Ask Sellforte MCP: Which Meta campaigns should we change first?
Welcome to the Sellforte MCP Use Case Library. Each post takes one question that marketing teams really ask their AI assistant through Sellforte MCP, which is now in general availability and accessible to all marketers worldwide. It strips out anything customer-specific and walks through what the assistant does with it, step by step. We run every question on our demo account so the numbers you see are illustrative, not anyone's real results. The idea is simple: steal the prompts, try them on your own model, and see what you get back. We hope you find this series inspiring!
-Juha, CEO at Sellforte
Turning a page of Meta recommendations into a plan
You open Sellforte's Performance Recommendations, and there's a row for every Meta campaign and ad set: a new daily budget here, a cut there. Useful, but it doesn't say where to start or why. Performance marketers who run several Meta campaigns across markets hit this every week, usually with a second question close behind: does this still hold if my real goal is new customers?
What you ask
Summarize the current recommendations for Meta Ads: how many campaigns to increase or decrease, the biggest budget changes, and which to act on first. Explain what the recommendations are based on.
Can we focus these on new customers? Show me new-customer return per Meta channel from the MMM for the same window, and tell me where it agrees with the recommendations and where it doesn't.
What Sellforte MCP does
Reads the recommendation snapshot
The assistant pulls the recommendations dataset filtered to Meta Ads. It's a snapshot over a fixed window, and the answer names it: 28 days, 2026-10-05 to 2026-11-01.
On the demo account there are 10 Meta campaigns and ad sets with a recommendation: 5 to increase, 3 to decrease and 2 to hold. The total daily spend stays the same at $6 891. The engine's own estimate of the net effect is about $570 more incremental sales a day, just from moving money around inside Meta.
Here are the biggest moves, in weekly terms (illustrative demo data, USD):
| Campaign / ad set | Action | Weekly spend now → after | miROAS | Weekly incremental sales impact |
|---|---|---|---|---|
| Advantage+ DE (broad) | Increase 8.9% | 6 830 → 7 435 | 3.79 | +2 845 |
| Prospecting US | Increase 12.3% | 2 848 → 3 199 | 4.69 | +1 831 |
| Advantage+ US | Increase 3.3% | 9 016 → 9 314 | 4.19 | +1 492 |
| Retargeting US (30-day) | Decrease 32.3% | 1 404 → 951 | 1.92 | −942 |
| Retargeting DE (30-day) | Decrease 32.3% | 3 190 → 2 161 | 2.02 | −2 233 |
The order it suggests is simple: do the big scale-ups first, where the expected gain is largest, and fund them by trimming the two retargeting ad sets, which have the lowest marginal return on the list. For each row you also get the exact setting to change, for example "set the daily budget to 1 062.21".
One row is worth a closer look. On Advantage+ US the recommended daily budget goes down, from 1 431 to 1 331, while spend goes up 3.3%. The assistant flags it rather than smoothing it over: the current budget is set higher than the campaign actually spends, so the lower cap still lets spend rise.
Explains what's behind the numbers
For the "based on what" part, it goes to the Sellforte Support Center instead of improvising. Recommendations rest on marginal incremental ROAS (miROAS) from the marketing mix model: what the next dollar in that campaign or ad set is expected to return at its current saturation level. Ad platform bidding, by contrast, optimizes on the platform's own attribution.
Puts new-customer return next to it
The recommendations have no customer-type dimension; they optimize incremental sales overall, and the assistant says so. Instead it pulls new-customer return per Meta channel from the MMM for the same 28 days (illustrative demo data, USD):
| Meta channel | Spend | New-customer incremental sales | New-customer ROI |
|---|---|---|---|
| Awareness | 4 730 | 17 125 | 3.62 |
| Sales | 56 682 | 199 274 | 3.52 |
| Prospecting | 37 284 | 127 340 | 3.42 |
| Traffic | 17 235 | 49 370 | 2.86 |
| Advantage+ | 63 383 | 156 418 | 2.47 |
| Retargeting | 18 376 | 17 888 | 0.97 |
Where they agree: retargeting is trimmed, and it also brings in the fewest new customers per dollar. Prospecting US gets scaled, and prospecting ranks high on new customers too. Where they don't: Advantage+ gets the biggest increase but sits second from the bottom on new-customer return. The assistant also reminds you what this table is. It's a historical average, and spend is the channel's full total, so it's a partial ROI. It isn't the return on the next dollar.
Why it matters
An average ROI tells you what a campaign earned. It doesn't tell you what the next dollar will earn, and that's the number a budget move should rest on. Starting from miROAS means you scale where there's still room and trim where the curve has flattened, campaign by campaign.
The new-customer view keeps you honest about your goal. If acquisition is the priority this quarter, you might still take the Advantage+ increase but watch it closely, or lean harder on prospecting.
Try these follow-ups
- Which of these recommendations have the bid or budget setting moving the opposite way from spend? Explain each one.
- Show the same recommendations for Google Ads and list the five biggest changes each way.
- For the campaigns we're scaling, how has their daily spend and incremental sales moved over the last four weeks?
Related
- Sellforte MCP
- Sellforte Performance
- Media Buyer Agent
- Sellforte home
- Marketing mix modeling
- Use case: Did our budget changes pay off?
Try it yourself
Want to see this on your own campaigns? Book a demo, or read how to connect Sellforte MCP to your LLM client. You'll find more about the launch in our Sellforte MCP launch post.
New use cases land in this library regularly, so check back soon.
