I hear some version of that line almost every week. On the surface it sounds reasonable. You scale to a comfortable number on Meta's in-platform data. You keep your reporting cheap and simple, and only when you're ready to branch into new channels do you go shopping for real measurement. 

Attribution gets filed under problems for the future, as many folks are loath to pay SaaS subscription fees. Yeah, I see you. But that’s a discussion for another day. 

Many of the businesses that come to us are spending millions a month on ads, so I understand where the assumption comes from. 

But some of our most compelling growth stories started much, much smaller. These advertisers saw the vision and had a plan. They knew they could grow from zero to hundreds of millions, so they put the right foundation in from day one. 

The facts: smaller advertisers cannot scale as easily on in-platform data as they used to. 

Our own reporting shows that advertisers spending under $50K a month saw noticeably worse year-over-year improvement this past May, and that gap has held for most of the year. The K-shaped marketing economy is real. The big spenders on Meta have the budget and team structure to quickly adjust, while smaller teams are struggling.

It isn't hard to see why. Meta is pouring everything into Andromeda, and Mark Zuckerberg keeps talking openly about pulling humans out of the advertising loop entirely. 

My favorite thing to say: starting an ad account has never been easier, but scaling one profitably has never been harder. Acquisition costs are climbing on every channel, not just Meta. 

So I wanted to take this “you don’t need attribution until x spend level” fallacy head on. I sat down with Tim Peacock, who leads SMB and agency partnerships at Northbeam. He works with sub-$50k-month-ad-spend businesses every day.

"SMB businesses need good data too, and if anything, it's probably more important for smaller businesses to have the right attribution in place so they can grow," Peacock said. "Because without that, it's tough to know where to spend your marketing dollars."

For the record, there is no spend minimum to use Northbeam. We have brought on businesses spending as little as $15k a month. Most smaller advertisers look at the price of Northbeam and wonder: is it worth it at my current level of spend?

Here are eleven reasons the answer is almost always yes.

If you've ever built a bootstrapped ad campaign from nothing, you know what it feels like when a test goes sideways. There goes ten grand out the door. You probably know how many units you had to sell to even fund the test in the first place. It hurts. 

When you're spending under $50k a month, there is no room for misallocation. If you screw up, that’s a learning tax you just paid to some ad platform. If you didn’t make a profit, you can’t reinvest next month, and you lost ten grand for nothing.  

"At that level of spend, there's a lot of opportunity for waste," Peacock said. "And of course, the more you scale over time, the easier it is to look back and see you overspent $500 on this campaign."

The problem is structural. These ad platforms only have so much information on your ads, and they will not talk to each other. That means it’s easy for them to give you the wrong information on which ads are working. In-platform numbers say things are going well, but your revenue in Shopify says otherwise. (More on this later.)

In-platform data misclassifies roughly half of your ads every single week. Based on our research, around 17% of the “winners” according to in-platform data are actually underperforming, and about 29% of the losers you're poised to cut are quietly overperforming. 

The ads driving your growth are often not the ones you think. If you're not sure why this happens, paste this into Claude and read the answer slowly:

Explain what walled gardens are in the context of social media advertising platforms. Explain how these ad platforms cannot track how users move across one another in the customer journey, and what that means for my ability to measure my advertising performance.

The short version: being small does not exempt you from the same messy, cross-platform customer journeys the big spenders deal with. You face all of it, just at a smaller dollar figure. 

So ask yourself an honest question. Are you spending more on failed tests than you would on a Northbeam subscription? Probably. And if you had Northbeam, you’d make back the subscription cost ten-fold each month anyways. 

Everyone assumes they can ride UTMs and GA4 for a while. There is a whole market of attribution tools out there that cost pennies. And as consumers ourselves, we fall for the common fallacy of assuming that you can get the same quality commodity for cheaper somewhere else.  

Then you hit a wall, and we see it land in the same spot almost every time, right as businesses climb into that $50K to $100K range. Your experiments stop making sense. You can’t break a ceiling on ad performance. All your growth plateaus.

The reason for this is attribution depth. You simply cannot pull enough accurate attribution data out of UTMs and GA4 to run experiments at any real level of sophistication. Worse, a lot of the cheaper attribution platforms are not even producing independent performance data. They are rolling last-touch platform numbers into a prettier dashboard and handing it back to you. If you want to audit your own attribution, use this. 

Northbeam gives you completely independent ad performance data, isolated and separate from whatever the ad platforms are claiming. That independence is what makes it more comprehensive, more conservative, and a far better benchmark to measure against.

If you are scaling your accounts with the kind of ferocity you see from the 9 Operators crowd, or Andrew Foxwell's collective, or anyone running the Alex Hormozi playbook, you cannot afford to start on a foundation you'll rip out in six months. Build on infrastructure that can carry you to a million dollars a month. This is necessary if you have ambitions of becoming a high-growth, high-experimentation performance machine. 

As with many things in life, in attribution you can only have two of the three: 

(I would argue Northbeam is extremely affordable in the long run, but that’s another post.)

Most of the enterprise businesses that come to us, the ones growing from $20m to $100m in revenue, are chasing marginal gains. They are fighting for percentage points across complicated omnichannel journeys, stitching together connected TV and a dozen touchpoints, and they need first-party data because they've already squeezed everything they can out of basic single-channel attribution.

As a small business, you’re playing a different game. Northbeam can turn you into a mega-scaler quickly.

"You would see lift faster. You would grow faster, because you're not one of these mega-conglomerate businesses looking for another 5% growth," Peacock said. "A lot of our starter businesses can grow 100% year over year, 50% year over year, by using Northbeam."

This is not hypothetical. Plenty of names you'd recognize started small with us. Grüns is the example I point to most. When they came on board they were spending $50k to $60k a month. Two years later they're spending $500k to a million a month, and the business is worth $500 million. Grüns chose to run a Northbeam-powered performance team from day one. Businesses like Ridge, HexClad, and Kizik tell similar stories. 

Straight up: if you want to absolutely smoke your competition, big or small, start with good attribution. 

Another misconception I can tackle: the data does not get worse because your spend is lower. 

A business spending $10k a month runs on the exact same first-party tracking infrastructure, the same pixel, and the same multi-touch attribution models as a business spending $100m a month.

There is no watered-down starter tier. You get enterprise-grade attribution that stays with you for your entire heroic arc, which means you can punch dramatically above your weight. That matters enormously when you're bidding against better-funded, more-established competitors for the same audiences. If they are not using Northbeam and you are, you can play to your strengths much better. Faster experiments, faster iterations, better testing – all the stuff that these big advertising teams can’t do fast, especially without Northbeam. 

"Northbeam is in the game for getting businesses the most quality data set," Peacock said, "so that whether you're spending $10,000 or $100 million a month, you can trust the data in its entirety."

Every ad platform is a walled garden. It can only see what happens inside its own ecosystem, so it optimizes for the conversions it can see and takes credit for the ones it wants. Meta grades Meta. Google grades Google. Every platform is grading its own homework and, surprise, every platform gives itself an A. They need you to spend more money, after all. 

That’s why it’s a big deal when we, as a third-party, come out and demonstrate how certain ad platforms are overperforming. You know it’s real. 

This is not a big-business problem. If anything it's worse for small businesses, because a small team is far more likely to be making major budget calls entirely on what Meta reports, with no independent check, no incrementality testing, and no media mix modeling. 

Think about it: no humans are only using one app all day. Your customers are watching connected TV, scrolling Instagram, getting a postcard in the mail, bouncing over to TikTok, opening an email. Every one of those touchpoints can only really grade itself, because none of these platforms are built to talk to each other or share user data. 

None of them can see the true full customer journey, so they can’t tell you the truth about your ad performance. How are you supposed to run good ad experiments off that?

Northbeam is platform-agnostic by design. Think of us as Switzerland sitting between the ad platforms, tracking the actual customer journey as your prospects move across all of them. 

"We don't care where you're spending your money," Peacock said. "Our goal is just to help better tell the story of what's truly driving your growth."

This one is bigger than it looks. Waiting until you're big enough to care about attribution means rebuilding your entire tracking foundation later, at the exact moment switching costs are highest and a pile of bad habits is already baked in.

We watch the same movie play out constantly. 

Start with Northbeam from day one and none of that happens. Your UTMs, your pixels, and your attribution models are correct from the beginning. The data only gets sharper the longer it collects. 

You are planting trees you'll sit under in five years. If you genuinely want to reach the scale of a Grüns or a Ridge, you have to use the same tools they do, and you want the algorithm learning on clean data for as long as possible. Build the good habits from day one. 

Multi-touch attribution is the core of what we do, and it's the most valuable piece for businesses of every size. But the Northbeam product suite grows right alongside your business.

Once you're in the $100k to $250k a month range, you'll start thinking about incrementality testing, especially as you push into new channels. Northbeam's incrementality is built in and runs off the exact same dataset, so you're not onboarding a separate provider and hauling your data into some new platform to do it. It also means your incrementality is built on real multi-touch data rather than in-platform numbers, which, as we've already established, are not trustworthy for measuring performance. A lot of incrementality vendors you’ve heard of quietly run on platform data. Yours won't.

Cross $500k a month in spend and media mix modeling enters the conversation. That is where you get sharper forecasting, quarter-by-quarter and even weekly budget shifts, and real channel-by-channel planning. Northbeam's MMM+ handles that, and no surprise, it runs on the same first-party data as your attribution. You get all three parts of the measurement pyramid, attribution, incrementality, and MMM, inside one tool as you grow into them.

Testing a new channel is the hardest, most expensive part of scaling, in my opinion. You have to learn a new culture, a new creative format, and a new set of rules, all while spending money you're not yet sure will work. How many of you have had misadventures with TikTok?

Northbeam built partnerships with the major platforms to unlock matched credits that small businesses normally cannot get. As a Northbeam customer, you get access to credits like these:

For a small business testing a new channel, that is a competitive advantage. Stack a few of them and you've effectively subsidized your Northbeam subscription several times over.

A lot of smaller businesses live in high-AOV or heavy-consideration categories. Think expensive home goods, premium wellness, high-ticket apparel, luxury. Your customers do not decide to buy in an afternoon. They take weeks, sometimes months. 

If your product has a long consideration cycle, your entire business is structured in a way that in-platform data literally cannot track.

Northbeam runs an unlimited lookback window, which means it captures the full path to purchase, including touchpoints that happened up to three months before the sale. You can know what ads are driving each stage of your funnel. 

We have businesses that start seeding their Black Friday and Cyber Monday audiences in February. Give yourself an infinite lookback window and you can finally see which ads people actually touched on the way to buying. This means you can run the experiments that grow you instead of guessing in the dark. 

Some of our most successful businesses are running long, complex consideration cycles or high-AOV products that look terrible on a one-day-click basis. If that is your business, this reason alone is worth the call.

This will be familiar. Add up what revenue Meta says it drove, what Google says it drove, and what Klaviyo says it drove, and the total almost always lands at 150% to 200% of what actually hit your Shopify. Every platform is claiming credit for the same sale. None of them are talking to each other, so none of them are doing the math to remove the overlap. I’ve wasted tons of time trying to dedupe this data, personally.

"You're kind of just sitting there like, 'Well, where should we put our next dollar? Should we spend more on Meta? Should we spend more on Klaviyo? What is actually driving revenue here?'" Peacock said. "Because you don't really know. It's 200% of what shows in your Shopify. How do you make good media-buying decisions based off revenue that doesn't exist?"

Northbeam is the only multi-touch attribution platform that has done the engineering to fully deduplicate. Whatever model you use, last touch, clicks only, or clicks plus deterministic views, your revenue in Northbeam matches your Shopify exactly, down to the cent. 

For a small business, this is existential, and Peacock has watched it play out more times than he can count.

"I've seen countless times where businesses scale and then fall off, because none of it makes sense and it doesn't add up to their actual cash in the bank," Peacock said. "So they start making decisions based on an inflated number, not the true number. And they quickly go down the wrong path."

Everyone can find a hundred reasons their Meta ads aren't working I am not here to argue about which channels are good, which tactics are dead, or whose creative framework wins. 

But here is the choice in front of you. If you want to follow the path of the hyperscaling names in this space, why wouldn’t you use the same tools as them? Even a business spending $20k a month with a goal of hitting $100k by year end will get there faster with Northbeam.

The value of Northbeam for a small business was never just accurate attribution. It’s the infrastructure, the habits you build, and the historical data. This saves you money while letting you reinvest in good, statistically significant creative tests. Do this right, and Northbeam won’t just pay for itself: it will teach you how to scale your business. 

So the real question. Do you actually want to scale? If so, let’s get on the phone. 

🚨 Meta went dark June 12: 100k+ outage reports and Ads Manager disruptions, with advertisers told to pause. Another day another dollar amirite?

🎬 Instagram opened Reels post-view ads to every advertiser worldwide. They run after Reels over 60 seconds, with a 5-second countdown and skip.

📺 Snapchat and dentsu's "Brand Reset" study found one Snapchat Commercial exposure drove +4.4% more sales over three months. Is Snapchat back? Take the Snapchat Commercials, leave the AR glasses.

🔎 SparkToro says 68% of US Google searches now end without a click, up from 60% in 2024. Man, just when I was starting to figure out paid search.

🤖 OpenAI is testing multi-advertiser ad units in ChatGPT, sold via second-price auction. Prediction: nobody likes it.

Love this newsletter? Please forward it to your friends, coworkers, or Evan Spiegel. Subscribe here to get this data in your inbox every week.