Your guide for growth. Dozens of agents watch every corner of your store, from acquisition to lifecycle, and tell you each morning what is leaking and what to focus on.
How did the store do yesterday?
€14,160 from 120 orders, AOV €118. Meta spend was €980, so a true ROAS of 4.10×. One thing to look at: checkout completion fell to 31.3%.
Checkout completion fell to 31.3%, from 41.2% a week ago. Mobile only, so it is not traffic.
+€16,900/mo if fixedEvery night runNorth goes through your whole store and keeps only what is costing you money. By morning you have the problem, what it is worth in euros, and the campaign behind it. Not a dashboard to read. One thing to do.
Checkout completion fell to 31.3% from 41.2%, putting €16,900/mo of revenue at risk. Klaviyo is connected but no abandonment flow has run in 121 days. Add an express checkout option this week, then switch that flow back on.
Only 31.3% of your 1,180 checkout sessions complete.
Measured over 7 days, against the 41.2% this store ran a week ago.
Mobile only. Desktop completion did not move, so this is not a traffic problem.
€16,900/mo is what the old completion rate would be worth here
9,840 carts left in the last 30 days with no follow-up.
Klaviyo is connected, but the abandonment flow has been off for 121 days.
The traffic behind those carts is already paid for, so this is recovery rather than acquisition.
€14,200/mo is what a working flow recovers at your cart value
Acquisition, funnel, spend, lifecycle, inventory, profit. Each agent owns one question, checks it against your reconciled data, and only raises a card when there is money attached to the answer.
Finds exactly where visitors drop, whether it's traffic quality, product page, cart friction or checkout, and prices the leak in euros rather than percentages.
Scale it, cut it, watch it. Every campaign and ad set gets a decision, not a chart, with the exact creative behind the number. You approve. runNorth never spends on its own.
Detects missing or inactive abandoned-cart, checkout and win-back flows, and estimates what an idle flow is costing you this month.
Tracks burn rate and stock runway per SKU, and blocks a "scale this campaign" recommendation outright if the product behind it is about to sell out.
True COGS, marketing efficiency ratio and net profit after ad spend. The number your accountant cares about, refreshed daily.
Spots the warm audiences you already own and aren't using, the buyers you keep paying to reach again, and the visitors no campaign is following up at all.
Connect runNorth to Claude and your reconciled data answers in plain language. Not an export, not a screenshot you have to explain. Claude queries the live store, runs the same agents, and shows its working.
Which channel actually brings me new customers, not repeat buyers?
Paid Meta is the only channel genuinely acquiring. 68.1% of its orders are first-ever customers, against 32.7% store-wide.
Google is mostly branded search, so it's catching demand you already created. Cutting Meta would cut the top of the funnel for everything else.
The rN columns are runNorth's own attribution, your first-party pixel matched to Shopify orders, sitting next to Meta's self-reported figures so the gap is in front of you rather than hidden behind it.
| Campaign | Spend | Cost / cart | Meta conv. | Meta ROAS | rN conv. | rN ROAS |
|---|
Your first-party pixel matched to Shopify orders. Meta's own figures are left exactly as Meta reports them.
Meta's figure can be wrong in either direction: inflated by generous attribution windows, or blind to revenue its own pixel never saw. runNorth checks it twice, with a first-party pixel that catches the visitors Meta misses, cross-referenced against the orders Shopify actually recorded. Not an estimate. Reconciled.
this week's attributed revenue, self-reported
real orders, real revenue, ground truth
the real overclaim measured on a live store, against Meta's own reported figure
of visitors ever come back, and 85% of those returns are another paid click you bought twice
of 283 checkout abandoners who never returned went on to buy anything, ever
One real person: arrived from an abandoned-cart email, viewed the product, added to cart, started checkout, and left. Meta never saw them. runNorth did, and knows what the email was worth.
Or what drags them down. Pick the day it turned, choose a window either side, and runNorth compares the two periods all the way down the chain, from sessions to profit, then names what actually did it.
The sale worked and it cost you everything it made. Traffic nearly tripled and orders rose 137%, but the discount took 14 points off your margin and Meta spend tripled to hold it there. You finished the fortnight €480 down.
I've been in D2C e-commerce for 10 years, and many of the decisions I made were based on data that was incomplete or simply wrong. When your data is unreliable, it is impossible to make a calculated decision that leads to growth.
That is what led me to build runNorth: the agentic operating system for D2C brands on Shopify. It captures the full customer journey, first click to last, and identifies what drives sales and where revenue is leaking. Dozens of agents work across the funnel, from acquisition to lifecycle, scaling what works and flagging what's broken.
The point is to spend less time figuring out what happened, and more time acting on it.
Built for Shopify brands doing €200K to €5M a year and spending €3K to €30K a month on Meta ads. Running without ads yet? Half the signal disappears, so talk to us first.
The smallest leak in last night's report was €11,300 a month. That is nine years of runNorth, found while you slept, and runNorth then tracks whether the fix actually worked.