Same $31.8M. Move it like this and the model expects $11.3M more revenue.
Not a bigger budget — a different shape. The plan takes money out of the channels whose next unit of spend is weakest and puts it into the ones that still have room. Every move below is sized by the response curves on the previous screen, and nothing is moved further than the cap you set.
Where the money moves
$4.1M reallocated · 13% of the budget · total unchanged at $31.8M
Money flows until every channel's next unit of spend returns the same amount — $1.01 here. That is what "optimal" means, and it is why the winner stops where it does rather than taking everything. TV — Broadcast loses the largest share because its next unit was the weakest at $0.97, not because its overall return was lowest.
If you’d rather not move that much
Push a channel far past anything you have actually spent on it and we are guessing rather than measuring. A tighter cap gives up some of the gain and buys a plan that survives contact with the agency contracts.
Or change the total, not just the shape
100% · $31.8MLeave it at 100% to shuffle the money you already spend. Push it up and the planner will show you where the extra stops paying — it solves against the curves, so it cannot promise a ceiling the data does not reach.
Takes the biggest number on offer and ignores how badly it could go.
The number that matters most
97%
the odds this plan genuinely beats what you’re doing now
Not the odds it hits $11.3M exactly — the odds it comes out ahead at all. We rerun the plan against every version of the model your data supports and count how often it wins. Roughly one time in twenty, it doesn't.
Even the bottom of this range comes out ahead — which is unusual, and worth checking against how far you let channels move.
Channels this plan will not touch
Affiliate, Retail Media are pinned at current spend and greyed in the ribbon above.
These channels' spend does not vary enough, did not run for enough of the window, or moves too closely with the rest of the model for the data to say what they returned. Their numbers come mostly from the starting assumption. They stay in the model, so they do not distort the others, but the budget planner will not move them.
You will get no recommendation for these channels — deliberately. A plan that reallocated budget on the strength of an assumption would look exactly like one built on evidence.
3 of 6 channels wanted to go further and the ±30% cap stopped them. Letting them go further means trusting spend levels you have never tried.
What you walk into the meeting with
A spend table your finance team can paste, with the odds, the cap and the withheld channels written into the top of the file. The caveats travel with it — they are not an appendix, and they are not something the person who opens it has to remember to ask for.
The plan, channel by channel
last column is revenue per unit of spend| Channel | Spending now | Spend instead | Change | The move | Next unit buys |
|---|---|---|---|---|---|
| TV — Broadcastwanted less | $10.6M | $7.4M | -30% | $0.97 | |
| Paid Socialwanted more | $7.3M | $9.5M | +30% | $1.03 | |
| Paid Searchwanted more | $6.3M | $8.2M | +30% | $5.77 | |
| Programmatic Display | $4.1M | $3.2M | -22% | $1.01 | |
| Affiliateleft alone | $2.2M | $2.2M | 0% | $0.79 | |
| Retail Medialeft alone | $1.4M | $1.4M | 0% | $0.86 |
How stable is this plan?
The allocation was re-solved independently against 60 individual posterior draws. Where those solutions disagree, the recommendation is being driven by parameters the data did not pin down — so the spread below is the honest uncertainty on the plan itself, not on the outcome it predicts.
- TV — Broadcast$7.4M–$12.1M±32%
- Paid Social$5.1M–$9.5M±23%
- Paid Search$8.1M–$8.2M±1%
- Programmatic Display$2.9M–$5.3M±38%
- Affiliate$2.2M–$2.2M±0%
- Retail Media$1.4M–$1.4M±0%