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
Closing the loop on last cycle's budget changes
You moved budget two weeks ago. Maybe you acted on a recommendation, maybe it was a hunch. Now the next round of changes is due, and before you make it you want to know whether the last one actually worked. That's a question performance marketers managing several markets face every cycle, and platform dashboards can't answer it on incremental terms.
What you ask
We raised the budgets of our Meta Sales and Meta Retargeting campaigns in Germany and the US on 2026-09-06. Compare spend, incremental sales and ROI for those campaigns in the four weeks after the change with the four weeks before. Do the same for the Meta Prospecting campaigns we didn't change, and tell me whether the changed ones improved more.
Show me the Sales and Prospecting campaigns week by week from 2026-08-09 to 2026-10-03. Did the gain hold?
What Sellforte MCP does
You tell the assistant which campaigns you changed and when. It takes it from there.
First it lists the Meta campaigns with their before-and-after numbers to confirm the exact campaign names. Then it runs two comparisons from the model, using equal four-week windows: 2026-08-09 to 2026-09-05 before, and 2026-09-06 to 2026-10-03 after. One covers the four campaigns you changed. The other covers the two Prospecting campaigns you left alone, which act as a reference group. The query layer calculates the changes, so the assistant isn't doing the arithmetic itself.
| Group | Spend before (USD) | Spend after (USD) | Incremental sales before (USD) | Incremental sales after (USD) | ROI before | ROI after |
|---|
| Changed: Sales and Retargeting, DE and USA | 49 849 | 62 790 | 570 157 | 871 744 | 11.44 | 13.88 |
| Unchanged: Prospecting, DE and USA | 28 034 | 27 908 | 79 716 | 88 525 | 2.84 | 3.17 |
Illustrative demo data. Demo account figures are in USD.
The first read looks good. Spend on the changed campaigns went up 25.96%, incremental sales up 52.9% and ROI up 2.45 points. The reference group, with flat spend (−0.45%), also improved: incremental sales up 11.05% and ROI up 0.33. So part of the lift is something that happened to Meta as a whole in those weeks, and part is on top of that.
Inside the changed group, the two Meta Sales campaigns carried the result. Sales DE went from an ROI of 12.43 to 16.45. The Retargeting campaigns nearly tripled their spend (+177.64%), but their ROI barely moved, from 1.93 to 2.10 in Germany. More money there bought roughly the same return per dollar as before.
Did the gain hold?
The follow-up splits the same eight weeks by week. Here's the Meta Sales campaign in Germany:
| Week starting | Spend (USD) | Incremental sales (USD) | ROI |
|---|
| 2026-08-09 | 6 959 | 96 281 | 13.83 |
| 2026-08-16 | 8 529 | 91 255 | 10.70 |
| 2026-08-23 | 7 426 | 117 549 | 15.83 |
| 2026-08-30 | 9 288 | 95 323 | 10.26 |
| 2026-09-06 | 12 616 | 231 613 | 18.36 |
| 2026-09-13 | 9 353 | 224 528 | 24.01 |
| 2026-09-20 | 8 218 | 60 557 | 7.37 |
| 2026-09-27 | 6 713 | 90 361 | 13.46 |
Illustrative demo data. Demo account figures are in USD.
That changes the story. Almost all of the gain sits in the first two weeks after the change. By the third week incremental sales dropped below every week before the change, and by the fourth, spend was back under where it started. The US Sales campaign follows the same shape. The Prospecting campaigns, meanwhile, show no jump like that after 2026-09-06.
So the honest verdict is "it worked, briefly". Before scaling the change again, you'd want to know what happened in those two weeks and why spend fell back afterwards.
Why it matters
The best estimate of what your next budget increase will buy is what the last one bought. That sounds obvious, but most teams skip the check and move straight to the next round. Comparing the changed campaigns with ones you didn't touch separates your decision from whatever was happening to the whole platform at the same time.
It's still not an experiment. Seasonality, a promotion or a competitor's move can show up in one group and not the other, and the most recent week is always still filling in because of carryover. If a decision is big enough to need proof, run a proper incrementality test. For a two-week optimization cycle, though, this check is quick and keeps each round honest.
Try these follow-ups
- Split the result by country. Did Germany and the US respond the same way?
- Compare the changed campaigns with all other Meta campaigns instead of just Prospecting.
- Which of these campaigns does Sellforte currently recommend scaling or cutting?
- Repeat the comparison with two-week windows so I can check the next cycle the same way.
Related
Try it yourself
Curious what your last round of changes really bought? Book a demo, or if you already use Sellforte, see how to connect Sellforte MCP to your LLM client.
We add new use cases to this library regularly.
