Four Tools, Four Numbers

Every week someone at the agency rebuilds the same report for every client. Export Meta, export GA4, check Shopify, paste it all into slides, then explain why none of it matches.

Why reporting eats the week

It takes hours per client. In a Databox survey, 58% of respondents said preparing for a client reporting meeting takes about four hours per client, and 77% pull data from five different tools.¹

The numbers never match. Ads Manager and GA4 "will never match up," mostly because GA4 gives no credit to someone who saw an ad and bought without clicking.² Shopify uses its own default, last non-direct click.³

The rules moved this year. From 12 January 2026, Meta's API stopped returning 7-day and 28-day view attribution.⁴ Then Meta narrowed click-through to link clicks only and moved likes, comments and shares into a separate one-day window.⁵ Month-on-month numbers can shift with nothing real changing.

Clients ask about creatives. Ads Manager reports by campaign and ad set. Which creative is working takes manual regrouping, every time.

What we built

• One nightly pull. Meta, Google Ads, GA4 and Shopify into one database.
• Attribution pinned. Every pull names its attribution settings, so a change on Meta's side can't quietly change the numbers.
• The gap, measured. Meta's reported revenue is reconciled against Shopify orders. The gap shows as a ratio and gets flagged when it moves.
• Read by creative. Results are grouped by creative using ad names and post IDs, so one video in six ad sets is one row.
• A summary in plain words. What changed, the top and bottom three creatives, what's fatiguing and what to test next. The model writes only from numbers the system computed.
• Delivered where clients look. The client's Slack channel and a shared sheet, weekly and monthly.

What it can't do

• It explains the gap. It doesn't close it. No two sources will ever agree.
• Clean names matter. Grouping by creative works best alongside our launcher, which writes every name the same way.