What I’m hearing
“You can't button push your way to better performance anymore. Forget targeting and settings, it’s all about the ad creative”.
This has been a common sentiment in the industry for years now. A narrative started by Meta and picked up quickly by agencies who used it as a convenient excuse to stop optimising their accounts.
Whilst I do agree that good creative is a necessary ingredient for growth, the statement "you can't button push your way to better performance anymore" is at best naïve and at worst lazy. It assumes an account is structurally perfect - that every setting has been considered, every test has been run. I've audited hundreds of e-commerce ad accounts and I can tell you that's rarely the case.
What I’m seeing
Meta's algorithm is engineered to be greedy and report as much revenue as possible. If we aren't being smart about targeting and settings, who wins? Meta - because they can serve ads to people who were already going to buy and claim the revenue. And also, Meta - because they're the ones who profit when advertisers increase budgets off the back of overinflated platform numbers.
So whilst the industry has decided that button pushing is over, I'm here to say there's still plenty of buttons worth pushing. Here's my list.
What I’m doing instead
(Buckle up, your meta rep would hate this part…)
#1 Excluding email subscribers and active site visitors from targeting.
Why? These people are already engaged with your brand through other channels. Don't hand Meta budget to cannibalise existing traffic and claim the revenue for themselves. The platform numbers may not look as strong, but your CAC should improve as you reduce spend on low-incrementality users and re-allocate it to new users.
Exceptions? If the purchase decision is long, non-urgent, or low-trust, you may need to remarket to site visitors to nurture them to purchase. If so, isolate remarketing into a separate campaign and run a conversion lift study to get a read on incrementality.
#2 Optimising to and reporting on click-through revenue only.
Why? View-through revenue can really exaggerate Meta's performance, particularly when remarketing audiences are in play. My stance is simple - you can't follow the view-to-sale journey in Google Analytics, but you can follow the click-to-sale journey. So put your dollars behind journeys you can actually track. Click-through data is more trustworthy and will lead to better budget allocation decisions.
Exceptions? New brands with small audiences don't need to worry about this. It matters most for established brands with large volumes of active site visitors.
#3 Manual placements over automatic placements every time.
Why? Lower quality placements like Audience Network send really low quality traffic to your site - usually low-intent clickers, accidental clicks, and bots. You're also at risk of automatically serving ads in new placements that Meta adds to the mix (Threads anyone?) - and who wants that.
Not a true exception, but an idea - if your agency or rep has sold you on Automatic Placements, ask them to tag placements in your UTM tracking. That way you can slice the data in Google Analytics and check the bounce rates yourself.
#4 Conversion objectives drive better traffic than Traffic objectives do.
Why? In every test I've run, a Conversion objective with a product page view optimisation event outperforms link clicks or landing page views for both session quality and volume. If you want to grow mid-funnel audiences with people who might actually come back and buy, Conversion objectives are the only way to go.
Exceptions? None. Traffic objectives are just not it.
And perhaps my most unpopular opinion #5 Don't bother with brand awareness ads on Meta.
In the last 5 years, brands have been sold on the need for "full funnel" campaign strategies to grow online revenue. I've overseen countless causal impact studies and funnel A/B tests trying to prove the value of upper funnel campaigns - and let me tell you, very rarely have they resulted in an uplift in revenue or new customer growth, even over long time periods.
There are many reasons those experiments fail (a topic for another newsletter), but my perspective is simple - that money is better spent getting people to your site.
And if brand awareness is genuinely a big business focus, go where attention spans are longer and ad formats drive stronger recall: influencers, BVOD, host-read podcasts, and - dare I say it - traditional media (in a newsletter about digital marketing, shocking!)
The button I’ll be pushing next
Incremental vs Standard attribution models. I’ll be running this as a conversion lift study to find out if this actually does what it says on the label.
Final take
I get it - button pushing isn't hot right now. But intentional account structure, smart targeting, and settings selection still matter. A lot.
Do you know what actually pushes my buttons? Brands paying performance marketers for strategy and advice, only to get "that creative didn't work" on repeat - with no actionable insight to show for it. We can do better than that.
Disagree? You'd better push the reply button and let me know.
Cheers,
Sarah Arvela Webb
Brought to you by shower thoughts, a glass of red and 10 years on the tools.