In March 2021 we wrote a blog post that looked at view-through conversions versus click-through conversions for a basket of US-based brands. Much has changed since then including:

• iOS14.5 and the ensuing performance marketing armageddon
• FB switched from date-of-click to date-of-conversion attribution logic (for example, a conversion today is always attributed to today, whereas in the past a conversion today could have been attributed to a click date prior)
• 28-day click attribution has disappeared, though I understand it may be making a comeback (again thanks to SKAN)

In that post, our conclusion was as follows:

If view-through conversions are at 40%+ of default conversions you ought to run an incrementality test ASAP.

The basic idea here is that when view-through conversions account for a significant percentage of attributed conversions there is probably a lot of waste. In those cases, one really ought to do some form of incrementality testing (perhaps using a third-party tool like Measured/Rockerbox/INCRMNTAL/Northbeam, etc. or perhaps in a DIY, geographic test form).

Now that attribution has changed on FB I wanted to re-run these numbers so that I had an updated benchmark. For our analysis, we looked at 70, US-focused ad accounts which we loosely categorized as eCommerce (we included some telemedicine brands herein). These accounts spent roughly $12m in total in a 30-day period.

Just like in our first post, I calculated the percentage of attributed purchases using the default attribution window (7-day click + 1-day view) that were attributed as 1-day view-through conversions. This is what that data looks like in aggregate:

On average, 34% of conversions were attributed to the 1-day view window. That feels about right to me.

We did have a number of accounts with significantly higher %s:

Obviously, any one of these brands above that is making decisions based solely on Facebook's default attribution window is probably in for a rude awakening when they run an incrementality test. I've found this is especially true for Stories/Reels/Snap/TikTok formats which drive up view counts tremendously due to their fundamental nature (ie lots of swiping/posts + minimal engagement per post but often >3 seconds and thus counted as a view).

PS: I don't mean to suggest that view-through conversions are implicitly bad or not accretive. In many instances, we've seen view-through conversions be very much validated by incrementality testing.

‍