Why platform numbers can’t be the source of truth
Meta and Google mark their own homework. They over- or under-attribute because they don’t know: attribution is statistical modelling, not observation, and every platform models in its own favour. GA4 samples data on top of that. Use any one of them as your source of truth and you end up making decisions that look great inside the platform and wrong for the business. So we treat platform numbers as what they are: a useful leading signal for optimising inside each platform, but never the verdict on whether marketing is working. The verdict comes from your own numbers: what left the bank, and what came back.The growth sheet
The core artefact is unglamorous on purpose. One view, often literally a spreadsheet, with a row per day: every new lead or order in, from every channel and every form, and every ad dollar out, from every platform, with traffic, conversion rate, spend, and revenue side by side. Why so simple? Because this is easy to overcomplicate. You can spend months wiring six tools into a CRM and end up with a dashboard nobody trusts anyway. A daily sheet everyone can read beats a beautiful integration nobody believes. When you change a budget, swap creative, or touch a landing page, you see the effect here, not in a platform report arguing for its own commission. It’s also built to outlive us. We use it constantly, but we’ll also encourage you to read it the way we do, because it makes every marketing decision you’ll ever make faster.What we set up, in order
The daily ledger
Leads or orders in, spend out, per day, across all channels, with MER (revenue divided by ad spend) on top. This exists by the end of week one, and nothing else proceeds without it.
Close rate and lag
For lead generation: what percentage of leads become customers, and how long that takes. For ecommerce: how quickly a first order becomes a second, and what a customer is worth over time. Once these are known, the sheet becomes a forecast. Today’s leads or orders are next quarter’s revenue at a known rate, and spend decisions stop being guesses about a lag you can’t see.
Qualified signal back to the platforms
This is the loop most accounts never close: telling Meta and Google which leads or customers were good, through offline conversions, CAPI, and enhanced conversions. Without it, the algorithm optimises for whoever fills in a form, including the people your sales team dreads calling. With it, every month of spend teaches the platform your definition of quality. It’s routinely the highest-leverage fix in the entire audit.
The North Star
One business metric the whole system aims at, whether that’s average order value, average deal size, or qualified leads per rep. Whatever moves your economics most. It’s often not ‘more leads’. Doubling the value of the average customer can justify several times the ad spend at the same lead volume, because your most expensive resource, a broker’s hours or a fitter’s day, suddenly earns more per unit. We agree it up front, instrument it, and point everything at it.
Why this ordering matters so muchEvery downstream decision, from which creative won to whether budget should rise to which channel gets the next dollar, is only as good as this layer. We’ve audited accounts that ran for 18 months on numbers nobody had checked. The first week of an engagement is spent here so no other week gets wasted.
What we need from you
- Access. Read-only for the audit, admin at kickoff, covering analytics, ad platforms, CRM, and forms. The getting started checklist has the full list.
- Honest unit economics. Margin, average order or deal value, close rate if you know it, and what a customer is worth to you. These numbers set every budget line.
- Sales reality. The information your sales side collects, because self-reported form answers are often wrong, and comparing form data against sales truth is where lead quality gets measured. Sometimes the fix turns out to be a sales process change, not an ads change, and we’ll say so.