Sena records marketing ROI with real-world zero-party data.
The formula is settled: incremental profit attributable to marketing, divided by marketing cost. Three properties of the available evidence make the numerator the whole difficulty.
Spend is invoiced, so cost is known to the currency unit. Return is estimated from whatever measurement the channel supports. That asymmetry gives every ROI figure the appearance of precision it inherits from one side only.
Reach, impressions, clicks, video completions, and site sessions are counted because they are countable. Each one is a step before a purchase, and the step from the last proxy to the till is the part left modelled.
Where the category sells through physical outlets, the transaction happens where zero tracking reaches. Attribution then runs on modelled uplift, and the model is calibrated against the same history it is predicting.
Sena records the purchase the shopper was aiming at and asks the shopper who made it what moved them. The cost side and the media plan stay where they are.
Recorded purchase in each market, so the return rests on a transaction over a proxy for one.
The same shopper states what prompted the purchase, so credit comes from the buyer.
Photographs and receipts capture purchases in physical outlets, where digital measurement holds zero rows.
Marketing ROI states the incremental profit produced by marketing spend, divided by that spend. It is a finance question answered with marketing evidence, which is why it sits in the finance function and gets argued about in the marketing one.
A full analysis needs four numbers, and only the first is straightforward.
Invoiced spend, including production and agency fees. Known exactly.
Sales in the period and territory the spend covered. Available, and it includes everything that would have happened anyway.
The part of that revenue the spending caused. This is the contested number.
Gross profit on the incremental units, over revenue, which is where a positive ROI frequently turns negative.
Step three is where every method differs, and step four is where the answer changes sign. An analysis that ends at revenue reports a return the margin can reverse.
The measures in common use divide into three groups. Exposure metrics count reach, impressions, and frequency. Response metrics count clicks, sessions, searches, and inquiries. Outcome metrics count purchases, revenue, and margin. Reporting climbs that list as the measurement gets harder, and the ROI question needs the third group.
A calculator applies the formula correctly and takes the incremental share as an input. Where that input comes from is the entire question, and a calculator supplies zero of it. A figure entered as an assumption produces an ROI carrying exactly the confidence of the assumption.
Three moves change the ratio, and they are ordered by how measurable they are.
Measurable, once the response is read per market over per region.
Measurable ahead of the spend, through stated response.
Measurable, and the smallest of the three in most categories.
| Output | What it settles | Where it goes wrong |
|---|---|---|
| Purchase in the market | What was bought, per market. | Approximated from shipment data. |
| Stated prompt | What moved the shopper. | Assigned by attribution model. |
| Availability at sale | Whether the item was buyable. | Assumed present. |
| Response per market | Where the spend works hardest. | Averaged across a region. |
A measured return rests on four collected inputs, captured per market across the campaign window.
| Input | What it answers |
|---|---|
| Receipts | What was bought during the campaign, and at what price. |
| Store captures | Whether the promoted item was available to buy. |
| Geo-verified photos | The outlet and the display, dated and placed. |
| Stated preference | What prompted the purchase, in the shopper's own account. |
The team's own numbers join separately. Spend files, the media plan, sales history, and the attribution model connect through 250+ integrations, so cost and measured return sit together.
Three questions sit outside the media stack and the ledger together.
Real people share what they buy and prefer, under explicit consent. Sena captures it directly at the source, so every figure traces back to where it came from whenever a number comes under question.
Purchase and prompt come from the same buyer, and the market and week travel with every figure.
One budget resolves into the markets it ran in, each with its own measured response.
Credit for a purchase comes from the shopper who made it, over a model.
An absent item and a weak message produce the same flat line until photographs divide them.
Marketing ROI built on it ends where the tracking ends, and in a category selling through physical outlets that is well short of the till. Sena builds every figure from real-world signals captured across the campaign window, from the display photographed in the outlet through to the shopper naming what prompted the purchase.
The item bought and the price paid are recorded during the campaign window, so the numerator is a transaction.
Reads whether the promoted item and its display were present, off images captured in real outlets.
Signal arrives from the consumer network under explicit consent. What prompted a purchase is reported by the buyer.
Spend files, the media plan, sales history, and the attribution model connect over 250+ integrations, so cost and measured return sit together.
Market, week, and source buyers ride on each figure, so a return claim opens onto the people who produced it.
Spend history, campaign windows, and prior period sales, reaching back across every scanned cycle.
Four teams argue from the same ROI figure, and each one needs a different part of the numerator.
The ROI figure. Needs the incremental share measured.
The channel mix. Needs response read per market.
The proposition. Needs the prompt stated by buyers.
The in-store execution. Needs availability confirmed at the outlet.
The same measured numerator, read against whatever each industry counts as a response.
Purchase uplift per market, with shelf availability confirmed.
Occasion-led response, checked against the display in the outlet.
Recommendation against advertising, split on the stated prompt.
Application uplift per market, credited from the buyer's account.
One mechanism, applied per market across a campaign window. Each step is documented, which is what carries an ROI figure past a finance review.
The same shoppers supply the purchase and the prompt behind it, so return and cause arrive attached.
Credit comes from the buyer's own account of what moved them, over an allocation rule.
Measured incremental purchase over invoiced spend, per market, with margin applied to the units.
Most marketing ROI runs on an exact cost and an estimated return, and the estimate carries every proxy in the chain. Sena measures the return at the purchase.
| Capability area | Typical setup | Sena |
|---|---|---|
| The numerator | Proxy events, modeled to sales. | Purchase recorded at the till. |
| Attribution | Assigned by a model. | Stated by the shopper who bought. |
| Offline purchase | Outside every tracker. | Captured by receipt and photograph. |
| A flat result | Read as weak creative. | Divided from absent stock. |
| Reporting level | One campaign, one number. | Response measured per market. |
| Margin on uplift | Applied late or omitted. | Applied to the measured units. |
| Evidence in a review | A model output. | Open any figure onto its buyers. |
Three ROI questions that end where digital measurement ends.
Read what shoppers bought during the window, so the numerator rests on a transaction.
Ask buyers what prompted the purchase, so trade and media spend get separated by the shopper.
Confirm whether the promoted item was on the shelf, so weak creative and absent stock divide.
A measured numerator settles what the spend produced. The four decisions below are what a finance team runs around it.
Every finance decision Sena answers, in one place
The spend side read
Reads what the spend itself costs to place and serve.
Explore cost intelligence 02The measurement underneath
Supplies the margin and revenue measurement the ratio uses.
Explore financial analytics 03What the return becomes
Carries the measured response forward into the revenue read.
Explore revenue predictive analytics 04Where the budget comes from
Holds the plan the marketing budget was committed inside.
Explore financial planning and analysisThe walkthrough takes one campaign window in one region, reports recorded purchase per market inside it, separates the markets where the item was absent, and reads the stated prompt from buyers while the team watches.
Name the campaign and the markets it ran in, and we will measure the return at the purchase.