7 things I love about Facebook ads right now
We are solidly into the ecom summer slump. If my Instagram Stories are to be believed then the majority of London is now in Greece. And across client teams, it seems about one third of any team is away at any given time.
What happens when all these people go away? Ever since an early boss told a room of us ‘never to bring external factors as an excuse for poor performance’, I’ve held that mantra close to my heart. But at a certain point the data becomes overwhelming. Summer is slow.
Not only that but there’s a lot of chat at the moment – as there often is when times are tough – about CPMs. “CPMs are up” when what people actually mean is “Facebook is expensive.”
Some of our CPMs are up, some are the same, one or two are down. CPAs are a mixture of all three as well – and as is always the case, there’s no statistical correlation across a broad dataset of the two.
But summer is slow. And for those of us in work looking at the numbers every day, it can be disheartening.
But it’s not all doom and gloom. As Ballpoint heads into our first annual offsite in a few weeks I’ve been in a reflective mood, and I’m feeling upbeat about Facebook Ads at the moment.
This also coincides with a fairly major platform update that’s coming. So here’s a mix of things we love and some new stuff on the horizon.
Today’s post is a celebration: 7 things I love about Facebook ads at the moment.
1/ 1-day engaged view
I’m a big fan of viewthrough conversions. If viewthrough conversions weren’t a thing then no brand ever would have spent a penny on TV. All digital has done is is build more data into that process.
But plain old 1-day view has its downsides. For various reasons, Meta often gets viewthrough attributions wrong. We work to a healthy barometer that viewthroughs should be about 10-20% of overall mix, and if it’s much higher than that it’s worth experimenting with setup.
And so 1-day engaged view is the dream.
Engaged view requires 10 seconds of more of a video to be viewed to be considered. There’s no hiding here. But it reveals a much truer view.
Go check your balance. See what current comparative CPAs are, and then run yourself a test to see how a different optimisation approach could work for you.
2/ Cost caps
I’ve always found cost caps to be one of those proposed silver bullets that in practice never deliver. I am far from a Cost Cap absolutist, and I’m probably more skeptic than much else, but we’ve had good success in some areas this year.
In particular they’ve meant summer has been far more efficient than other periods.
The biggest mental shift is moving away from the idea that every month should be a growth month. It’s not. Some months you will have fewer customers in market.
This psychologically becomes tricky if you don’t have the confidence in your ads. If you’re left wondering: ‘is it the time of the year – or is it our marketing?’ then you’ll be switching these off in no time. But if you’ve got high ad confidence, periods like this can be a beneficial time to be in them.
3/ Awareness campaigns for remarketing
Awareness campaigns can be pound-for-pound much more cost effective than conversions ones. Fewer people are bidding for this campaign objective, and so there’s less competition.
Now, you can’t measure this – most of the time – on Meta. You need to measure this off platform with an incrementality test. But we’ve ran five of these over the last year, and four of them have shown the same thing. (Read my post I did on this from earlier this year).
How to measure your own test?
Split your audience off platform. We do this in Weld to auto-assign all users into cohorts, then choose how we send the cohorts back to Meta
Run one cohort on a conversions campaign, another on a reach, and a third as a holdout
Wait 4-12 weeks for enough data
Sum contribution margin 2 per cohort
Subtract the Meta spend from the 2 Meta cohorts
Check conversion rates for statistical significance
Measure CM3 and choose your remarketing solution
4/ Northbeam Apex for data in platform
Northbeam and Meta have been testing out using Northbeam data to feed the Meta algorithm over the last few months for a private beta. The results have been very cool.
For those using Northbeam, they’ll likely have a different view of where acquisitions are coming from to what the platforms report.
Now, attribution is just one part of your measurement framework. I can’t understate that enough. But this report has been put together taking incrementality into account. The uplifts aren’t just a change in reporting, but actually helping the algo find better customers.
Northbeam is much less used in the UK than Triplewhale, which is coming soon, but the report is really promising. Read the whole docs here.
There’s risks here. And the risks are it creates a deeper love of click-based attribution – which we shouldn’t be reliant on. But if you take a holistic view, then having better data at every point is vital.
What’s not promising is using GA4 for this. Don’t use GA4 for any attribution, ever.
5/ Conversion value rules aka bid multipliers
If you have some customer types who are more valuable to you than others, conversion value rules could help.
As an agency, we’re way more focused on Jobs to be Done customer segments than we are demographic ones. But that said, there are segments for most accounts which produce better LTV.
Conversion value rules allow you to bid more/less for different cohorts of people.
Now bid multipliers have been available to many whitelisted brand for a long time via the API, but this should hopefully bring more ability to the masses.
6/ Incremental conversion attribution setting
Another incoming new feature. This would be ranking top of my list if we could see yet who this would be available for. It looks like it’s only going to be for those who have already run conversion lifts, which rules out the majority of advertisers.
But for those who have, this is a new way to bid towards actual incremental conversions. One might argue that this should be the default setting, which hopefully one day may well be.
7/ Breadth of creative types working
Last one, but one that just fills me with joy is the breadth of creative concept types currently working.
A year ago it felt like some creative was often getting much narrower: certain UGC formats were copy and paste, certain static styles too.
But if I look agency-wide right now, we’ve got the broadest selection of creative concept types we’ve ever run and it’s helping across everyone.
This stuff is all audience specific and it requires the starting point to be: what do my target customers usually see in their feeds. We’ve got content that looks like it’s from Real Housewives of Clapton that kills it for one brand – but would be a total fail for others.
Know your audience, stop copying other brands, listen to your customer. Succeed.
August can be a time of frustration in e-commerce and DTC, but it doesn’t need to be. There’s a lot happening on platform to advertisers benefit, and creative is more exciting than ever. As ever, if you’re hitting a ceiling with your growth and wondering: “is it August, or is it actually our marketing?” then we can help you find an answer to that. Drop me an email direct here.