In February 2021, I tweeted about a theory that talked about objectives/optimizations in lower-funnel campaigns. The concept was that, for retargeting campaigns, you could set either your objective to reach or your optimization to link clicks/impressions to save big money on the high CPMs of purchase optimizations.
Sounds strange, right? Think of it this way: You have already paid the high CPMs of a conversion objective in your top-of-funnel campaigns to get these people to this position in your funnel. They’ve already proven, through their actions, that they are qualified to some degree. They have signalled to you through their engagement with your ad and website that they are of higher purchase intent than a cold audience. Now that they sit lower in your funnel, I propose that you no longer need to pay Meta exorbitant CPMs to re-reach and convert them.
The reason I say this is two-part.
First, using add to cart abandonment, as an example, we can see that these people have already made it low in your funnel. Did they stop because Meta’s algorithm behind what served them an ad wasn’t perfectly tailored to that moment in their life? No, they did engage with the ad and got close to purchasing after all. Instead, something at checkout likely prevented them from purchasing. These people have self-selected themselves as potential buyers. There’s no reason to assume that we need the power of Meta’s algorithm to display an ad that addresses shipping concerns or brand CTAs more efficiently than simply reaching this person one more time.
Second, focusing solely on purchase objectives in retargeting can limit the people you’re reaching. If Meta’s algorithm determines that Person A is of higher purchase intent than Person B, Person A will get the ad first, and Person B will wait until that determination changes. The problem is that Meta’s algorithm is 1) missing a lot of data to make these decisions in a sensitive, person-to-person manner, and 2) can’t pick up on subtleties of UX on your website. For this reason, we may be unintentionally excluding people who have self-selected themselves as potentially qualified buyers from our retargeting campaigns.
So, this is all well and good, but what do we do about it? The concept is simple enough: Change our campaign objective to reach. This way, we can save money on CPMs by using a cheaper objective while hitting the same high-intent users.
When running this type of structure, monitor your frequency. If this gets too high, your savings may be nullified. You can easily scare away potential customers with too many ads at too high a frequency.
Without Meta prioritizing only the ‘low-hanging fruit’, you may have to take some of this into your own hands. Revisiting one of my all-time favourite bottom-of-funnel strategies like breaking out lookback windows might work well with this new objective, though it does come at the cost of consolidation. Breaking out lookback windows, for those wondering, would look something like this:
BOFU Campaign - Reach Objective
- Ad Set 1 - Add To Cart Abandoners, excluding Purchasers - Last 3 Days
- Ad Set 2 - Add To Cart Abandoners, excluding Purchasers - Last 7 Days, excluding Last 3 Days
- Ad Set 3 - Add To Cart Abandoners, excluding Purchasers - Last 14 Days, excluding Last 7 Days
- Ad Set 4 - Add To Cart Abandoners, excluding Purchasers - Last 30 Days, excluding Last 14 Days
- Etc, etc.
The idea is that, if using either caps or ABO, you can control which level of recency is being prioritized since Meta isn’t doing that algorithmically anymore (or at least to the same degree). Users who performed a cart abandonment action in the last three days are generally more likely to convert than those 30 days ago, so putting more budget or a more loose cap on the 3-day audience makes sense. Likewise, it makes sense to put a more restrictive budget or cap on the 30-day audience, mainly because the 30-day audience will be bigger than the 3-day audience.
It’s crucial to consider the outcomes of new strategies - the good and the bad. Using a more upper-funnel objective may work great, but there’s a possibility that it fails. Ponder on how it may fail, and produce a means of addressing that ahead of time. Revisiting a strategy like lookback windows can be a way to get ahead of frequency/budget running wild.
The concept that I’ve written about above is a thought experiment. Every brand is unique and requires a different approach. Take the idea above, think about how you can apply it practically, and revise as needed.
It might be worth doing a follow-up newsletter discussing expectation vs reality on concepts like this. If you end up testing this in one of your accounts, let me know! I’d love to hear about how it performed for you and what learnings you could glean from it. Maybe it performs exceptionally well; perhaps it fails. We will never know if we don’t test it, and that’s the beauty of marketing. The sky is the limit in terms of room for innovative thought. Testing, and especially failing, is the only path towards finding the gold.