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
A promotion has just ended and the next one of the same kind is already on the calendar. If you plan promotions, this is the moment to learn from it, while the details are fresh: what did it actually add, how did it sit next to media, and was it better or worse than last time? Without a model, you usually end up comparing gross sales during the promo with the week before, which mixes in everything else that happened.
Our Shopping Week ran from 2026-10-05 to 2026-10-11. How much incremental sales did it drive compared with media, and how does it compare with last year's Shopping Week from 2025-10-06 to 2025-10-12?
Break this year's Shopping Week down by discount level. Which discounts paid back?
We ran both prompts on our demo account. All figures below are illustrative demo data in USD.
Sellforte's promotions data is organized by promotion type, discount level, product category, customer segment, sales channel and country, by day. There's no single "promotion ID", so the assistant looked the promotion up by its type and dates. A quick daily check confirmed Shopping Week had activity on exactly those seven days in both years.
Next it split total sales for 2026-10-05 to 2026-10-11 into base, media, promotions and weather:
| Component | Sales |
|---|---|
| Base | 1 555 203 |
| Media | 1 117 419 |
| All promotions | 846 650 |
| of which Shopping Week | 170 660 |
| Weather | -56 443 |
| Total | 3 462 828 |
Media brought in 32.3% of the week's sales and promotions 24.4%. But Shopping Week wasn't the only promotion running. Product promotions (472 766), vouchers (130 062) and a Sale Sunday (73 161) overlapped with it. So Shopping Week itself added 170 660, a fairly small slice of a busy week.
| Metric | Shopping Week 2025 | Shopping Week 2026 | Change |
|---|---|---|---|
| Promo investment | 54 051 | 108 087 | +99.97% |
| Incremental sales | 73 865 | 170 660 | +131.04% |
| Incremental margin | 27 290 | 63 170 | +131.48% |
| ROI | 1.37 | 1.58 | +15.54% |
| Margin ROI | 0.505 | 0.584 | +15.76% |
This year's Shopping Week was roughly twice as big and a bit more efficient. Every dollar of promo investment brought back 1.58 in incremental sales, up from 1.37. On margin, though, the picture is more sober: margin ROI was 0.584, so each dollar invested returned about 58 cents of incremental margin.
The follow-up is where the planning advice comes from. Shallow discounts did the heavy lifting:
| Discount | Promo investment | Incremental sales | ROI | Margin ROI |
|---|---|---|---|---|
| 10% | 5 165 | 14 070 | 2.72 | 1.27 |
| 15% | 44 089 | 110 741 | 2.51 | 1.09 |
| 20% | 10 504 | 21 042 | 2.00 | 0.801 |
| 60% | 26 993 | 10 087 | 0.374 | -0.0747 |
The 15% discount alone produced 110 741 of the 170 660. The 60% discount took a quarter of the investment, returned 10 087 in sales and slightly negative incremental margin. Every discount level from 50% upward came in below 1 on ROI.
A promotion can look like a hit on the sales chart and still be the wrong one to repeat. Here, the headline says "twice as big as last year", but the useful answer is narrower: keep the shallow discounts, and question the deep ones before the next Shopping Week goes into the calendar.
There's a second check worth adding. A sales event can pull sales forward from the days around it rather than create new ones. Comparing the event window with the days before it, and the days after, is the simplest way to see whether that happened.
One limit to know: because promotions are found by type and dates, two promotions of the same type running at the same time can't be told apart. Name the type, dates and any category or market up front.
Want to review your own promotions this way? Book a demo, or if you're a Sellforte customer, follow these steps to connect Sellforte MCP to your LLM client. More use cases are added to this library regularly.