In-App Contextual Advertising In An IDFA-Less World

With the iOS 14 announcement, a lot of app publishers are fretting about their revenue. When cookies were deprecated on Safari, publishers saw monetization rates drop well below 50%. For most smaller app publishers, life without IDFA will be the kiss of death. Larger publishers may have the resources and the scale to weather the storm. With addressability becoming less prevalent, apps will be more reliant on contextual ad products to make up the loss in revenue. How might contextual advertising take shape in-app? Let’s explore.

The web context is quite clear. When a user is on a particular URL reading a particular piece of content, the content gives you a sense of the potential mindset of the user on that page. There are NLP and semantic analysis startups that have been very successful in scanning the content of the web and assigning a taxonomy of context to specific URLs. Grapeshot was probably the shining success story here.

Within apps, context is not the same. There is no open Web where you can verifiably match a URL to a specific piece of content. At most, you can map a user to a specific app and that app in itself has context. Creating context beyond that will be based on trust. Unfortunately under iOS 14, trust will be in short supply. With less addressability, click fraud will be far more prevalent. Media buyers will need to trust that publishers are accurately representing context. The publisher and the enabler of context need to be trusted.

What targeting elements might apps put forward as context for advertisers?

Location data

Even if there is no IDFA present, knowing that a user is at a specific GPS coordinate could be a really powerful signal to target. Note that under iOS 14, users will be able to allow “precise” or “approximate” location. “Approximate” will locate a user within a 10-square mile radius which will mean lower fidelity and less effective targeting – depending on the use case. Other related data attributes, such as how fast a user may be travelling (via the phone’s gyroscope), could be useful.

Content

Where an app is focused on developing content, all the same semantic analysis that powers context on the Web can be applied, but with less transparency. Given the social nature of mobile devices, there will be a lot of consumer-generated content that will be usable for processing and contextual advertising. This includes formats like video and images, and you can expect that computer vision technologies will  be in high demand to create contextual meta data on this content.

Device characteristics

Knowing how much storage someone has on their device, their carrier, SDK footprint, language settings, battery level, and network speed could all come into play to develop a unique contextual understanding of the user.

AdTech data

Context regarding someone’s ad consumption relative to their session could be really valuable. Knowing how many ads a user has seen and the duration of their session will paint an interesting contextual picture.

More to come

Having been following the developments at Apple, we at TrueData have been preparing for an IDFA-less world for some time. Rest assured we are at work (in stealth mode) on new ways to help our partners increase the value of their ad inventory. We will have more to share in the near future.

We are now supporting our agency partners with free Platform Credentials.