Ask Sellforte MCP: Why did our Direct Mail ROI drop last quarter?
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
Finding out what really pulled Direct Mail's return down
The dashboard shows one channel's ROI sliding from one quarter to the next, and someone wants to know why by Friday. Was it the budget, a weak campaign, a shift in mix, or just softer demand? Channel and marketing managers get this question constantly, and the honest answer usually needs a few layers of digging before anyone starts guessing.
What you ask
Compare Q3 2026 with Q2 2026 by media group. Which group's ROI dropped?
Why did Direct Mail ROI drop between Q2 and Q3 2026? Was it spend, campaigns, markets or the baseline? Rank the drivers, and keep what the data shows separate from interpretation.
What Sellforte MCP does
Spots the outlier
First the assistant compares the two quarters by media group. On the demo account, total media spend fell 10.2% while incremental sales held steady (+0.8%), so overall ROI went up from 3.67 to 4.11. Almost every group improved. Direct Mail is the clear exception among the big ones (illustrative demo data, USD):
| Direct Mail | Q2 2026 | Q3 2026 | Change |
|---|---|---|---|
| Spend | 520 000 | 502 857 | −3.3% |
| Incremental sales | 2 403 021 | 2 106 266 | −12.4% |
| ROI (MMM) | 4.62 | 4.19 | −0.43 |
Spend barely moved, but incremental sales fell almost four times as fast. That's what drags the ROI down.
Splits the drop into spend and efficiency
The assistant then does a quick split of its own: spend times ROI. If Q3's spend had kept Q2's ROI of 4.62, it would have returned about 2 323 000. So of the roughly 297 000 lost, about 80 000 comes from spending a little less, and about 217 000 from each dollar earning less. It's a simple bit of arithmetic, not a model output, and the assistant labels it that way. Still, it tells you where to look: efficiency, not budget.
Rules out campaigns and markets
Next it drills down by country and campaign. On the demo data there's nothing to split: Direct Mail is a single campaign running in one market, Germany. So campaign mix and market mix can't be the explanation here. That's a useful result, even if it's a short one.
Checks the timing
Going month by month shows the shape of it (illustrative demo data, USD):
| Month | Spend | Incremental sales | ROI (MMM) |
|---|---|---|---|
| 2026-04 | 157 143 | 834 487 | 5.31 |
| 2026-05 | 122 857 | 740 775 | 6.03 |
| 2026-06 | 240 000 | 827 759 | 3.45 |
| 2026-07 | 160 000 | 561 592 | 3.51 |
| 2026-08 | 14 286 | 19 072 | 1.34 |
| 2026-09 | 328 571 | 1 525 601 | 4.64 |
Direct Mail is lumpy. Spend comes in drops, and each month's return varies a lot. Q2 had two strong months in April and May. Q3 had a weaker July, an almost empty August, and then the biggest month of the half-year in September, which did well but not well enough to lift the quarter back to Q2's level.
Checks the baseline
Last, it checks whether softer demand could be behind it. For Germany, the model's base sales actually rose 4.4% from Q2 to Q3. Promotions contributed 21.1% less, and weather swung from a small positive to a negative contribution. So weaker underlying demand doesn't explain Direct Mail's drop. If anything, the base was stronger.
The ranked answer, then: the drop comes from lower return per dollar, concentrated in July and August, with lower spend a distant second. Campaign mix, market mix and baseline don't explain it. Why those drops returned less, whether it's the offer, the targeting or the timing, isn't in the data. The assistant says so and treats any reason as a hypothesis to check with the team that runs the mailings.
Why it matters
A weaker quarter has many possible parents, just like a strong one. Without separating them, it's easy to blame the budget or the season and move on, and then make the same call next quarter. Here, the checks clear the baseline and the mix, and point at the individual drops. That's a much better conversation to have with whoever plans them.
It also helps to keep the data and the story apart. The numbers say which month and how much. The reasons are your team's to test.
Try these follow-ups
- Compare each Direct Mail month in 2026 with the same month in 2025. Is the Q3 pattern new?
- Which media groups improved ROI the most from Q2 to Q3, and was it from lower spend or higher incremental sales?
- Split Direct Mail's Q3 incremental sales by new and returning customers.
- Show Germany's total sales for Q3 split into base, promotions, weather and media, week by week.
Related
- Sellforte MCP
- Marketing mix modeling
- Sellforte dashboard
- Use case: Can you build our monthly management report?
Try it yourself
Got a channel whose numbers you can't quite explain? Book a demo, or check how to connect Sellforte MCP to your LLM client. For the bigger picture, read the Sellforte MCP launch post.
Fresh use cases are added to this library regularly.
