Skip to content
← The intelligence layer

A short list of what to do next, ranked by what it is worth.

Recommendations are drafted actions waiting for a decision. Each one carries the measured difference it rests on, the spend it affects, the expected effect, a confidence grade and the constitution rule it clears. Ranking is by euros at stake against your objective rather than by percentage change, so a 3% inefficiency on €19,800 outranks a 40% swing on €400. Holding steady is a recommendation too, and it is written down when it is the right call.

In practice

Demo data

What should I do first this week?

Start with the Google value signal, not the budget move. Google recorded 41 leads last week and the CRM qualified 3, while the conversion signal sent back to Google still counts all 41 equally — that is €17,200 a month optimising toward the wrong outcome. Evidence: CRM opportunity stages over the 14-day window and the conversion feed configuration on the account. Business impact: every day it stays unchanged, Google's bid strategy keeps buying the cheapest form fill. Confidence is high on the misconfiguration and medium on the size of the effect, since one week of 41 leads is a small sample. An alternative explanation is a qualification backlog in the CRM rather than worse traffic; the queue was checked and is clear. Recommended order is ACT-2294, then the €1,200/day move in ACT-2291, then the TikTok creative replacement — and hold LinkedIn, which needs no action at all. Constitution check: ACT-2294 clears the 30 qualified-outcome threshold at 34, and the creative replacement waits for approval because new creative always does.

Evidence thresholds gate the queue

A recommendation does not surface until the measurement behind it clears the minimum in your constitution — 30 conversions before scaling, in this account. Below that it appears as a watch item, because a budget moved on nine conversions never leaves the learning phase.

Holds are recommended explicitly

A channel can cost several times the account average per lead and still be the one to leave alone. Where your CRM records its opportunity rate well above the blended rate, the recommendation is to hold its budget steady and say so in writing. The channel only looks expensive on the metric it is worst at.

An exploration reserve is standing

Twenty per cent of budget is reserved for new channels, concepts, audiences, offers, markets and landing pages, and recommendations compete inside that reserve separately. Exploration is allocated first, not from whatever is left at month end.

Demo data

Behind one answer

Answer
Meta. Its customers are credited with €14,800 more collected 90-day revenue than Google's on 15% less spend.
Evidence
311 customers, 90-day window, invoiced value from the finance system, last-non-direct attribution.
Business impact
Moving €1,200/day would add an estimated 31 customers this month.
Confidence
High — 14 days above the minimum evidence threshold.
Alternative explanation
Meta's audience skews to annual plans. Some of the gap is plan mix rather than channel quality.
Recommended action
Move €1,200/day from Google Prospecting to Meta Prospecting.
Constitution check
Passes — within the €1,500/day movement limit.

One goal. Every channel. Real business truth.

Connect your ad accounts, say what a good customer is worth, and let the Growth Brain work the objective inside limits you set.

No ad spend commitment · Connect one channel to start · [email protected]