A reconciled, automated P&L for a small brand: sales, every variable cost, and overhead from the books, combined into the real net profit the founder can finally trust, refreshed daily.
The founder did not actually know if the brand was profitable. Sales sat in Shopify, ad spend in two ad accounts, COGS and shipping in invoices, fees buried in payouts, returns somewhere else, and overhead in the accounting software. Once a month someone stitched it into a spreadsheet that was stale before it was finished and probably wrong. Every real decision rode on a number nobody trusted.
01 The Challenge
The pieces of the brand's profit lived in a dozen places: Shopify for sales and refunds, Meta and Google for ad spend, supplier invoices for cost of goods, the 3PL for shipping, the payment processor's payouts for fees, and QuickBooks for overhead. Nobody could see them together. The monthly profit spreadsheet was a heroic manual stitch-up, out of date the day it was done and riddled with guesses. So the founder was making real decisions, spend more here, launch this SKU, on a bottom-line number they did not believe.
Growing revenue means nothing if you can't see the profit underneath it.
02 The Approach
The governing rule: every dollar of revenue carries every cost, COGS, shipping, payment fees, ad spend, returns, and the overhead from the books, before profit is computed, and the result is reconciled against payouts and the bank so it ties out. Refreshed daily, not monthly.
The decision that makes this its own build is right-sizing. A brand this size does not need a data warehouse and an analytics team; that would be over-engineering. The best fit is managed connectors for the standard tools, Postgres for the modest data, dbt for the P&L model, and an off-the-shelf dashboard, delivered lean.
One honest boundary, set up front: this produces a trustworthy, reconciled, automated profit number, not an exact-to-the-penny figure. It is only as precise as the bookkeeping behind it, so reconciliation and clearly stated assumptions are part of the build.
What we deliberately did not do: no enterprise data warehouse, no custom-built dashboard (Metabase does the job), no replacing Shopify or the accounting software, and no machine learning, this is aggregation and reconciliation done well.
03 The Build
Managed connectors (Airbyte) pull Shopify orders and refunds, ad spend from Meta and Google, payment-processor fees, and the books from QuickBooks Online; a small Python connector handles the 3PL shipping-cost feed. Everything lands in PostgreSQL. dbt builds the P&L model: order-level contribution margin (revenue minus COGS, shipping, fees, and returns), rolled up and combined with operating costs from QuickBooks into true net profit, then reconciled against payouts and the bank so the number ties out. Metabase gives the founder a daily P&L and margin broken down by product, channel, and month. It runs on a simple daily schedule.
Deliberately light: no data warehouse, no custom dashboard, and no heavy orchestrator, a small brand does not need them.
04 The Results
For the first time the founder has one reconciled profit number, refreshed daily, instead of a monthly guess. The real net margin came in below what the spreadsheet had implied, and seeing margin by product and channel exposed the leaks: a couple of hero SKUs that actually lost money once shipping and returns were counted, and a channel whose true cost ate its margin. Repricing and cutting those lifted net margin about five points. Decisions now run on a number the founder believes.
05 What's Next
The reconciled P&L is the foundation everything else builds on. As the brand grows, it graduates naturally: SKU-level pricing decisions, then the channel-level contribution-margin engineering and the attribution work that a larger brand needs, all plugging into a profit number that already ties out. Built lean now, it grows with the brand instead of being rebuilt.
A warehouse-native cost model that loads every order with its real variable costs, so CAC, payback, and contribution margin are finally true by channel, campaign, and cohort.
One idempotent pipeline that turns retail purchase orders into shipped, invoiced, reconciled revenue across multiple 3PLs, with humans touching only the exceptions.
Numbers you can trust, operations that run themselves, all built into your stack and owned by you. Tell us where you want the edge, and we'll build it.
You'll hear from an engineer the same working day, not a sales sequence.
A short call, then a plan, and we start with Proof Week or a full build.
A fixed fee agreed before anything starts, for Proof Week or the full build.