Ask Sellforte MCP: How should we split next year's media budget?
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 fixed media budget into next year's channel split
Every autumn someone gets handed a number: "here's next year's media budget, tell us where it goes." It's usually the marketing director or head of media, and the plan has to survive a CFO review. Last year's split plus or minus a bit is easy to write and hard to defend, so you want the model to do the first draft.
What you ask
How should I allocate media for 2027 if I have $17 500 000 for media? Start from what we spent and earned over the last 12 months.
What does the last 10% of budget actually buy us?
What Sellforte MCP does
Starts from the last 12 months
First the assistant pulls the marketing mix model's results for 2025-11-02 to 2026-11-01: spend, incremental sales and ROI by media group. On the demo account that's 17 492 953 in spend and 66 492 139 in incremental sales, an ROI of 3.80. So the budget in the prompt is roughly a flat year.
| Media group (illustrative demo data, USD) | Spend | Incremental sales | ROI |
|---|---|---|---|
| Cross-network | 5 397 867 | 23 788 264 | 4.41 |
| Paid Social | 3 216 237 | 17 413 163 | 5.41 |
| Direct Mail | 2 005 714 | 8 375 177 | 4.18 |
| 1 737 692 | 3 442 620 | 1.98 | |
| TV | 1 690 984 | 2 354 961 | 1.39 |
| Paid Search | 1 452 504 | 4 815 966 | 3.32 |
| All media groups | 17 492 953 | 66 492 139 | 3.80 |
That's history, though. Average ROI tells you where money worked on average, not where the next dollar goes furthest. For that you need the optimizer.
Proposes scenarios and asks before saving them
Next, the assistant suggests two optimizer scenarios for the year: a business-as-usual plan with every channel locked at its current spend, and an optimized plan at the same total, planned by advertising channel and country. Creating scenarios saves them in your Sellforte account, so the assistant confirms the setup with you before it runs. In this demo run we stopped at that point and didn't create new ones.
If you have a sales goal rather than a fixed amount, the optimizer can also plan to a total sales target, an incremental sales target or a target ROI. You give it the figure to reach.
Reads saved plans at different budget levels
To show what the answer looks like, the assistant read three plans already saved on the demo account for the Q4 window (2026-09-27 to 2027-01-09). It flagged all three as stale: they were created before the model's latest refresh, so a real plan would start by re-creating them.
| Saved plan (illustrative demo data, USD) | Spend | Incremental sales | ROI |
|---|---|---|---|
| Business as usual | 4 276 232 | 20 044 353 | 4.69 |
| Optimized, same budget | 4 276 231 | 20 155 674 | 4.71 |
| Optimized, 10% less budget | 3 848 609 | 18 423 017 | 4.79 |
At the same budget, the optimized plan adds 111 321 in incremental sales (+0.56%). It gets there by cutting Direct Mail by 107 247 and TV by 46 003 (both about 22% down) and moving the money mostly into Google Performance Max (+68 720), Google YouTube (+17 853, almost 20% up) and Meta Sales (+10 555). Email stays where it is, because it's locked in the plan's limits.
The follow-up is where it gets interesting. Going from the 10%-smaller plan to the full one adds 427 623 in spend and 1 732 657 in incremental sales. The total ROI dips slightly, from 4.79 to 4.71, which is what you'd expect: each extra dollar earns a bit less than the average one. On these numbers the last slice of budget still returns roughly 4 dollars of incremental sales per dollar.
Why it matters
An annual plan really has two questions in it: how big should the budget be, and how should it be split? The same-budget comparison answers the second one. The step between budget levels answers the first, because the difference in incremental sales between two levels is what the extra money buys. That's the number your CFO actually needs.
It also pays to know how much of next year's sales will come without media at all, from the baseline and promotions, before you argue about the media line. And keep the optimizer's limits in mind: it plans by channel and country, not by campaign name, and it can't set a separate growth target per sales channel such as online versus stores.
Try these follow-ups
- Create the same plan at 90%, 100% and 110% of the budget and show what each step adds in incremental sales.
- Which channels does the optimized plan grow and cut, by country?
- What budget would we need to reach $70 000 000 in incremental sales next year?
- How much of next year's sales does the plan expect from the baseline and promotions, without media?
Related
- Sellforte MCP
- Sellforte Optimizer
- Media Planner Agent
- Multi-market MMM
- Ask Sellforte MCP: Where should we move budget mid-month?
Try it yourself
Want to see it on your own model? Book a demo, or read how to connect Sellforte MCP to your LLM client. New use cases land in this library regularly.
