The platform for machine-learning powered advertising attribution.
Joined March 2021
- Tweets2.4K
- Following481
- Followers5.7K
- Likes1.8K
Pinned Tweet
Need some Black Friday Cyber Monday inspiration?
All recordings from our 2026 Conference are on YouTube 🗽
youtube.com/playlist?list=PL…
Clickthrough rates on Meta Ads are up more than 40% year-over-year since January.
Meta is pushing toward a high-engagement future.
This is ALL placements on Meta Ads (Facebook, WhatsApp, Instagram) across all 1,000+ ad accounts we track, around $6 billion in ad spend.
What we can see is that average clickthrough rates have continued increasing nonstop since January 2025.
CTRs cooled off around Prime Day 2026 but have started improving again.
What's going on here, and why does this matter?
A few hypotheses:
1. Since iOS 14.5, Meta has realized that Apple's goal is to restrict as much tracking as possible. iOS 27 continues this trend.
Therefore, Meta is shifting their algorithm, product development, and strategic goals towards engagement.
It will be difficult for Apple to restrict or obscure actual clicks, meaning Meta can retain a high-intent (in theory) signal.
2. Advertisers, especially the folks seeing this message, have been pushing Meta to provide new customer acquisition more accurately and cheaply.
Driving up clicks gives Meta more data.
3. GEM, Lattice and Andromeda need massive amount of signal to work properly. If you know anything about data science you know that more data = more accuracy in machine learning.
In-app browser behavior via clickthroughs gives these models a ton of signal, so Meta's optimizing for it.
4. CTRs are, to some degree, a vanity metric. They're useful for measuring hook performance, but we all know that having good hooks is a bragging right. Nobody wants to talk about what's happening AFTER the clickthrough.
So nobody's sounding the alarm on this because we over-index on CTRs and we're just accepting the win.
5. Meta's goal is to build a userbase that engages heavily with ads. Users who engage with ads are more likely to buy from them.
At a time when ads are deeply despised (on places like YouTube, for example) Meta's making sure that users are getting served ads they actually want to click.
6. If Meta can build a userbase that engages with ads, then when Meta pivots fully to their AI-generated in-house ads creation engines, they'll have a strong sense of what users WANT to see, and therefore those ads will perform as well or BETTER than human-made content.
This is the holy grail for them. Tin hat conspiracy time: Meta is using all your ads to train their engine so they can replace you with AI in the next 18 months.
7. We are a hair away from ads appearing in DMs. It's the final frontier. Meta is gonna do it. It's a matter of when. If this CTR behavior begins to flag after they pivot to AI ad creation, I bet it starts.
What are you seeing in your ads? Are CTRs improving over time?
All data, charts and analysis by @bryanbumgardner 🛰️
That's a real screenshot from a @northbeam dashboard, by the way.
Are you measuring "conversion lag?"
If you aren't you're probably leaving money on the table this BFCM. @BryanBumgardner explains, guide in the thread
This week's Media Buyer Index is in:
It feels like Black Friday came early this year.
This week's data from @northbeam:
Nothing is moving like it's supposed to. This data is from right before Election Day, so post-political spending release isn't visible yet.
Still, Meta conversion rates are dropping even while CTR is improving?
This feels like people researching but holding off on buying until they see a BFCM discount.
Why don't you just give them a discount, then? Why do we have to wait until after Thanksgiving?
If Amazon can do Big Deal Days on a random weekend in October, you can start giving away your BFCM discount now, who cares?
AppLovin performance has been chaotic since Oct 1, but I'm not really concerned about that. Usually BFCM on AppLovin is pretty decent because it's a more captive audience.
TikTok is going wild, though. More people are on TikTok Shops than ever, and since BFCM is truly an Amazon holiday, TikTok benefits.
The TikTok -> Amazon customer purchasing pattern is biblically powerful. Except this year, it feels like more mainstream advertisers are catching up.
I expect TikTok performance to continue to improve through the end of this year into next.
TikTok has my attention. It's capturing more share, Shops is taking off, and brands are starting to take it more seriously than ever. Especially as a source of high-quality creative and content.
Next Thursday's charts are gonna look nuts.
Legend of ecomm @JasonPanzer @9operators hopped on a quick dish with @MAustinHarrison about acquiring net new customers.
Reposting this because BFCM is coming and we need a reality check.
Timeless lessons in here:
This was a great mastermind 💪 catch us in LA next time
Ecomm operators: there's a new paradigm for Black Friday Cyber Monday. 👀
Not everybody will survive it.
I hopped out to LA for a quick panel with @1800DTC repping @northbeam.
Here's what we talked about:
1. Profitability is the most important thing, period.
Nobody is hyperscaling at a loss anymore. You must be first-order profitable to really compete.
2. Maximize margins without sacrificing product quality.
Healthy margins give you enough wiggle room to scale. Most businesses have such challenging cash positions that they can’t really move budget up or down on their paid ads during Black Friday.
3. Awareness channels matter most.
It’s too late to seed any real brand discovery this year. The smartest operators start thinking about Black Friday in the summertime and begin their demand generation then.
4. Black Friday is a demand capture event.
This builds on the previous point. If you’re scrambling to tell the story of your product over Black Friday weekend, you’re going to have a bad time. Customers are super distracted over this weekend, you aren't going to be able to really teach them about your brand them in such a compressed space.
5. Don’t sleep on smaller advertising channels.
If your measurement is solid, meaning you use a third-party measurement tool instead of in-platform data, you can properly scale channels like TikTok and YouTube ads. Plenty of businesses are finding arbitrage in those white spaces where most advertisers struggle.
Northbeam retweeted
This is the kind of niche stuff I learn working with people at @northbeam.
If you're turning off "underperforming" ads before they provide their full ROI, you're leaving money on the table.
That's the power of conversion lag.
Essentially, there is an average length of time it takes someone to purchase after they first discover your product.
If you turn off ads BEFORE that length of time has fully passed, you're losing signal on which ads are actually working.
For example: if it takes 14 days for someone to purchase after seeing your ads, but you're evaluating ads on a seven day performance window, then you're making choices with only half the signal.
That's conversion lag.
It's a specific and specialized concept, but conversion lag can add up. I've seen brands lost >$500,000 in revenue over Black Friday by cutting ads too early.
Northbeam retweeted
Hot Take from @connorrolain of @hexclad :
The best marketing teams know when to push the gas on creative volume and when to ease off.
Early on, volume is your friend. And once you find a winner, you need to lean into it.
But as you mature, iterating on the same winning concept hits diminishing returns. Instead, you have to try new angles to attract new audiences.
And once that new audience is in your funnel, your tried-and-true ads start working again.
Connor is one of the best in the world at this 👇
Most ad platforms are designed to show your ads to the cheapest and mostly likely person to convert.
Usually, that's existing customers. But we want to target new customers, so what can you do?
Match lists, CAPI, all these things still can't get match rates high enough.
Ads KEEP showing to already-existing customers.
You run incrementality tests, and they work, but you're still waiting weeks for results.
The shorter answer: you need to give your ad platforms (like Meta or TikTok) better information about your customers.
If you had that, you'd be able to relentlessly target exclusively new customers.
ROAS would skyrocket alongside efficiency.
You need Northbeam Apex.
Northbeam retweeted
Yappin about @northbeam Apex with @andrewfoxwell and a room full of people way smarter than me
Marketers: You are probably going to overspend on Black Friday.
Overspending on ads, I mean -- not on your own personal shopping.
We can't help you with the latter.
But @couuor, @codyplof, and @MAustinHarrison are here to help you with the former.
@M__Operators strikes gold again
These guys use @northbeam, by the way 🤓
Great advice from @M__Operators legends @codyplof and @couuor.
It's been a pleasure watching you guys thrive over multiple BFCMs now.
Here's to dozens more 🥂
Last week's performance trends, according to our data.
As political spending moves around, so too does the chaos across ad platforms.
We are in the final week before spending spikes. Typically we see ad spend skyrocket on November 1 in preparation for Black Friday Cyber Monday.
But we also just had Prime Big Deal Days, so folks are recovering from that - wasted spend there caused the dramatic week over week improvements in performance on "All Other Platforms" which is about 30% Amazon.
Political spending is important, but not as much as you think. The @nytimes reports that of the billions spent, most falls in Texas, Ohio and Michigan.
Don't expect your CPMs to improve TOO much after election day.
What are you seeing in your data?
To get these charts in your inbox every Thursday, subscribe here: themediabuyer.northbeam.io/
