As of November 7, 2026, the most useful retail disclosure of the week is not a sales number. Assaí, Brazil's largest cash-and-carry operator, reported third-quarter results on November 5 with adjusted recurring net income of R$248 million, up 55.2%, while same-store sales moved just 0.1%. What sits underneath that gap is a store network being rewired: eleven pharmacy units opened in São Paulo between July and November, a target of 25 by December, last-mile delivery sales up 204% across 104 stores, and a private-label range moving from 140 SKUs toward 250. For anyone tracking on-shelf availability, the pharmacy aisle is where the next stockout problem will surface.
The eleven units that changed the basket
The pharmacy move is usually read as diversification. Operationally it is closer to a stress test. A club format is built around pallet-sized flows, a narrow assortment and predictable replenishment. Pharmacy adds small packs, expiry dates, regulated categories and a much higher SKU count per square metre. Those four characteristics are exactly the ones that break shelf-level accuracy first, because a mis-scanned box of 20 units matters far more than a missing pallet.
Assaí is not running this experiment in a quiet quarter. Gross revenue reached R$21.3 billion, up 2.7%, gross margin widened 45 basis points to 17.2%, and selling, general and administrative expenses grew 4.2%, below the 4.4% twelve-month inflation print. Nine store openings accounted for 3.6 percentage points of that expense growth, against 0.6 points from same-store operations. In other words, the cost of extending the network is already visible in the P&L before the pharmacy units have a full quarter behind them.
Where a club format breaks first
Store-level availability in a cash-and-carry network fails differently from a supermarket. The customer is often a small retailer buying for resale, which means an empty facing is not a lost impulse purchase but a broken supply promise to that reseller's own customers. Adding a pharmacy counter to the same building raises the number of independent replenishment rhythms the store has to run in parallel.
Three formats, three failure modes
| Store role | What gets added | Replenishment rhythm | First failure to watch |
|---|---|---|---|
| Club floor | Pallet and case packs, narrow assortment | Weekly, high volume per SKU | Pallet ends empty while back stock holds units |
| Pharmacy counter | Small packs, expiry-dated, regulated | Two to three times a week | Expiry-driven delisting that reads as a stockout |
| Pickup and last-mile node | Online orders drawn from the same shelf | Daily, order-driven | Online orders reserved against stock the shelf no longer has |
The third row is where the 204% figure becomes interesting. When last-mile volume grows that fast across 104 stores, the same physical shelf is suddenly serving two demand streams with different priorities. A store that reports stock correctly at 8am can be wrong by 11am if pickers and walk-in customers draw from the same facing without a reservation layer. Retailers that scale on-demand delivery usually discover this in the second month, not the first, because early volumes are too small to expose the conflict.
What to measure before adding more stores
The instinct after a 204% quarter is to add stores. The better sequence is to prove that the existing 104 can hold availability under load. Two numbers carry most of the signal: the share of online orders cancelled after picking started, and the difference between system stock and shelf stock measured in a sample of fast-moving SKUs. If the first rises while the second widens, adding stores multiplies the problem rather than the revenue.
Reading the 204% last-mile number
Last-mile growth of 204% is a percentage off a small base, and Assaí does not disclose the absolute value. That makes it a directional signal rather than a size signal. The directional message is still important: a wholesale operator whose customers historically drove to a warehouse is now willing to send goods to them, and is using the existing store network as the origin point instead of building separate fulfilment centres.
That choice has a cost structure attached. Fulfilment from stores consumes labour hours that were previously spent on the floor, and it consumes shelf stock that walk-in customers can see. Industry benchmarks put delivery at roughly 10% to 13% of sales and store picking plus related operations at 11% to 14% for omnichannel grocery, which is why operators that scale store-based delivery usually end up revisiting minimum order values and delivery windows before they revisit marketing.
| Signal | Q3 reading | Why it matters for availability |
|---|---|---|
| Same-store sales | +0.1% | Volume is flat, so any stockout is a direct loss rather than a timing shift |
| October same-store sales | +3.9%, best in 15 months | Campaign-led lift concentrates demand on fewer SKUs |
| Customer traffic | +2.1%, above 40 million a month | More visits to the same shelf space raises picking conflicts |
| Private-label SKUs | 140, target 250 | Each new SKU competes for the same facings |
The private-label variable
A jump from 140 to 250 own-brand SKUs inside a single year changes shelf arithmetic. Own brands are usually placed next to the national brand they are meant to replace, so each launch forces a re-plan of the facing rather than a simple addition. When the launches land in the same quarter as a pharmacy rollout and a delivery expansion, the store team is re-planning three assortments at once, and planogram compliance is the first thing to slip.
None of this argues against the strategy. Assaí cut leverage to 2.15x from 3.03x, generated R$3.3 billion of operating cash flow over twelve months, and kept expense growth below inflation while opening stores. A stronger balance sheet is exactly what funds the shelf-level work that a pharmacy and delivery rollout demands. The risk is sequencing: if availability data is still reconciled monthly, the network will be adding nodes faster than it can see them.
FAQ
Why does a pharmacy rollout affect on-shelf availability in the grocery aisles?
Because both assortments compete for the same labour hours, the same back room and the same replenishment windows. Pharmacy adds expiry-dated small packs that need more frequent counting, which pulls store time away from the pallet flows the club format depends on.
Is the 204% last-mile growth a reliable indicator of demand?
It shows direction, not size. Assaí does not disclose the absolute value, and part of the growth may come from a small base, so it should be read alongside order value, cancellation rate and picking time per store.
What is the single best metric for store-level availability?
The gap between system stock and shelf stock on a fixed sample of fast-moving SKUs, measured more than once a day. It catches the failures that a monthly reconciliation cannot see, especially when online orders and walk-in customers draw from the same facing.
Should a retailer add delivery stores before fixing stock accuracy?
No. Adding stores multiplies whichever behaviour already exists, so a widening stock gap becomes a larger stock gap. Prove the existing network first, then expand.
