Sena reads the shelf and the shoppers in front of it from real-world signals.
Shelf space optimization decides how many facings each item receives, in which outlets, and at which height. Three properties of the usual inputs push space toward what already sold.
Allocation runs against the plan. What each store assembled differs from it, by item and by facing count, and the gap goes unmeasured between audits.
Facings follow last period's sales share. An item held at two facings sells what two facings permit, so the allocation confirms itself and the constraint reads as demand.
An empty face sells zero. The item then loses share, loses facings at the next reset, and the original cause disappears into the sales history.
Sena is the decision AI with access to real-world data. It photographs the shelf as it stood in real outlets and records what shoppers bought in front of it, then asks those shoppers what they would take when a facing is empty, so space follows demand ahead of its own history.
Images from real outlets return the facing count each item held, outlet by outlet, dated.
The stated response covers what an item would sell at more faces, which its own history leaves out.
An empty facing is recorded as an empty facing, so a stock failure reads as a stock failure.
Shelf space optimization allocates a finite number of facings across a range so the category returns more from the same fixture. The shelf is a hard constraint, so every extra facing for one item removes one from another, and the decision is comparative at all times.
A planogram is the plan, and the shelf is what the store built. The two diverge by item and by facing count, and the divergence is where a space decision loses its basis.
Allocation against an unverified plan optimizes a fixture that exists on paper.
| Output | What it settles | Where it goes wrong |
|---|---|---|
| Facing count | How many facings each item holds. | Set on the last period's share. |
| Position | Which shelf height and block. | Fixed by supplier negotiation. |
| Space to sell | Whether space matches contribution. | Backward-looking by construction. |
| Localization | Which fixture each outlet runs. | Set by store size, over shoppers. |
The four are usually settled at an annual reset and left until the next one. An item whose demand rose in month two waits ten months for the space to follow.
On-shelf availability measures whether an item was there to be bought. Three reads separate the causes.
Whether the facing was empty when shoppers were present, dated and placed.
Whether the item was in the store or off the shelf, which is a replenishment failure over a supply one.
What the shopper took instead, or whether the trip left the category.
Planogram compliance measures the distance between the plan and the fixture. Compliance is commonly self-reported or audited quarterly across a handful of outlets, which produces a number that is hard to dispute and hard to act on. Image capture at the outlet level returns the facing count and position as they stood, so compliance becomes a measured figure per store, ahead of an estimate per chain.
Four collected inputs stand behind the facing read, per outlet.
| Input | What it answers |
|---|---|
| Receipts | What shoppers bought at that fixture, dated to the day. |
| Store captures | Which items the outlet carried, and at what facing count. |
| Geo-verified photos | Every face on the fixture, counted and dated. |
| Stated preference | What a shopper takes when a facing sits empty. |
The team's own numbers join separately. Planograms, sales history, replenishment records, and space-to-sales reports connect through 250+ integrations, so the plan meets the shelf.
Three questions live beyond a planogram file, and each one changes the facing.
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.
Shelf captures and stated preference arrive from the same markets, with the outlet and the week on every figure.
Capture and stated response together show whether share reflects the shelf or the shopper.
An empty facing carries the sale it cost, dated and placed.
The same item at a different facing count elsewhere gives the counterfactual directly.
Shelf space optimization built on planograms and sales history can allocate against the plan and struggles to allocate against the shelf. Sena builds every figure from real-world signals captured from the fixture photographed in a real outlet through to the receipt showing what the shopper took.
Purchase at the fixture is recorded at the till, so a facing decision meets the demand in front of it.
Counts the facings each item held and reads its position off images captured in real stores, so compliance becomes a measured figure.
Signal arrives from the consumer network under explicit consent. What a shopper takes when a facing is empty comes from that shopper.
Planograms, sales history, replenishment records, and space-to-sales reports connect over 250+ integrations, so the plan meets the shelf.
The outlet and capture date sit on every figure, so a facing recommendation opens onto the images and purchases behind it.
Planograms, reset histories, and compliance audits from prior cycles load in, outlet by outlet.
Four teams work the same fixture, and each one needs a different cut of it before they can act.
The space plan. Needs facings set against measured demand.
The fixture in store. Needs compliance measured per outlet.
Availability. Needs empty facings dated and located.
The reset negotiation. Needs a space case the retailer accepts.
The same measured shelf, read against whatever each industry puts in front of a shopper.
Facing count against demand, per outlet, with availability separated.
Chilled against ambient placement, read by occasion.
Counter and gondola position, read against own-label blocks.
Demonstration space against the tier, read against the trade-up path.
One mechanism, applied per market and per category. Each step is documented, which is what carries a space case through a retailer reset.
Images from real outlets return the facing count, the position, and the empty gaps, dated.
Sena splits the recorded share into what the shelf permitted and what shoppers wanted, so a ceiling reads as a ceiling.
Stated demand above the current facing count sizes the case for moving space, item by item.
Most shelf space optimization runs on a planogram and a sales history, and the first describes a plan while the second describes what that plan permitted. Sena reads the shelf itself.
| Capability area | Typical setup | Sena |
|---|---|---|
| The fixture | The planogram as designed. | The shelf as built, photographed per outlet. |
| Facing count | Set on the last period's share. | Set against demand above the current ceiling. |
| Compliance | Self-reported or audited across a few outlets. | Counted from images, outlet by outlet. |
| An empty facing | A zero in the sales line. | Recorded as a gap, dated and placed. |
| Demand above the limit | Beyond reach. | Stated by shoppers, with the accepted alternative. |
| A competing fixture | Beyond reach. | The same item's facings, captured elsewhere. |
| Evidence at a reset | A space-to-sales chart. | Open any face onto the images behind it. |
Three space decisions where the shelf as built changes the answer.
Separate a ceiling effect from a demand limit so a flat share holds its face on evidence.
Count facings and positions from images, so the plan and the fixture reconcile store by store.
Attach the lost sale to an empty facing, dated and located.
A measured fixture settles how the space is split. The four decisions below are what a category team runs around it.
Every category decision Sena answers, in one place
Which items the category carries
Decides which items exist before the space read allocates them.
Page in progress 02The discipline, eight steps
Runs the process the space decision sits inside, end to end.
Page in progress 03The verdict on a period
Reports what the space read returned once the period closed.
Page in progress 04The measurement, every cycle
Reports what the space allocation returned per facing, weekly.
Page in progressThe walkthrough takes one category in one market, counts the facings each item held across a set of outlets, and separates a flat share into a shelf ceiling and a demand limit while the team watches.
Name the category and the markets it sells in, and we will count the shelf as it stands.