by Trivas.aiYou open Northbeam. You open Shopify. Same date range, same store, same everything. The revenue numbers don't match.
Most people assume they did something wrong. A filter's off, a channel's excluded, someone forgot to sync a date range. But this is why Northbeam shows different numbers than Shopify almost every single time you compare them: it's not a bug, it's a structural difference in how each platform counts a sale.
Shopify records orders as they happen inside its own system. Someone checks out, payment gets captured, an order ID gets created. Done. Northbeam does something completely different: it models attribution based on ad platform signals and its own tracking pixel, trying to answer a harder question than "did a sale happen." It's asking "what caused this sale."
Those are two different jobs. One is bookkeeping. The other is a guess, an educated one, but a guess.
The rest of this post walks through exactly where that gap comes from, how big it usually is before you should worry, and which number you should actually be reporting to your team.
Shopify's number is a ledger. An order was placed, payment was captured, revenue got recognized at checkout. There's no modeling involved, no judgment call. It happened or it didn't.
Northbeam's number is a model output. It assigns credit to marketing touchpoints using its own attribution logic, usually some blend of last-touch, multi-touch, or a proprietary MTA model layered on top of pixel data. That's a meaningfully different thing to measure, and it means Northbeam can only attribute revenue it can actually track.
Here's where it gets messy: a returning customer who comes back through dark social, types your URL directly into their browser, or clicks through from an organic search result doesn't leave the same trail an ad click does. Northbeam either has to guess which channel gets credit, or it drops that order from attributed revenue entirely. Shopify has no such problem. It just sees an order.
So the gap you're staring at isn't really a Shopify-vs-Northbeam gap. It's an attribution gap. Shopify shows you what happened. Northbeam shows you what it thinks caused it. Those will never be identical numbers, and expecting them to be is the first mistake most teams make when they start comparing dashboards.
Once you accept the gap is structural, it helps to know exactly where it's coming from. Five causes show up over and over:
None of these individually explain a massive gap. Stacked together across a month of orders, they add up fast.
A rough sanity check: a 5-15% variance between Shopify's total revenue and Northbeam's total attributed revenue is normal for most DTC brands. That's the cost of doing attribution at all.
Anything above 20-25% usually isn't a modeling quirk anymore. It's a tracking or integration problem, and it's worth treating it that way instead of shrugging it off as "just how attribution works."
To isolate it properly, don't compare Shopify's total against Northbeam's paid-channel-only number, that's comparing two different things and will always look worse than it is. Compare Shopify's total store revenue against Northbeam's total attributed revenue across all channels for the exact same window. That's the apples-to-apples check.
If the gap is still oversized after that comparison, check two things first: whether the Northbeam pixel is actually installed and firing correctly across every page, and whether your attribution window setting matches what you think it's set to. Those two misconfigurations account for most of the "why is this gap so huge" cases teams run into.
Shopify is the source of truth for total revenue, cash collected, and order counts. Use it for financial reporting, use it for inventory planning, use it any time someone asks "how much did we actually sell."
Northbeam, and this goes for any attribution tool, is directional. It's built for relative comparison: is Meta outperforming Google this month, is this campaign more efficient than that one. It was never built to be your absolute revenue number, and treating it that way is where teams get burned.
The practical rule: never hand Northbeam's revenue figure to finance or leadership as if it's the store total. Reconcile it against Shopify first, every time.
And this isn't a Northbeam-specific problem. The same gap shows up with Triple Whale, Polar, and basically every other attribution tool on the market, because the root cause, modeled data versus recorded data, is the same across the whole category. If you're weighing Northbeam against alternatives, it's worth seeing how each one handles this exact tradeoff, which is something we break down in our comparison of Northbeam, Polar, and Trivas.
You won't get Shopify and Northbeam to match perfectly. But you can shrink the unexplained portion of the gap down to something predictable.
Start with the pixel install. Confirm it fires on every page, including checkout and thank-you pages, not just the homepage and product pages. A surprising number of "mystery gaps" trace back to a pixel that silently stopped firing on one template months ago.
Next, align attribution windows across every tool you're using, where the platform allows it. If Northbeam is set to 7-day click and your ad platforms are reporting on a different window, you're comparing numbers that were never going to match.
Then build a single dashboard that pulls raw Shopify order data alongside your ad platform and attribution data, on a shared timezone and shared date logic. When the reconciliation is visible in one place instead of buried across two browser tabs, the gap stops feeling like a mystery and starts being a known, explainable number.
This is the layer Trivas is built around. Our BI reporting pulls Shopify, Meta, Google, and GA4 data into a single Redshift-based warehouse, so instead of guessing why the numbers don't match, you can actually see the line items causing the divergence. For Shopify stores specifically, our Shopify integration handles the order-level sync so the "source of truth" side of the comparison is never in question.
Some gap between Shopify and Northbeam is expected. That's not a failure state, it's just what happens when a ledger and a model try to describe the same event. The goal was never a perfect match. It's understanding the gap well enough to explain it in one sentence instead of shrugging at it in a meeting.
Trivas sits on top of your Shopify and ad platform data to give founders and growth teams one reconciled view, instead of two dashboards quietly disagreeing with each other. If you're already comparing attribution tools, it's worth looking at how Northbeam, Polar, and Trivas actually differ on this exact problem. And if you'd rather just see it on your own store data than read about it, starting a trial is the fastest way to find out where your own gap is coming from.
explore more insights
3 min read
3 min read
3 min read
.d53b12e5.png&w=3840&q=75 "Trivas.ai")