Photo by Khwanchai Phanthong on Pexels. Grocery delivery at a doorway.
Imagine adding a loaf, apples and washing-up liquid to an online grocery order. The basket gets bigger; the van still stops at the same door. That looks like a useful sale for the supermarket. Whether it is a better sale depends on the goods margin and the extra work required to deliver it.
Online growth is visible in recent supermarket disclosures. The financial question sits behind the checkout: how much activity can an existing picking operation and delivery route absorb before the next order needs more capacity?
Fact: three disclosures, three different measures
Tesco (LSE: TSCO) reported UK online sales growth of 8.4% in its 8 October interim results, covering the 26 weeks ended 29 August 2026. It also reported an approximately 10% recent increase in available UK delivery slots. Neither figure is a standalone online operating margin.
J Sainsbury (LSE: SBRY) reported Groceries Online sales growth of 12.5% for the 16 weeks to 20 June 2026. Its definition includes Sainsburys.co.uk and OnDemand sales serviced by supermarket and convenience locations. That combined channel cannot be treated as a measure of the large weekly delivery alone.
Ocado Group (LSE: OCDO) provides a different view in its 2026 half-year results. It reports Ocado Retail orders up 13%, fulfilment-centre costs including labour at 5.7% of sales, and an EBITDA margin excluding Hatfield fees of 5.0%. EBITDA is earnings before interest, tax, depreciation and amortisation; it leaves out costs relevant to the final return. Ocado Retail is reported as an associate, so its grocery sales are not interchangeable with Ocado Group revenue.
These disclosures demonstrate activity and, in Ocado Retail’s case, defined operating measures. Their periods, business boundaries and accounting definitions differ. Ranking the three businesses by those percentages would turn unlike measurements into a false comparison.
Interpretation: one basket has several costs
The useful unit is a completed order. Begin with what remains from selling the goods after their purchase cost. Then consider picking, packing, delivery, customer support and refunds, using consistent definitions. A delivery fee may help pay for the service, but its presence does not establish that it covers the cost.
A larger basket can spread the cost of stopping at a door across more sales. Yet more goods can also take longer to pick, fill more containers or require separate temperature handling. Basket value alone does not describe that workload. Price rises can enlarge the receipt without adding an item.
There is a similar distinction in the delivery route. An extra nearby stop within existing paid hours has different economics from an order that requires another shift. A retailer can improve average utilisation while some individual orders remain expensive. It can also accept a weaker initial order in pursuit of a repeat customer. That strategy needs evidence of subsequent purchasing, rather than an assumed retention rate.
Speculation: a fictional order makes the trade-off visible
Consider an invented retailer selling an £80 basket. Assume the goods leave £20 after their purchase cost: a 25% goods margin. Assume picking, packing and delivery together cost £12 per order. The simplified remainder is £8 before overheads and other excluded costs. It is not operating profit.
Hypothetical arithmetic
Same basket, different service bill
Goods margin
£80 basket × 25%
Remainder
£20 less £12 service cost
Costlier delivery
£20 less £15 service cost
Keep the basket and goods margin unchanged, but raise the service cost to £15. Only £5 remains. That is £3 less, or 37.5% below the first remainder, despite identical sales. No real retailer’s order margin is being estimated here.
Alternatively, let the basket rise to £100 at the same assumed goods-margin rate. It leaves £25 before service costs. At £15 to fulfil and deliver, the remainder becomes £10. Extra spending has helped, but the result depends on both assumptions holding. Discounts, product mix or additional handling could change it.
The example excludes delivery fees, VAT, refunds, marketing, central overheads, property and equipment costs, financing and tax. It isolates a mechanism; it cannot value a delivery business or substitute for reported accounts.
Who could benefit?
Interpretation: Tesco and Sainsbury’s could retain more from incremental online orders over the next several quarters if existing capacity can serve them efficiently and customers return without persistent extra discounts. Their disclosed channel growth establishes demand, not the size of that incremental financial benefit. Group-level materiality depends on online sales mix and costs that these headline growth figures do not isolate.
Ocado Group has a related but distinct opportunity through fulfilment technology and logistics. Its half-year disclosure separates Technology Solutions and Logistics from the associate’s retail performance. Better throughput could support the service proposition and partner economics over several years. Converting that into a group benefit requires retained contracts, fees and controlled spending; an efficient grocery basket does not automatically produce a technology contract.
Who’s at risk?
Interpretation: The same three issuers face execution risk. Tesco’s extra slots need customers and economical service. Sainsbury’s combined online measure needs a cost account that distinguishes different shopping missions. Ocado Group needs partner demand and spending discipline alongside operational efficiency. More orders can bring more work before enough revenue remains to pay for the next investment.
Speculation: If a retailer buys growth with discounts while delivery costs rise, sales could improve and the contribution per order weaken. If repeat orders fill spare capacity at sustainable prices, the opposite could happen. Neither outcome is established by the reported growth rates. Near-term service costs may appear within a quarter; capacity investment and contract returns take longer. These mechanisms make no prediction about share prices.
What would change the assessment?
The strongest evidence would connect consistently defined order volumes and basket values to fulfilment costs, repeat purchasing and cash needed for capacity. Lower cost per order with reliable service would support the case for useful growth. More capacity alongside persistent underutilisation would weaken it.
Sources were checked on 16 October 2026. This analysis does not have route-level cost records or comparable online profit figures for all three businesses. Sainsbury’s has announced interim results for 22 October; those results will provide a further test of the older trading evidence used here.
The customer gets the groceries. The financial test is whether serving that basket leaves enough to support the operation and its next investment.
This is informational editorial analysis, not investment advice or a recommendation concerning any security.