Put the next euro where it still returns.
TEA takes your spend history and the response curves and computes the allocation that returns the most within your budget. Not just the optimum, but why: what moves, what it gains, what it costs.
Invitation only while in beta. Tell us who you are and we get in touch.
Today against the recommendation, six channels
weekly spend, same total budget, sample market- Spending today
- Recommended, up
- Recommended, down
An illustration on one of our sample markets. On your file the recommendation comes with the fit quality of the curves it stands on.
Gain and loss, channel by channel
Every move has a price and a return. The cuts cost less than the additions gain, because the money leaves the flat part of one curve and lands on the climb of another. The waterfall adds up to the net uplift, and nothing is hidden in a total.
Hover a bar for the move behind it. In the product, the channel opens with its curve, so the reason for the move is one click from the move.
Revenue impact of the reallocation
a year, sample marketAnd if the budget moves?
The optimum is recomputed at every budget, so the question “what would ten per cent less cost us” has a number before the meeting rather than after it. The curve flattens on the right because the channels do.
The slider is on the page. Drag the total and the allocation follows, with the model’s recommendation held still beside it so nothing moves under your hand.
Expected uplift as the total budget moves
against today’s revenueWhat the analysis hands back
A recommendation with its reasons
Current against recommended for every channel, the gain on each move, and the curve each move is read from. A number you can defend because you can show where it comes from.
Three scenarios, not one
Conservative, recommended and aggressive, with the expected outcome for each. The plan that gets approved is rarely the optimum, and now it is a chosen distance from it.
Constraints you set
A contract that cannot be cut, a floor on presence, a cap on a platform. The optimum is found within them, and the gain given up for each constraint is shown.
Drag it yourself
Macro tuning on the total, micro tuning per channel. The recommendation never moves while you drag, so what you compare against stays still.
What your file needs
A date
One row per week, a year at least. Two years makes the curves steadier.
week date
The number the spend should explain
Sales, units, revenue. One column.
sales revenue
At least two channels
Spend per channel in euro, GRPs or impressions. An allocation needs somewhere to move
the money to. tv_spend search_spend social_spend
The curves are fitted first and the allocation is computed on them. If a curve is thin, the recommendation says so instead of hiding it under a confident number.
How it works
-
Drop the file
CSV or Excel. The Data Scout maps the date, the KPI and the channels, and says whether each channel has enough variation to carry a curve.
-
Set the budget and the constraints
The total, the floors, the caps. Or none: the unconstrained optimum is the one to argue from.
-
Read the allocation
The moves, the gains, the three scenarios, the slider. Export the recommendation as a document, with the curves it stands on inside it.
What budget allocation is, in a paragraph
Given a response curve per channel, the allocation that returns the most is the one where the next euro returns the same in every channel. Anywhere else, moving a euro from the channel with the lower marginal return to the one with the higher gains something, and the optimiser keeps moving euros until it cannot. The optimum is not a guess and not a weighting: it is a point the curves agree on.
What makes it usable is everything around the point: the constraints a real plan carries, the three scenarios at different distances from the optimum, the slider on the total, and the fit quality of the curves, declared beside the recommendation. TEA computes the point in seconds and spends the rest of the page on the reasons.
How is the optimum found?
From the response curves. The allocation that returns the most is the one where the next euro returns the same in every channel: anywhere else, moving a euro from the channel with the lower marginal return to the higher one gains something. TEA solves for that point under your total budget and any floor or cap you set.
Can I lock a channel?
Yes. A contract that cannot be cut, a minimum presence, a cap on a platform: set the constraint and the optimum is found within it. The gain you give up for the constraint is shown too.
What if my curves are thin?
Then the recommendation is thin, and the page says so. A curve fitted on few spend levels carries a wide band, and the optimum inherits it. TEA declares the fit quality beside the recommendation rather than hiding it under a confident number.
Does it replace the planner?
No. It replaces the spreadsheet the planner built at midnight. The recommendation comes with the reason for every move, three scenarios, and a slider on the total budget, so the conversation starts from a number instead of ending in one.
Where it comes from
The next euro, with a reason.
Drop the file, read the allocation, take it to the meeting. Free while in beta, by invitation.
Join the beta