AUDIENCE ANALYTICS, FACE ANALYSIS
How to Choose the Right Audience Analytics Solution for Your DOOH Network
The right audience analytics solution helps ensure the success of your digital out of home (DOOH) network. It has a big impact on your return on investment (ROI), since your analytics data affects how easily you are able to attract advertisers to your network, what type of advertisers you are able to attract, and how much you are able to charge. Now you might be wondering how it’s possible that an Audience Analytics solution can be so powerful? To answer this question, let’s first have a look at what exactly an audience analytics solution is and what it does.
What is an Audience Analytics solution in the context of DOOH?
In the context of DOOH, an audience analytics solution provides data on the people who come into contact with the screens within your network. The type of data that such a solution might gather includes demographic data such as age and gender, people count, impression and viewer count, attention time, and even sentiment data.
The power of Audience Analytics
When audience analytics technology is integrated with a CMS, it provides powerful opportunities for both network providers and advertisers to deliver highly relevant and tailored messages, which increases audience engagement and enhances the overall effectiveness of the digital signage network.
Furthermore, if you, as a media owner, are charging for advertising space based on the number of impressions, it’s essential to know exactly how many impressions your screens are getting, and that this data is accurate. In addition, if you know that the demographic profiles of your viewers resemble the target profiles of your advertisers, you can charge a higher price per impression.
What are the options? Wi-Fi vs Bluetooth vs Face Analysis
When it comes to choosing a DOOH audience analytics solution, you have various options. These options differ in terms of accuracy and ease of implementation, and there is no single best option. You need to consider your own set of circumstances and requirements, and then decide what will work best for you. In some cases, you may even want to use a combination of solutions.
Here are some of your main options:
Wi-Fi tracking tracks people within a given area by picking up their smartphones signals when searching for Wi-Fi. This technology can provide a good general idea of how many people are moving through a specific area, where they spend more time (dwell time), and so on. Depending on how the tracking is set up, and whether it is done in combination with free Wi-Fi access, it may be possible to collect demographic data and possibly even email addresses and device data (e.g. through captive portals).
Some software solutions (e.g. Skyfii’s Wi-Fi tracking software) can also send push notifications with offers or discounts directly to devices found within specific areas.
Some drawbacks of this type of tracking are price of implementation, required calibration, limited features (only people count, not actual views can be tracked), and gaps in data. Since Wi-Fi must be turned on for this type of tracking to work, and since not everyone carries a smartphone with them at all times, this is not the most accurate form of tracking.
Bluetooth tracking works in a similar way to Wi-Fi tracking, and requires a smartphone’s Bluetooth to be turned on in order to be tracked. It does, however, offer a few advantages over Wi-Fi. It’s cheaper to set up, and it provides more accurate location tracking. It can also be linked to Google Analytics for attribution modeling purposes. One of the downsides, however, is that Bluetooth cannot be used to collect email addresses and other data by allowing people to log in for free internet access.
Face analysis is generally more accurate than Wi-Fi and Bluetooth tracking because it does not rely on people’s devices but it rather analyzes every individual that enters the camera field of view. It works by placing a camera (or cameras) above a digital display and capturing footage of the audience as they walk past. This footage is analyzed in real-time (no video recorded) by AI-powered Face Analysis Software, after which the footage is automatically discarded to maintain anonymity. The only thing remaining is raw audience data.
Another major advantage of face analysis is that it can determine whether people actually looked at an ad, or whether they merely passed by it (opportunity to see (OTS).
Choosing an Audience Analytics option
For the most accurate results, we recommend using a face analysis solution, possibly in combination with either Bluetooth and/or Wi-Fi tracking if you require the option to send push notifications to your audience when they are within a certain radius.
We know that choosing a face analysis solution can be a challenging task, and that’s why we’ve put together a list of questions you can use to make sure you’re considering all the crucial points.
Questions to ask when choosing a face analysis solution:
- How accurate is it? Can it detect faces at extreme angles and at large enough distances?
- What type of hardware will I need? Does the technology’s requirements match what I’m willing to invest?
- What platforms (operating systems) can it run on?
- How many camera feeds can the solution support?
- Can it give me the metrics that my advertisers want to know? For example, age, gender, people count, OTS, attention time?
- What integrations are compatible with the solution? Will I be able to integrate it with a dashboard and/or CMS?
- Do I need to develop anything myself? If yes, what skills and resources will I need?
- What type of support is provided by the technology provider?
- What will the initial investment be?
- What about data protection? Does the solution come with privacy by default?
Sightcorp’s solution and what it can do for you
Sightcorp’s anonymous audience analytics solutions for DOOH work with face analysis technology and are designed to run on affordable hardware. Since they are based on deep learning, they can deliver high levels of accuracy, even in challenging scenarios such as extreme angles.
Depending on your use case, there are two products for you to choose from: DeepSight Toolkit and DeepSight SDK.
DeepSight Toolkitis a plug-and-play solution that you can use right away, with no hard coding required. We also have ready integrations with dashboards and CMSs, to make it even more convenient for you to get up and running quickly.
DeepSight SDK, on the other hand, gives you the opportunity to integrate face analysis into your own solution using C++.
Both the Toolkit and the SDK incorporate privacy by design, and allow for anonymous data aggregation.
If you’d like to learn more about these products, how they work, and what the benefits are, you can try them out for two weeks for free.