With policy updates coming to iOS in 2021, Facebook is making a number of changes to their marketing platform. For web-based, ecommerce advertisers, we’d like to summarize those with the largest impact:
- Conversion Time Reporting: Currently conversions report on Impression Time (the time the ad budget was spent) even if conversions are triggered up to 28 days later. Facebook has said reporting will now “default” to the date of the conversion. We’re waiting to hear if any Impression Time data will still be available. If not, this will impact advertisers’ ability to tie conversions back to the specific date of spend. Advertisers may want to frequently relaunch ads, so conversions and revenue are more tightly cohorted into the time period when the ads ran.
- No more 28-day click or 7-day view attribution windows: Facebook’s UI will no longer display conversion that came in more than 7 days after clicking an ad, and no more than 1 day after viewing an ad. While many advertisers currently use these attribution windows, many have already moved to 1-day or 7-day post click windows. Combined with the Conversion Time Reporting, this is less of an issue than if you were looking at a full 28 day window. But it still makes it harder to recognize the full value of your paid campaigns.
- User Breakdowns: Facebook will no longer offer user-level breakdowns on ad performance, including age, gender, and region (we’re still waiting to hear if Country will be supported). To maintain visibility into these segment’s performance, advertisers will need to break their ad sets into smaller audiences. For example separate ad sets targeting men and women.
- 8 Conversion Events: Facebook is limiting advertisers to 8 standard or custom conversion events for reporting or optimization. That means events like Purchase, AddToCart, Lead, or any custom events we’ve created. Additional events can be used for building retargeting audiences, but they wont appear as ad conversion metrics.
- Conversion Lift Studies: These will no longer include iOS users. We expect Facebook to use statistical analysis to derive results using the data they continue to collect, but with such a large chunk of users missing, it remains to be seen how accurate or actionable this product will be.
While these changes are a bit scary, there are steps advertisers can take, to minimize their negative impact:
- Implement Facebook’s Conversion API: This server-side API sends Facebook conversion data they can tie back to their users’ ad interactions. This bypasses Apple’s restrictions on what data can be sent directly from the user’s app. The match rate will not be as exact as the current methodology, but it will help a lot. If you’re on Shopify, they recently released a Conversion API integration that works for Purchase events. You’ll want to activate Customer Data Sharing, and set the level to Maximum. For other advertisers, you can contact adMixt. In some cases, Facebook will fund our efforts to help you integrate with this API.
- Verify your Domain with Facebook: Follow these instructions to verify your domain. This is required to manage the 8 conversion events you’re authorizing Facebook to track. If you’re only using standard events, and not using 8 of them, this may not be required. But it’s still a good idea.
- If you have a native mobile app, install the latest Facebook SDK: According to Facebook documentation, only Facebook iOS SDK versions 8.1 or higher will properly send App Event data to Facebook. The latest version of the Facebook SDK can be found on GitHub.
- Pull Breakdown Data Now: It remains to be seen if historical user-level breakdowns will be available after this update goes live. But it’s available now, and advertisers should take advantage of it. Pull breakdowns for Age, Gender, Region, and Country, at the most granular level that will be useful. For example, pull those for your acquisition campaigns separate from your retention campaigns. Save that data to plan your future strategies.
- Set Baseline Attribution Backfill Windows: Some advertisers see 80% of their conversions on day one (the first 24 hours after a user interacts with an ad). Others see less than 50%, with significant volumes of conversions coming in by days 7 and 28. While all these windows are available, document your rates for each stage in your marketing funnel. Upper-funnel acquisition prospecting will typically have longer backfill windows than past purchaser retention campaigns. Establishing your baselines will help you estimate the true value of your campaigns. You can also compare these metrics to last-click data you see in your order management system (like Shopify) or third-party analytics.
While these changes focus on iOS users, it remains to be seen how Facebook will incorporate them alongside Android and web users where more robust data is available. In some cases, like with attribution windows, they’ve decided to implement the lowest-common denominator in their UI, and restrict data to 7-day post click. But with other features there may be different levels of detail for different types of users. We’ll have to wait and see.
As with all changes in technology, some brands will adapt more easily than others. While many of our clients could be effected, there are often ways to lessen the impact so they fare proportionally better than their competition. In an auction-based system like Facebook, sometimes that means costs actually decrease. So every challenge can be an opportunity for profitable growth, as long as we stay ahead of the curve.
If you’d like more information on the coming changes, or how to mitigate their impact, reach out to the adMixt team.
