iOS 14 and its impact on the location-based advertising market
by Mark Slade, CEO, Location Sciences
On Sep 19, 2019, Apple released iOS 13 which introduced new privacy and transparency settings. For the first time, consumers were actively told by the operating system that apps were using their location data in the foreground and background.
In our 2019 study, The State of Location Advertising, we recorded a 68% drop in relevant location data needed for scaled behavioural audience development and accuracy of targeting. Despite this dramatic change in data sharing by users, there seemed to be very little acknowledgement of the scale or cost impact within the buyer and seller markets. This was a clear indication of hidden, maybe overlooked, data challenges that already existed in the market for location data.
This past week, Apple has announced continued operating changes by way of its Fall 2020 release of iOS 14. Similar to iOS 13, this new operating system update will add greater consumer choice and transparency features that will have have a significant impact on location advertisers – particularly when it comes to behavioural audiences and precise location targeting at scale.
What are the most important changes to the location market in respect to iOS14?
Apple’s IDFA limitations make it far more difficult to accurately target based on previous location behaviours and to attribute a visit conversion
While Apple didn’t completely turn off IDFA (Apples equivalent of in-app cookies), its new limitations raise a number of considerable barriers to effectiveness. Specifically, all Apple users will now be presented with a new permission box whenever an app wants to use that user’s data for ad personalization
Apple’s identifier is attached to all the data points received via an app. It allows companies to build profiles of devices based on their behaviour. This could be ad interaction, offline location behaviour, or identifying which apps a user has on their device – amongst many others. Since its inception, it has been crucial for advertisers to deliver targeted advertising and limit wasted ad spend.
The ability to opt out of sharing this identifier has long been present but buried away in settings. Previously described as a way to receive more relevant ads, it now uses a harsher tone; the use of tracking versus relevancy is likely to make users more privacy conscious and more likely to opt out of sharing this identifier.
As users begin to opt out, advertisers and suppliers that rely on app-based performance attribution will also be affected. With a reduced ability to match IDs, they will find it increasingly difficult to attribute installs and store visits.
2. Apple’s updated privacy settings introduces “Approximate Location”
As an advancement to iOS 13 updates, which gave users more location data sharing choices and regular reminders, with iOS 14, Apple has segmented the globe into its own proprietary 10 square mile radius. This “Approximate Location” is what will now be passed to adtech companies if the user restricts to this option.
Clearly, there will be many apps, such as travel apps, that will still require precise location – but there are many out there, such as weather apps, where the user will deem an approximate location sufficient enough. This further limits the sharing of precise location advertising data for suppliers and marketers as we look ahead. However at a macro level, we could see more sharing of location data as many consumers previously adverse to sharing precise data might feel more comfortable with sharing approximate location.
You can see in the image below, that if the centre of the Apple defined area was Central Park they could in fact be almost anywhere in Manhattan. Advertisers and suppliers who rely on this data will lose precise location confidence for any users who now choose to share their location data as ‘approximate’.
What does it all mean for location advertisers and suppliers?
Challenges to third-party behavioural audiences and attribution puts greater emphasis on real-time advertising
If a user opts out of tracking, it will mean that the developers, third-party data collectors and advertising SDKs will no longer receive the IDFA of the device. They will no longer be able to aggregate user data across different sources, track how users previously interacted with ads or match with user’s offline location behaviour.
For location companies using their DSP to sniff out device IDs and location events in the bid-stream to build audiences, the lack of IDFAs makes it very difficult to group together location events against individual users to build a profile – for example, a device has been to Walmart three times in the last two months.
From a brand safety and consumer perspective, this is the right thing. It has always been questionable that a location DSP can hoover up all these location events without even serving an advert and therefore being referenced properly with consent properly given.
For location companies using third-party SDK data to build audiences, the biggest challenge is going to be binding together this audience data and the media in the DSP. If you are not the audience provider or the media owner, you will be unable to use the workaround of IDFV (identity for vendors). This may push audience and media supply vendors to partner up or possibly start adopting standardised identifiers between them in partner networks.
The same challenge exists at the end of a campaign for store visitation attribution. Location attribution uses a panel of GPS data to bind location events or visits via an IDFA to a campaign. The reduction in IDFAs and precise location events will mean it’s much more difficult to both attribute a visit and to match footfall events to the campaign that delivered them.
The market will of course continue to evolve. Suppliers will look to augment precise location data with other data signals such as purchase data, wifi pings, surveys and check ins. The mixture and makeup of these data sources will be critical to success.
What we can be sure of however, is that the double whammy of reduction in IDFAs and precise location data will cause a sharp decline in audience volumes. Given these market changes and different approaches by suppliers, the onus needs to be on suppliers to open up their black boxes to show their mix of signals used and data collection methodologies. The exchange for this, of course, will be an adjustment to the price. The more precise the data sources are, the more valuable they will be to advertisers.
Ironically, this reduction in behavioural targeting could mean increased demand for real-time proximity targeting in programmatic which requires less granularity in accuracy. For example, QSR companies won’t need to know the device ID is in the restaurant at a point in time but near the restaurant when the media is served.
Without the myriad of behavioural targeting options brands will use the signals left such as approximate location and IP location to help build more context into their targeting. Serving an advert in the right context or location is not a new thing. Let’s remember it has worked very well for the out-of-home industry for years, so simplification may be a good thing for the market. Only time will tell.
2. The creation of a new class of location data – “Approximate Location”
Approximate location data is effectively a new class of location data – sitting right between Accurate GPS and IP data. Although it cannot be used for micro-targeting and store level attribution, it still has great value to advertisers. Brands will now have a choice to target based on precise, approximate or IP data signals. Understanding the type of data now being used to target users and attribute performance will become even more important.
It is critical that brands work with their agencies and suppliers to get the right mix of different types of location data and, as an industry, that this is priced based on a functioning, transparent market and the performance it delivers.
Agency holding company Group M and MRC have recently released new guidance suggesting the need for location data verification in the wake of COVID-19 and the previous iOS 13 impact on data quality. Without total transparency into these source data sets, it becomes impossible to optimize pricing and outcomes within location marketing.
Conclusions: what does this all mean for the location-based advertising market?
The location market will no doubt rise to the challenge of iOS 14 and I welcome the new form of data class. I believe it will be welcomed by consumers which will, in turn, increase the volume of location data being shared – albeit Approximate location.
There is no doubt that the market is a sensitive one for brand safety in terms of consumer privacy. I have to believe that Apple’s changes in iOS 14, and its drive for better consumer transparency in their data, will push the location data market into delivering better industry transparency. The end result will be increased confidence and advertising dollars.
Rather than this being the end of the location-based advertising market (as some are touting) it’s simply a reset in terms of the supply, pricing, and types of accuracy in data.
Let’s not forget that locally targeted ads (with relevant local messages) have worked well for decades. And when you use accurate consented location data, this can be invaluable to retail, QSR and auto companies in delivering ROI.
The introduction of iOS14 could indeed be a watershed moment in the industry. With a lack of third-party audience segments, brands will come back to some basic principles – serve the advert on content your users like, in areas you know your customers are located, at the right time.