Meta's "Walled Garden" is coming

Meta is changing. Are you ready to change with it?

To recap what we’ve seen in the last year:

The goal is pretty clear: less manual control for advertisers, more automated decision-making by Meta’s machine learning. Zuckerberg’s vision is to cut out the “middleman”: fully automated Meta advertising, from creative to media buying.

If a brand can just feed money into Meta and get results, then what becomes of us marketers? Let’s assume Meta can actually pull this off.

How do we prepare for a fully-automated Meta Ads platform?

One answer: if the old “levers” we pulled as marketers are disappearing, we find new ones.

Meta’s Custom Attribution has emerged as one such lever.

Meta’s new “Custom Attribution” solution allows you to optimize your Meta ads on performance measured by an external attribution tool. Northbeam Apex is our version of that connection: you feed your Northbeam data back into Meta, and your Meta ad algorithms optimize on Northbeam performance numbers instead of Meta’s numbers.

The early results from Apex are awesome. It’s included at no extra cost for all Northbeam customers, and you can turn it on right now.

But why are we talking about this in the context of Meta fully automating advertising?

Because Apex is a lever you’ll always have, no matter how automated Meta becomes.

Think about it. If we lose all ability to really customize our ads or targeting, then the few remaining opportunities to influence our Meta ads become far more important.

But first, let’s talk about a phrase I say a lot, that I don’t think most marketers say enough: “walled gardens.”

As an advertising platform, Meta is a “walled garden.”

This means that Meta’s ad algorithms (the machine learning that decides which users to show your ads) can only optimize on user behaviors that happen inside Meta.

Imagine that giant walls tower around Meta, and Meta can see everything that happens inside their platform, but they cannot see outside of their platform. Hence, a garden with giant, tall stone walls around it.

Meta cannot see or optimize against datapoints from outside their “garden,” so to speak. The algorithm cannot see over the walls into other gardens (Google, CTV, TikTok, etc.)

Until every waking moment of your customers digital life happens within Meta’s “garden” — watching Meta TV, shopping in Meta Shops, texting in Messenger instead of iMessage — there will be some hours (and touchpoints) in your customers’ days that Meta cannot track.

This means Meta’s ad algorithm has holes that it cannot optimize against. It cannot see the entire customer journey.

Beyond Meta, your customers and prospects are seeing your TV ads, your emails, your search campaigns. But Google surely will not share their data with Meta, as Google is a walled garden as well.

This is not a flaw of Meta. All the ad platforms are walled gardens, isolated from one another, and will remain so unless U.S. anti-trust law changes. This is why all the ad platforms so often emulate one another: they are trying to become the biggest walled garden, tracking the widest variety of user behavior possible, all to support their ad algorithms.

Which leads me to the second problem: any machine learning model is only as good as the data it has.

A basic tenet of data science holds that machine learning models can only be as accurate as the data provided to them.

Or, as many scientists would say: “if you put garbage data in, you get out garbage results.”

Imagine Meta’s ad algorithm is a campfire, fueled by the data that Meta can see in-platform: clicks, engagements, some CAPI data you’ve provided. It’s already the best converting ad algorithm in the world… but what if you could provide even stronger fuel?

Northbeam Apex gives Meta’s ad algorithm more signal, more context, and more data than Meta can track alone. It’s like throwing gasoline on the fire.

Apex can help Meta understand what’s actually happening outside Meta’s walled garden. That means the best converting ad algorithm in the world now has even MORE information.

That’s why most people who turn on Apex instantly see double-digits improvement in Meta ROAS.

As Meta continues to automate their ad platform, they’re taking away the levers we used to pull as marketers. But one thing Meta can’t do easily: expand the reach of their walled garden.

Your customers have long, cross-platform customer journeys. They’re watching TV, seeing billboards, watching TikToks. Without Apex, Meta can’t see that behavior, and can’t optimize against it.

Now imagine if your entire ad process, from creative to targeting, was run by an algorithm that couldn’t see your entire customer journey.

Right now, you’re optimizing for that missing context by making creative that works across channels. You adjust your creative designs and strategy for each medium, so everything feels cohesive and part of one campaign. Meta won’t have the same data to do that.

If Meta hopes to fully automate their advertising, you need to be sure that machine learning model has the most accurate attribution data available. Garbage in, garbage out.

For marketers, the battle then becomes: “How can I get better data to feed back into Meta?”

Lucky for you, that’s as easy as flipping a switch in Northbeam and watching your data flow. (I think it’s three button presses, actually. But still.)

I believe Apex is the future of advertising on Meta.

As we lose manual functionality, the quality of external data you’re feeding into Meta will become the largest differentiator or performance between good versus bad marketing.

It’s kind of unbelievable that we’re offering it at no cost inside all Northbeam subscriptions, but once again, that’s because we believe in our mission: making the world’s most accurate ad attribution data.

⚖️ The FTC and 22 states sued Amazon over a hidden reserve price they say cost advertisers $20 billion. No surprise: it’s about money, not your performance.

🧩 Stop cloning your winning Meta ad and start writing one message per audience. Once again: your creative is the targeting!

📊 Adtaxi says 40% of Americans are actively switching to lower-cost brands. Haha, whoops, recession indicator?!

Love this newsletter? Please forward it to your friends, coworkers, or Target’s entire PR department. Subscribe here to get this data in your inbox every week.