Sena draws the category from the basket with real-world signals, so the category matches the set shoppers treat as one.
Product category analysis answers what a category contains, how it divides into segments, and which items compete with which. Three properties of the data most analyses run on set that structure before the work begins.
Category lines come from the reporting hierarchy, which was built for accounting and stock control. The analysis then measures a set assembled for a different purpose.
Which items compete is commonly inferred from price distance and pack format. Two items at the same price meet different needs, and two items at different prices are frequently direct substitutes.
Segmentation follows product attributes recorded in the item master: flavor, size, and format. Shoppers segment by occasion and by who the purchase is for, and those fields sit in zero item masters.
Sena is the decision AI with access to real-world data. It records every item a shopper bought on the same trip, across the whole market, and asks those shoppers what they would swap for what, so a category structure comes from the shoppers who buy it.
Items appearing together and items appearing as each other's replacement draw the category line.
Recorded swaps show which items compete, including pairs a price model would separate.
Stated purpose and basket context group items by the need they met.
Product category analysis establishes what a category is made of. It sets the boundary, divides the interior into segments, and maps which items compete with which. Everything downstream depends on it: a range decision, a space plan, and a share figure all assume a category definition, and most inherit one and leave it untested.
Category analytics is the continuous measurement of a category, run off connected systems every cycle. Product category analysis is the discrete exercise of answering one structural question about that category.
Category performance analysis is the verdict on a single closed period.
| Output | What it settles | Where it goes wrong |
|---|---|---|
| Boundary | Which items belong to the category. | Inherited from the reporting hierarchy. |
| Segments | How the interior divides. | Built from item attributes, over occasions. |
| Substitution | Which items compete. | Inferred from price distance. |
| Gaps | Which needs stay unmet. | Read against listings, over demand. |
The four are usually settled in one workshop and then treated as fixed for years. A category boundary drawn once and inherited thereafter is the single most consequential unexamined assumption in category work.
Substitution is the working core of the analysis, and it is measurable in three ways that disagree.
Items near each other on price are assumed to compete. Cheap to compute and frequently wrong.
Items appearing in the same basket are complements, and items appearing in alternate baskets are substitutes. Read from recorded trips.
What a shopper takes when a first choice is absent and the point at which the trip is abandoned. The only source for items the range has yet to carry.
Those four outputs rest on four collected inputs, per market.
| Input | What it answers |
|---|---|
| Receipts | Which items shared a trip and which replaced one another, dated. |
| Store captures | Which items sat in the same fixture, outlet by outlet. |
| Geo-verified photos | Which items shared a fixture, captured and dated. |
| Stated preference | What a shopper swaps for what, and when the trip ends. |
The team’s own numbers join separately. Item masters, sales history, planograms and segmentation files connect through 250+ integrations, so the recorded hierarchy meets the observed one.
Three questions live beyond an item master, and each changes the structure.
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.
Recorded trips and stated swaps reach Sena from the same shoppers, each figure carrying its market and its week.
Recorded swaps show whether the current definition matches how shoppers buy.
Basket context groups items by the need met, which frequently cuts across format.
The same category divides differently across borders, and each division carries its own set.
Product category analysis built on an item master and a sales file can describe the products precisely and struggles to draw the category. Sena captures real-world signals from the fixture photographed in a real outlet through to the receipt showing what shared the trip.
Every item on a trip is recorded together, so complements separate from substitutes in the basket itself.
Reads which items sat in the same fixture, off images captured in real stores, so the physical grouping meets the recorded one.
Signal arrives from the consumer network under explicit consent. What a shopper swaps for what is reported by that shopper.
Item masters, sales history, planograms, and segmentation files connect over 250+ integrations, so the recorded hierarchy meets the observed one.
Market and capture week travel with each figure, so a category boundary opens onto the trips that drew it.
Item masters, segmentation records, and planograms from prior cycles load in, reaching as far back as the category has been scanned.
Four teams work from the same structure, and each one needs a different cut of it before they can act.
The definition. Needs a boundary drawn on shopper evidence.
The competitive set. Needs the rivals shoppers swap to.
The gap. Needs unmet needs read against demand, over listings.
The retailer case. Needs a structure the buyer recognizes.
The same structural read, drawn against whatever each industry treats as one need.
Boundaries and segments drawn from the basket, per market.
In-home against out-of-home sets, split by occasion.
Symptom-led grouping, read against own-label substitution.
Tier and use-case sets, read against the trade-up path.
One mechanism, applied per market and per category. Each step is documented, which is what carries a category definition through a retailer review.
Receipts return every item on a trip, so items sharing a basket separate from items replacing one another.
Sena sets the boundary and the segments from that evidence, per market, so the structure matches observed buying.
Stated swaps cover items the range has yet to carry, so a gap reads against demand.
Most product category analysis runs on an item master and a sales file, which describe the products in detail and hold zero rows on how shoppers group them. Sena reads the grouping itself.
| Capability area | Typical setup | Sena |
|---|---|---|
| The boundary | Inherited from the reporting hierarchy. | Drawn from items shoppers swap between. |
| Segments | Built from item attributes. | Built from the occasion and the buyer. |
| Substitution | Inferred from price distance. | Recorded at the till, pair by pair. |
| An item outside the range | Absent. | The volume it already takes, stated and recorded. |
| Physical grouping | Taken from the planogram. | Photographed in the outlet as it stood. |
| Coverage | The accounts that report. | 190+ countries, per outlet type. |
| Evidence in a review | A workshop output. | Open any boundary onto the trips behind it. |
Three structural questions where recorded trips change the answer.
Test the current definition against what shoppers swap between, market by market.
Read recorded swaps to find the rivals a price model separates and shoppers treat as equivalent.
Read demand against listings to locate the segment the category serves thinly.
A tested structure sets what the category contains. The four decisions below are what a category team runs around it.
Every category decision Sena answers, in one place
Why the shopper chooses
Explains why a shopper moves between the items mapped here.
Explore category intelligence 02The measurement, every cycle
Runs each cycle, where this page answers one structural question.
Explore category analytics 03Which items the category carries
Acts on the substitution map by adding and cutting listings.
Explore assortment optimization 04The verdict on a period
Reports whether the category gained or lost over the period.
Explore category performance analysisThe walkthrough takes one category in one market, tests its current boundary against recorded swaps, and names the competitive set the price model separated while the team watches.
Name the category and the markets it sells in, and we will draw it from the basket.