Every channel gets tired. Drag one and see exactly where.
The first dollar you put into a channel works harder than the fifth. These curves are your model’s actual answer to “how much harder” — drag a channel’s spend and the revenue figure recalculates from the same maths the budget planner uses. Nothing is being simulated for show.
ChannelShapeSpend for the year — drag meRevenue backNext $1 returns
TV — Broadcast
tires slowly — still the best next dollar
$10.6M
$13.2M
as you ran it
$0.70
below break-even
Read it like this
“Next $1 returns” is the number that should drive a budget decision, not the average. TV — Broadcast has averaged $1.25 across everything you spent there, but the next dollar returns about $0.70.
Your year, as you have it set right now
$476.7M
total revenue · against $476.7M at the spend you actually ran
$31.8M
media spend · was $31.8M
$4.14
revenue per $1 of media
These figures include five channels at today's spend, with no slider: we could not measure what they return on their own.
Nothing is moved yet, so this is the run's own answer: $476.7M, the same figure on every other screen. Drag anything and watch how much less the revenue moves than the spend does.
The caveat that travels with this screen
These curves are the model’s best guess, drawn as single lines because that is what you can drag. The ranges behind them are wide enough that a small move is a direction, not a forecast. The budget planner puts the uncertainty back on — it reports the odds a plan beats what you do today rather than a single number.
What dragging does not change
Nothing here re-runs the model, and nothing here changes a verdict. Whether a channel has proven it pays is a property of the fit and stays where it was on the channels table. This screen is for exploring what the answer implies, not for arriving at a new one. revenue per unit of spend.
Not shownPaid Social, Paid Search, Programmatic Display, Affiliate and Retail Media, not found on their own. Why each one is missing
Two different questions, and people mix them up
“Was this channel worth it?” and “should I put more into it?” are not the same question. The first looks at everything you have already spent there. The second looks only at the next dollar — and the next dollar always earns less, because channels fill up. A channel with a brilliant history can still be the wrong place for more money. On this run we can only answer either question for TV — Broadcast — the other channels have no return of their own, for the reasons further down.
What your next dollar is worth, channel by channel
Each bar is the range we think the answer sits in, and the solid mark is our best single guess. A short bar means we are confident; a long one means the data did not pin it down. The vertical line is break-even — a bar sitting entirely to its right has proven it pays, and one that straddles it has not, whichever side its mark falls on.
revenue per unit of spend
TV — Broadcast
$0.70$0.28–$1.53
$0.18$1.63break-even
Not shownPaid Social, Paid Search, Programmatic Display, Affiliate and Retail Media, not found on their own. Why each one is missing
Keep funding these
No channel has proven the next dollar pays
The range of every channel we could measure crosses break-even, so we cannot yet say of any of them that more money would pay for itself. That is a statement about the data, not about the media.
The best of them
TV — Broadcast
Its best guess of $0.70 is the highest here, so it is where an extra dollar should go first — but the range still reaches down past break-even, so treat that as a bet, not a fact.
Where we are least sure
Your best bet is also your blurriest
TV — Broadcast's range runs from $0.28 to $1.53 — the widest here, and it is the same channel we just told you to back. Usually that means its spend barely moved, so the model had little to learn from. Vary it and this bar will tighten.
What each channel has returned across everything you already spent, next to what the next dollar would return. The gap between those two is what you have used up: early spend in a channel works harder than late spend, so a channel can have a great history and very little left to give.
revenue per unit of spend
Channel
Spend
So far
Next dollar
Already used up
Share of the effect
TV — Broadcast
$10.6M
$1.14$0.51–$2.34
$0.70$0.28–$1.53
−39%
2.5%1.1–5.2%
Paid Social
$7.3M
no figures of its own — counted only in the media total · why
Paid Search
$6.3M
no figures of its own — counted only in the media total · why
Programmatic Display
$4.1M
no figures of its own — counted only in the media total · why
Affiliate
$2.2M
no figures of its own — counted only in the media total · why
Retail Media
$1.4M
no figures of its own — counted only in the media total · why
Channels with no number of their own
Before fitting your data we hid returns we chose in your own spend and checked whether the setup could find them again. For these channels it could not, so we report no return, next-dollar figure or curve for them anywhere — a number here would be a guess we could not defend. Their spend still did something: it is counted in the total effect of your media, and the budget planner leaves it where it is.
Paid Social$7.3M spent
Counted only inside the total effect of your media.
When we hid known returns in your spend, the chosen setup did not find Paid Social's return reliably, so we do not report a return for it. It still counts toward the total effect of your media, and the budget planner will not move it.
What it costs you
You get no return and no budget advice for Paid Social. A test that changes its spend on its own would make it measurable.
A test you could run
The test under Paid Search tells the two apart.
Paid Search$6.3M spent
Counted only inside the total effect of your media.
When we hid known returns in your spend, the chosen setup did not find Paid Search's return reliably, so we do not report a return for it. It still counts toward the total effect of your media, and the budget planner will not move it.
What it costs you
You get no return and no budget advice for Paid Search. A test that changes its spend on its own would make it measurable.
A test you could run
To tell Paid Social and Paid Search apart, pause Paid Search, the one you spend less on, everywhere for 13 weeks, and keep Paid Social running as usual.
A test this long would add little: the range we could give for Paid Search would narrow by only about 5%.
Programmatic Display$4.1M spent
Counted only inside the total effect of your media.
When we hid known returns in your spend, the chosen setup did not find Programmatic Display's return reliably, so we do not report a return for it. It still counts toward the total effect of your media, and the budget planner will not move it.
What it costs you
You get no return and no budget advice for Programmatic Display. A test that changes its spend on its own would make it measurable.
A test you could run
Run this one first
To measure Programmatic Display on its own, pause it everywhere for 13 weeks.
A test this long would add little: the range we could give for Programmatic Display would narrow by only about 10%.
Affiliate$2.2M spent
Counted only inside the total effect of your media.
When we hid known returns in your spend, the chosen setup did not find Affiliate's return reliably, so we do not report a return for it. It still counts toward the total effect of your media, and the budget planner will not move it.
What it costs you
You get no return and no budget advice for Affiliate. A test that changes its spend on its own would make it measurable.
A test you could run
To measure Affiliate on its own, pause it everywhere for 13 weeks.
A test this long would add little: the range we could give for Affiliate would narrow by only about 5%.
Retail Media$1.4M spent
Counted only inside the total effect of your media.
When we hid known returns in your spend, the chosen setup did not find Retail Media's return reliably, so we do not report a return for it. It still counts toward the total effect of your media, and the budget planner will not move it.
What it costs you
You get no return and no budget advice for Retail Media. A test that changes its spend on its own would make it measurable.
A test you could run
To measure Retail Media on its own, pause it everywhere for 13 weeks.
A test this long would add little: the range we could give for Retail Media would narrow by only about 5%.
Pairs the model could not separate
These channels moved together across your history, so the fit could learn at most what they earned together.
revenue per unit of spend
Moved together, with no combined figure because it would stand in for a return we do not report: Paid Social and Paid Search; Paid Social and Programmatic Display; Paid Search and Programmatic Display.
What each channel would do at other budgets
Each curve runs that channel from spending nothing up to 2.0× what you spent, keeping the same on-and-off pattern rather than smearing the money evenly — so every point on it is a plan you could actually go and buy. The shaded band is how sure we are: where a curve flattens off is itself a guess, and the band shows how much room that guess has.
TV — Broadcast$10.6M now
Doubling spend would add roughly 50% more outcome from this channel — it is approaching its bend.
Not shownPaid Social, Paid Search, Programmatic Display, Affiliate and Retail Media, not found on their own. Why each one is missing