Sena asks the shoppers who made the purchase and brings the answers back with the real-world data behind them.
Category intelligence answers why the category behaves as it does: what a shopper is choosing between, what settles the choice, and what would change it. Three properties of the usual evidence base leave the reason unstated.
A sales file records the item bought. The reason gets reconstructed afterwards from price, promotion, and share movement, which are outputs of the choice ahead of causes of it.
Segments arrive as age, income, and location, because those fields exist. Two shoppers with identical demographics buy for different reasons, and two who share zero attributes buy for the same one.
Where shoppers are asked, the answer arrives with zero purchase attached. Stated importance and revealed choice diverge, and price is the clearest case: it ranks first when asked and predicts less than the reply implies.
Sena is the decision AI with access to real-world data. It records what shoppers bought across the whole category and asks those same shoppers what settled the choice, so the stated reason and the recorded purchase sit on one shopper.
The same shopper supplies the reason and the receipt, so a claim is checkable against a purchase.
Shoppers group by what settled the choice, so a segment predicts the next purchase.
The stated response covers what would change the choice, including options the market has yet to offer.
Category intelligence explains how a category works from the shopper outward. It answers what a shopper is choosing between, in what order the choice narrows, what settles it, and what would change it. The output is a decision structure ahead of a report.
The consumer decision tree is the working artifact of category intelligence. It sets out the order in which a shopper narrows the field, from the first split down to the item taken. A tree for the same category differs by market, which is what makes it worth drawing.
The attribute a shopper resolves before any other. Where this is an occasion in one market and a format in another, the two markets need different ranges.
Whether brand precedes price or price precedes format decides which items are genuinely in competition.
What settles the final pick between two acceptable items, which is where promotion and availability do their work.
| Output | What it settles | Where it goes wrong |
|---|---|---|
| Choice drivers | What settles the purchase. | Ranked from stated importance alone. |
| Decision tree | The order the choice narrows in. | Copied across markets unchanged. |
| Motive segments | Which shoppers behave alike. | Built from demographics on hand. |
| Switching triggers | What would change the choice. | Modelled from price distance. |
The four are commonly produced once, then quoted for years. A decision tree drawn in one market and applied across a region is the most frequent error in category work, and it survives because the tree looks stable while the markets underneath it move.
The driver read draws on four collected inputs, market by market.
| Input | What it answers |
|---|---|
| Receipts | What the shopper chose and what sat in the basket beside it. |
| Store captures | Which options were available at the moment of choice. |
| Geo-verified photos | The choice as it presented itself, dated and placed. |
| Stated preference | What settled the choice, and what would change it. |
The team's own numbers join separately. Segmentation files, sales history, price records, and prior category assessments connect through 250+ integrations, so recorded choice meets stated reason.
Three questions sit outside any transaction record, and each one changes the driver.
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.
One shopper supplies both the receipt and the reason, and each figure names its market and its week.
The same shopper's claim and purchase sit together, so an overstated driver is visible.
The order a choice narrows in differs by market, and each market carries its own.
Stated response measures how many shoppers a change would move ahead of the change.
Category intelligence built on a sales file can rank what was sold and struggles to explain it. Sena builds every figure on real-world signals captured when the question needs it, from the fixture photographed at the moment of choice through to the shopper stating what settled it.
The chosen item and the rest of the basket are recorded together, so a stated reason has a purchase to stand against.
Reads which options faced the shopper, off images captured in real outlets, so a choice is judged against what was available.
Signal arrives from the consumer network under explicit consent. What settled the choice is reported by the shopper who made it.
Segmentation files, sales history, price records, and prior category assessments connect over 250+ integrations, so recorded choice meets stated reason.
Each figure carries its market and its capture week, so a driver ranking opens onto the shoppers who supplied it.
Prior category assessments, segmentation files, and price histories load in, reaching back across every scanned cycle.
Four teams argue from the same driver read, and each one needs a different cut of it before they can act.
The category argument. Needs the decision tree per market.
The proposition. Needs the driver that settles the choice.
The activation. Needs the terminal step named, outlet by outlet.
The unmet need. Needs the counterfactual sized before the build.
The same paired read, run against whatever each industry treats as the deciding attribute.
The decision tree per market, with the terminal driver named.
Occasion as the first split, with format resolving below it.
Symptom and trust as drivers, read against own-label choice.
The switching trigger, read against the account already held.
One mechanism, applied per market and per category. Each step is documented, which is what carries a decision tree through a category review.
The same shoppers supply the receipt and the stated reason, so claim and purchase arrive attached.
Sena orders the drivers on recorded choice, with stated importance held beside it, so the two are comparable.
Stated response measures how many shoppers a change would move, per market.
Most category intelligence runs on a sales file with a stated read alongside, and the two describe different shoppers. Sena reports both on one.
| Capability area | Typical setup | Sena |
|---|---|---|
| The reason for a choice | Reconstructed from price and share. | Stated by the shopper who chose. |
| Claim against purchase | Two datasets, two shopper sets. | One shopper, both records. |
| Segments | Age, income, and location. | Grouped on what settled the choice. |
| The decision tree | Drawn once, applied across markets. | Held per market, from recorded choice. |
| What was available | Assumed present. | Photographed at the moment of choice. |
| A change the market omits | Beyond reach. | Sized by the stated response, before the build. |
| Evidence in a review | A commissioned conclusion. | Open any driver onto the shoppers behind it. |
Three questions where a stated reason and a recorded purchase disagree.
Hold what shoppers say against what they bought, so an overstated reason surfaces before a plan is built on it.
Read the order a choice narrows in, market by market, so five different shopper bases each get the range plan that fits.
Measure how many shoppers a change would move before anything is built.
A named driver settles why shoppers choose. The four decisions below are what a category team runs around it.
Every category decision Sena answers, in one place
What the category contains
Sets out the structure the choice drivers operate inside.
Explore product category analysis 02Where the category grows
Turns the choice drivers into a plan for growing the category.
Explore category growth strategy 03Which items the category carries
Applies the choice drivers to the range a retail category stocks.
Explore assortment optimization 04The discipline, eight steps
Puts the choice drivers to work inside the retail process.
Explore category management in retailThe walkthrough takes one category in one market, ranks its choice drivers on recorded purchase against stated importance, and draws the decision tree while the team watches.
Name the category and the markets it sells in, and we will rank the drivers on both records.