To dive deeper into the measurement method behind our new digital currency, Viewed Impressions, we interviewed Esben Hørning Spangsege, Head of Alpha at AFA Decaux. Esben works on innovation within the digital space and collaborates closely with external partners and the international data division at JCDecaux. He has played a central role in developing a new digital currency at AFA Decaux and has been deeply involved with the project for the past six years.
We sat down with Esben and asked him a series of questions to shed more light on the measurement method, gain a better understanding of how it works, and explore the underlying calculations and methodological validations behind it.
Why have we created a new digital currency?
We developed a new digital currency because Share-Of-Time (SOT), which we used previously, doesn't provide information about how many people actually see your campaign. Our new currency, Viewed Impressions, does exactly that.
Why is this currency more transparent, precise, and secure?
The reason we believe our new currency is more transparent is that it now accurately reports how many Viewed Impressions your campaign receives, not just how much screen time it gets. We believe this is a more transparent and meaningful currency.
Esben Hørning SpangsegeWe actually believe this is one of the most precise impression currencies in the world for outdoor advertising.
Head of Alpha, Digital Business Innovation
AFA Decaux
The precision refers to the fact that we measure every location, every hour, down to the second, every day, every week, and so on. This gives us a large amount of real-time data to document our Viewed Impressions. Additionally, we have received advice and support from various partners who have verified that our approach is statistically robust. So, we believe this is one of the most accurate impression currencies in the world for outdoor advertising.
Security comes from ensuring that, as an advertiser, you have confidence in your media purchase. With our approach, we measure impressions while your campaign is live, allowing us to document exactly how many Viewed Impressions your campaign receives. Just like with other media types such as TV, radio, and online, as an advertiser or media agency, you should only pay for what you actually get. This way, you have the assurance that the campaign delivers as promised.
How does our measurement differ from other types of outdoor measurements?
When measuring Viewed Impressions for outdoor advertising, the process can be divided into two main elements.
The first key element is measuring traffic flow, which involves collecting data to estimate how many people are present near a given screen at a specific time.
The next key element is incorporating data and modeling to adjust the traffic count, so only those people who are highly likely to see the screen at the time a specific campaign is being shown are included in the measurement.
Our measurement differs from most other types of outdoor measurements by using a highly precise method for documenting traffic flow. The second element, which involves adjusting the traffic count, follows the same principles and standards that have been globally adopted within the outdoor industry.
How do we measure traffic flow?
We measure traffic flow using sensors embedded in all our digital screens. These sensors detect mobile phones within a defined radius around each screen, capturing data down to the second.
This method contrasts with the common approach of using, for example, public transport passenger counts on a monthly or weekly basis, or municipal road traffic counts to estimate traffic near a given screen. The issue with those data sources is that they are often quite imprecise. For example, a municipal road traffic count could be several years old or conducted far from where the screen is located.
Esben Hørning SpangsegeWith our method, we have real-time data for the exact period a campaign runs. That’s why our currency is quite precise compared to the rest of the market – both nationally and internationally.
Head of Alpha, Digital Business Innovation
AFA Decaux
With our method, we continuously gather data for the exact time a campaign is running, making our currency much more accurate compared to the broader market, both nationally and internationally.
How can we be sure we are not measuring too many or too few people?
The number we need for our Viewed Impressions calculation is how many people are present, but our sensors measure how many mobile phones are detected. The two numbers are not identical.
For example, a person carrying two phones might walk past our screen, resulting in a double count. Conversely, someone might pass by without their phone being detected. So, we need to adjust the numbers before calculating the final Viewed Impressions.
To ensure we aren’t counting too many or too few, AllUnite, in collaboration with Nielsen, has conducted extensive manual counts in all environments where we have screens. They then compared these numbers with the mobile phone counts we recorded during the same time frames. Based on this data, AllUnite can model how many people are actually present.
In simple terms, the model creates rules like: if there are X number of mobile phones, there are Y number of people present. It is a bit more complicated than that, which is why we have compiled all the details in a methodology document that explains the method in more depth. The entire model has been verified by Nielsen.
Not everyone can see the screen or chooses to look at it. How do you account for that?
This brings us to the second key element in outdoor measurement, where we apply the same global standards and principles accepted across the outdoor industry for solid measurement practices.
First, we adjust the traffic flow numbers so that only those people who actually have the possibility of seeing the screen are included. For instance, if someone is approaching the screen from behind and has no chance of seeing it, they are filtered out of our dataset.
Next, we calculate what is called a Visibility Adjustment Coefficient. For our measurement, we use Ipsos to calculate this. When Ipsos calculates the visibility adjustment coefficients for each of our digital screens, they consider many factors and variables. These include obstructions that might block the view of the screen from certain angles, the size of the screen, and the various angles from which people might approach the screen.
Ipsos boils this down into a single number, which further reduces the traffic flow count, so we are left only with the people who are highly likely to see the screen.
Finally, the reduced traffic flow data is matched with the campaign scheduling at the specific time the ad is running, ensuring that the traffic flow is applied to the correct campaigns. This algorithm, developed by JCDecaux's Data Division, has been validated by the global media measurement organization, CESP.
We believe that, along with our partners, we have created a solid, transparent, and precise digital currency that offers security for your media investment with us.
Partners and their specialist competencies