, illustrated by the grey dotted line on the diagram. Adboost’s goal is simple: to launch the TV/search synchronisation only on the the groups of channels that are situated after that dotted line. It then eliminates all of the other groups with a lower performance that would have been chosen in any case by a traditional TV/search synchronisation.  
To determine precisely where this breaking point is, Adboost’s algorithm calculates the derivative (flashback to high school math) of the performance curve of the groups of channels.
  • Text Hover
  • Text Hover

Want to know more about Drive-to-Web analysis ?

Don’t waste any more time: discover the advice from our media analysts to evaluate your media plans efficiency and make the right choices on TV.