How closely do Coles and Woolworths price?

You are reading the September 2026 report from the archive. Go to the latest month (October 2026).

September 2026

Priced from the week of 2 Sep 2026 · 34 items × 7 capitals = 238 comparisons.

Updated 7 Oct 2026: every month was re-run after we found that a product our weekly capture had missed at one store let a dearer product stand in for it on that store's line. Where the product was in the store's previous capture and enough of the chain's other capitals still listed it, at one price, the line is now priced from it and marked in the breakdown (Basket Test rule R7 gives the conditions), and a line priced that way is set aside from the regular-price basis. Every line and store total that changed is listed, with its earlier figure, under Corrections on the Basket Test page; the counts, medians and cumulative figures on this page are those of the re-run. In September the shelf-price matching rate is unchanged at 47% (113 of 238): one more pair is identical in Sydney and one fewer in Darwin. On the regular-price basis it is 54% (107 of 198); the figure published on 15 Sep 2026 was 54% (108 of 200). That figure itself followed a re-check, on 15 Sep 2026, after we found that Woolworths member prices (a "Was $X" tag with no saving recorded) had been counted as regular prices; the first note published that day gave 53% (108 of 202).

By city

Items priced identically, out of 34, per capital. The regular-price columns count only the pairs where neither chain's price carried a recorded saving or a member-price tag that week, was carried forward from another week, or was taken from the chain's other capitals, so they are counted out of fewer pairs.

By category

Identical rate and median price gap this month, and the cumulative rate over the 13 published months (September 2025 to September 2026).

Items

Over all 91 comparisons to date (13 months × 7 capitals), the items that matched in at least nine comparisons out of ten — and the items that have not matched once to date.

Items that have not matched to date, or rarely match, can differ by assortment rather than by pricing: where the two chains' cheapest qualifying products come in different pack sizes or brands (rolled oats, sausages, tomato pasta sauce, toilet paper), the line prices differ even when the unit prices are close. Fresh produce mismatches are like-for-like (loose bananas against loose bananas).

Direction

Where the two were not identical: how many lines each chain undercut the other on, and the median gap on those lines. August 2026 in brackets.

The 6 largest Woolworths undercuts this month (median gap across the capitals where Woolworths was cheaper): Washed potatoes $2.50 (2 capitals); Greek-style natural yoghurt $2.50 (1 capital); Brown onions $1.60 (4 capitals); Vanilla ice cream $1.45 (6 capitals); Chicken breast fillets† $1.00 (2 capitals); Tasty cheese block† $0.90 (7 capitals).

† Woolworths' price on this item carried a member price or a recorded saving in at least one of the capitals counted, so the gap is promotional rather than a difference in regular prices.

September 2026 in plain language

Archive

Every published month, newest first. Each month’s full report (Markdown) and its data (JSON) are the committed files this page renders from; the per-item prices every figure is counted from are the Basket Test CSVs.

A month is published only after its report passes every check in the repository and a second, independent count agrees with it; the archive never changes a published month silently — every re-run is dated on the page.

Method

• Same basket, same stores and the same eight rules as the Basket Test — this page adds one count on top of it.
• “Identical” means the same price to the cent for the same item in the same city on the run date. “Identical” compares the price for the specified quantity: for items sold by weight, the unit price applied to the spec; for packaged items, the shelf price of the cheapest qualifying pack. Two chains can be identical on a by-weight item while their pack prices differ.
• Regular-price basis: the same comparison with every pair removed where either chain's price carried a recorded saving or a member-price tag that week, was carried forward from another week's capture, or was taken from the chain's other capitals because that store's capture missed the product — so a special on one side, a stale price, or a price not captured in that store, cannot count as the two chains pricing differently, or the same.
• Prices are the chains’ published prices for the named stores, not shelf audits. A line the Basket Test breakdown marks as carried from another week, or as priced from the chain’s other capitals, was not captured at that store that week (Basket Test rule R7): it counts on the shelf-price basis and is set aside from the regular-price basis.
• The page reports how often prices match. It does not say why.
• Every number above traces to a line in the committed monthly report; the per-item prices behind them are the CSVs published on the Basket Test page.