Skip to main content
FHAVI

Free resource

Reconcile the order book against the ad platforms

Store says 41 orders. Meta says 33 purchases. Google says 29 conversions. One of those gaps is normal and the others are not — and you cannot tell which from any single dashboard. This sheet separates them by day, because most gaps are concentrated on a handful of dates and not spread evenly.

Setup

Build the sheet before you draw conclusions

Use one row per day, export orders by creation date, and export platform conversions by the same date with the same attribution setting you actually report on. Mismatched date bases are the most common reason two exports 'disagree' when nothing is wrong.

01

Columns

  1. Date.
  2. Orders in the store (all statuses) and orders excluding cancellations and refunds — reporting these as one number hides the difference.
  3. Revenue actually charged, in the store currency.
  4. Meta purchases and Meta purchase value.
  5. Google conversions and conversion value.
  6. Gap, store minus each platform, as a count and a percentage.
  7. Test orders, cancelled orders and manual orders for the day — the legitimate explanations.
  8. Notes on anything that changed: campaign launch, discount code, checkout change, pixel deploy.

Diagnose

Four gaps, four different causes

02

Read the shape of the gap, not just its size

  1. Consistent percentage every day — usually attribution. Cross-device, view-through, or a channel with no tracking. Not a bug, and not fixable in code.
  2. A step change on one date that never recovers — a deploy, a checkout change or a tracking change on that date. This is the case the sheet exists for.
  3. Spikes on days with promotions — coupon-driven orders attributed to 'direct' or 'email'. Compare against campaign spend before assuming under-reporting.
  4. Platform higher than the store — the reverse gap. Duplicate events, a test order, or a purchase counted twice because browser and server fire with different event ids.

Act

What to do with each verdict

03

Then close the loop

  1. Attribution gap: accept it, and report on store revenue for decisions rather than platform revenue.
  2. Step change: bisect by date against your deploy log, then fix forward — do not 'recheck' the same broken state.
  3. Reverse gap: open the browser and server records for one affected order and compare event ids before touching anything.
  4. Consistent small gap only on mobile: check the pixel loads before the checkout redirect on that path.
  5. Re-run the sheet weekly. A gap you cannot explain in one line is a gap you will still have in three months.

Keep going

The rest of the library

Every checklist here is free and ungated. The research pages carry the reasoning behind them.

All resources · Research library

The gap has a date on it?

Send the two daily exports and your deploy log for the same week. We will find the date the number changed and what changed with it.

Prefer email? Write to md@fhavi.com