Optimisation
Recommendations and optimisation workflows
Every recommendation leads with the metric, the comparison period, the evidence, and the trade-offs.
Overview
Northwind Retail — Meta Ads · USD · UTC
| Point 1 | 38 USD |
|---|---|
| Point 2 | 41 USD |
| Point 3 | 39 USD |
| Point 4 | 44 USD |
| Point 5 | 46 USD |
| Point 6 | 45 USD |
| Point 7 | 48 USD |
| Point 1 | 52 purchases |
|---|---|
| Point 2 | 50 purchases |
| Point 3 | 51 purchases |
| Point 4 | 47 purchases |
| Point 5 | 45 purchases |
| Point 6 | 46 purchases |
| Point 7 | 44 purchases |
| Point 1 | 20 per purchase |
|---|---|
| Point 2 | 21 per purchase |
| Point 3 | 22 per purchase |
| Point 4 | 23 per purchase |
| Point 5 | 25 per purchase |
| Point 6 | 25 per purchase |
| Point 7 | 26 per purchase |
| Point 1 | 40 attributed revenue |
|---|---|
| Point 2 | 39 attributed revenue |
| Point 3 | 38 attributed revenue |
| Point 4 | 36 attributed revenue |
| Point 5 | 35 attributed revenue |
| Point 6 | 35 attributed revenue |
| Point 7 | 34 attributed revenue |
Cost per purchase
vs previous 7 days| Point 1 | 20 USD |
|---|---|
| Point 2 | 21 USD |
| Point 3 | 22 USD |
| Point 4 | 23 USD |
| Point 5 | 25 USD |
| Point 6 | 25 USD |
| Point 7 | 26 USD |
Spend distribution
Needs attention
- CriticalProspecting CPA rose 24% over 7 days
- Data quality1 ad account delivered no data after the last sync window
Recommended next step
Creative fatigue in broad prospecting ad sets
Current
“Bundle offer v2” active in 3 ad sets, $12,480 of 7-day spend
Proposed
Pause “Bundle offer v2” in the two broad ad sets and reallocate 15% of their daily budget to retargeting
Figures are illustrative sample data. Nothing here is presented as a guaranteed result, and no change is executed without review.
How it works
Input → analysis → output
The same sequence runs on every surface, so a conclusion can always be traced back to its inputs.
Input
A detected change or opportunity
Analysis
Evidence, ranked factors, and a proposed action with trade-offs
Output
An approved, tracked change with an outcome measurement plan
Scenario. A media buyer reviews a proposal to reallocate budget, checks the assumptions and reversibility, and approves it — or dismisses it with a reason.
What you can do
Capabilities and limits, stated plainly
We list what is shipped and what is not, so expectations match the product.
Supported
- Evidence panel with ranked contributing factors
- Current vs proposed comparison with estimated effect and assumptions
- Explicit approval step before any spend-affecting change
- Execution status driven by the platform response, plus an audit trail
Limitations
- Estimated impact is shown only when it can be calculated, and is directional
- Partial failures and stale data are surfaced rather than hidden
Proof
From a change in the data to a reviewable action
- 1
Select an account and date range
One account, one comparison window. Every number below stays anchored to the same selection.
Northwind Retail — Meta Ads · Last 7 days
- 2
Identify a material change
CPA moved well beyond normal week-over-week variation, so it is surfaced instead of waiting to be noticed.
CPA +11.2% vs previous 7 days
- 3
Inspect the supporting metrics
Spend, conversions, ad set mix, and creative performance are shown together with the affected scope.
3 ad sets · 1 creative · 7-day window
- 4
Review a suggested action
A specific proposal with evidence, assumptions, risk, and reversibility — ready to approve or dismiss.
Pause creative · reallocate 15% budget
Make your next advertising decision with better evidence.
Request access and connect an advertising account to see the workspace with your own data.