Budget moved to where it earns,
not where it was spent last quarter
Budgets usually get set by inertia, whatever got spent last quarter gets spent again, with small adjustments. We build the model that shows what each channel actually earns per dollar, by real outcome rather than platform-reported numbers, and forecast forward from there, the same comparison that once found one placement converting 2.7 times cheaper than another.
Where the money leaks today
Most marketing budgets get set by habit: last quarter’s split, nudged slightly based on which channel manager made the most noise in the planning meeting. The platforms themselves do not help, since each one reports its own performance in the best possible light, and a channel that looks efficient from inside its own dashboard can be quietly stealing credit from a channel that did the real work of creating intent.
We found a direct, measurable version of this gap comparing conversation cost across two placements on the same campaign: $17 on one, $46 on the other, a 2.7 times difference invisible from either platform’s own reporting and only visible once spend was compared against real CRM outcomes. In a separate expansion across four markets, we found two countries producing zero real purchases despite active ad spend, a forecasting failure that a model built on real outcomes caught immediately and a model built on platform metrics alone would have missed entirely.
Budget that follows last quarter’s split instead of this quarter’s real numbers is budget that is, by definition, not optimized.
What we do
We calculate real cost per outcome, lead, purchase, qualified conversation, by channel, using your CRM or order data rather than each platform’s self-reported conversions, since that is the only fair basis for comparing channels against each other. From that real baseline, we build a forecast for the next quarter under two or three spend scenarios, so the reallocation decision has an actual range of outcomes attached to it rather than a single confident guess.
We flag explicitly any channel that is getting spend without real conversion behind it, the zero-purchase-market pattern we have found directly in past work, since that is often the fastest, least controversial place to cut. Seasonality and any known upcoming change, a new competitor, a platform policy shift, a product launch, gets folded into the forecast rather than ignored because it complicates the model.
What we need from you
Access to every ad channel you spend on and your real CRM or order outcome data, since the entire value of this exercise depends on comparing spend against real results rather than platform-reported numbers. A short conversation about anything you already suspect, a channel you think is overrated, a market that feels soft, usually points straight at what the model will confirm or correct.
How we measure
The forecast’s accuracy against what actually happens over the following quarter, and the real dollar impact of the reallocation once it ships, checked against the model’s own prediction.
Price and timeline
| Option | Price | What it covers | Timeline |
|---|---|---|---|
| Launch or audit | from $1,000 | Real cost-per-outcome model, quarterly forecast, reallocation plan | 1 to 2 weeks |
| Monthly management | from $700 / month | Ongoing reforecasting as channel performance shifts | monthly, no lock-in |
| Full control, handover to your team | from $2,200 | Full model and forecasting framework handed to your team | 2 to 3 weeks |
Related
This builds directly on end-to-end marketing attribution and a unit economics and CAC/LTV model. The full build is on the performance marketing service page. For the automated layer, see ad budget allocation. Real examples: the Bali lead-routing project and the Latin America media-buying expansion.
Spending across more than one channel and not sure the split still makes sense? Get in touch and bring your current numbers.
FAQ
How much does budget allocation and forecasting cost?
From $1,000 for a real cost-per-outcome model by channel and a forecast with a specific reallocation recommendation, delivered in 1 to 2 weeks.
How often should this be redone?
Quarterly at minimum, since channel costs and conversion rates shift, and more often if you are running a new channel or market where the numbers have not stabilized yet.
What budget size is this worth doing for?
Any business spending across two or more channels benefits, since the main value is comparing them against each other on a fair basis; a single-channel budget has less to reallocate but still benefits from an honest forecast.
What do you need from us?
Access to your ad accounts across every channel you run, and your real order or CRM outcome data, since the model's value comes entirely from comparing spend against real results, not platform-reported conversions.
How do you report on the forecast?
A specific reallocation recommendation with the model behind it, plus a check-in against the forecast once a quarter of real spend has passed, so the next forecast gets sharper.