DOOH Advertising: Proving Your Network's Value Using Computer Vision Technology3 February 2022
As the world of DOOH advertising evolves, advertisers are coming to expect a similar level of proof of value and return on investment as in the online world.
They look to invest in ad space where they can get better results from targeting precise audiences; there are a number of ways in which real-time impressions (and concurrently ROI) have been tracked in DOOH in recent years.
Traffic and foot traffic data has provided an estimated answer as to how many pairs of eyeballs are on a piece of content. Third party data (for instance that about a particular building’s use) can indicate the types of people that are viewing a digital display.
WiFi tracking has presented another way for media owners to prove the value of their network. Through data from mobile phones, GPS systems and WiFi enabled cars, audience size can be measured. This is, however, big picture data as while wifi tracking can help in detecting people within a certain proximity to the digital display, there is no way to determine the true number of actual viewers.
Computer vision, audience analytics technology is the only data rich solution that provides granular detail akin to that found in online advertising. Through collecting anonymous yet detailed impression data, media owners can really get clear on the specifics of their ad space and increase trust and transparency with media buyers. What this trust and transparency translates to ultimately is more reorders and less unused inventory. A reliable source of data (such as that provided by computer vision systems) is key to solidifying relationships and reliably proving the value of an ad network.
In this article, we explore everything you need to know about obtaining this data and we explain the true benefits of using computer vision for DOOH audience measurement.
What Data Insights Do Media Buyers Want?
Put simply, media buyers look to receive data that makes advertising easier. They need to know that they are reaching the right type of audience for their product.
There will always be insights that can be gleaned from a particular location. For example, one could make an educated guess that a football stadium could be filled with people (mostly men) of a certain age who like football.
However, real-time data insights from computer vision software might in fact show that during certain days of the week, the stadium is filled with a different demographic. Perhaps there is a concert on at the weekend and younger people visit. Maybe there’s an event tailored for women every Tuesday. These are insights that you miss if you focus only on location and traffic volume.
Computer vision technology not only provides demographic insights pertaining to age and gender, but can even indicate levels of audience interest and attention. This audience behaviour data is completely transforming how media buyers advertise and how media owners prove value; in the next section, we will explore exactly how this looks.
How Can Computer Vision Be Used To Analyse DOOH Audiences?
Computer vision is a type of artificial intelligence which allows a computer to “see” video and picture content. Through machine learning (essentially being shown the same content repeatedly) the computer learns to make accurate judgements about what it is “seeing.”
When analysing digital out of home audiences, computer vision technology can look at big picture details such as how many people are in the frame close to the digital display and how they engage with the screen.
On a more granular level, it can also accurately predict age and gender and, using head pose tracking technology, assess levels of interest in the content displayed. This is powerful information for media buyers, who need to understand whether their ad content is performing effectively across the impressions they receive.
Raydiant's computer vision software takes this facial analysis one step further, by analyzing whether an audience member is smiling and therefore showing a heightened level of interest or positive mood.
Since the computer vision software works anonymously, no identifiable information is stored. With privacy-by-design built in. This type of facial analysis is therefore different to face recognition, which relies on non-anonymous data.
Why Do You Need Real-Time Data?
The insights that can be gained from computer vision systems can be collected in real-time. This offers huge advantages to media buyers; providing this information as a media owner allows you to objectively prove the value of your ad space and fuels programmatic trading.
Aside from allowing media buyers to see that they are getting as many impressions as they are paying for in buying with CPM, a key benefit of real-time data is that it adds assurance that their target audience is being met with content. With programmatic advertising, a media buyer’s content will only be shown in a slot in where the computer vision technology has determined that an appropriate audience is present. In short, they will not be paying for ad space that doesn’t deliver results.
With these real-time insights at hand, the impact of a campaign can be truly measured and content can be dynamically rendered to better target the audience known to be viewing the content. Computer vision systems mean no more missed opportunities to capture a certain demographic and far less wasted budget. Since budget is a hotly contended topic within the space, this is critical for both media buyers and owners.
The Benefits of Using Computer Vision Technology for DOOH Audience Measurement:
The benefits of using computer vision technology for audience measurement and proving the value of ad space are clear. Not only do media owners build and retain media buyer trust through objective insights around CPM but they also can attribute audiences to screens, something which other smart measurement technologies cannot provide. In short, media owners can get a lot more specific about the number and types of people seeing a piece of ad space. This inevitably allows them to sell more inventory.
To find out more about our computer vision technology or get started, click the link below.