We Priced the Same 727 Recipes at Two Stores. Here Is the Gap

CentsPerServing Team · July 25, 2026 · Original data

What we compared

We took every recipe in our corpus that priced cleanly at both of two Kroger stores in the same metro area - one downtown, one in a suburb - and compared them line by line. That is 727 recipes where every non-pantry ingredient matched a real product with a real shelf price at both stores, using each store's own catalog and package sizes.

Each store gets its own pricing run. The matcher picks the cheapest suitable product from that store's shelf for each ingredient, then rounds package counts up the way a shopper actually buys. Pantry staples like salt and oil are excluded. If an ingredient failed to price at either store, the recipe was dropped from this comparison entirely rather than half-counted. Prices were captured in July 2026. Our full cost distribution post covers how the pricing itself works.

The headline: 15 cents a serving

The median recipe cost $0.15 more per serving at one store than the other. On the paired set of 727 recipes, the median dinner ran $4.71 per serving at the downtown store and $4.70 at the suburban one. That is a rounding error at the level of a single meal, and it is the single most useful thing in this data.

The medians differ slightly from the $4.62 in our launch post because that figure covers the 712 USDA recipe cards priced at one store. This post uses the 727 recipes that priced cleanly at both, which is a slightly different set.

The spread, recipe by recipe

Medians hide tails, so here is the whole distribution of per-serving differences. A third of recipes came out identical to the cent, roughly four in five landed within fifty cents, and a small tail swung by dollars. That tail is where the interesting behavior lives.

Source: CentsPerServing corpus pricing, Kroger stores 01400513 and 01400421, July 2026. 727 recipes fully priced at both. Shares rounded.

The 90th percentile difference was $0.88 per serving and the 95th was $1.15. The largest was $6.03. So the honest summary is not "stores cost the same" or "stores differ a lot" - it is that most recipes are flat and a small minority are wildly store-dependent.

Identical products cost identical money

Here is the finding that reframes everything else. Our matcher picked 778 products that appear in baskets at both stores. Of those, 99.2% carried the exact same shelf price at both locations. Only six products were priced differently at all, and every one of them was a perishable.

Source: CentsPerServing corpus pricing, July 2026. The only six of 778 shared products with a price difference between the two stores.

Within one banner in one metro, shelf pricing is essentially centralized. Milk, eggs, and bananas get local treatment. Everything else - the canned tomatoes, the rice, the ground turkey - rings up the same on both sides of town.

The gap is assortment, not pricing

If identical products cost identical money, where does any gap come from? Assortment. Of 6,470 comparable ingredient lines, 83.8% resolved to the same product at the same price, 0.5% to the same product at a different price, and 15.7% to a different product entirely. That last bucket is the whole story.

Store catalogs are not clones. In this data, 112 products were picked only at the downtown store and 92 only at the suburban one. When a store does not carry the cheapest option, the next-cheapest suitable product takes its place, and that substitution can cost real money.

Divergence is also highly category-specific. Shelf-stable goods are nearly interchangeable between stores; fresh, refrigerated, and bakery items are not. The rate below is how often the two stores' cheapest suitable pick was a different product for the same ingredient.

Source: CentsPerServing corpus pricing, July 2026. 6,470 ingredient lines compared across 727 recipes at two Kroger stores.

This is why a lentil soup prices the same anywhere and a chicken dish does not. The more of your basket sits in the freezer aisle, the can aisle, or the rice shelf, the less your store choice matters.

One ingredient usually explains the whole gap

We looked at the 55 recipes that differed by $1 or more per serving and asked how concentrated the difference was. The answer: extremely. A single ingredient line accounted for a median 86% of the entire dollar gap, and in 75% of those recipes one line explained at least 70% of it.

The repeat offenders were eggs, potatoes, chicken breast, and lettuce. Potatoes are the cleanest example: the suburban store had a 5 lb bag of russets at $3.49, while at the downtown store the cheapest suitable bag was Kroger gold potatoes at $7.49. Same recipe, same quantity, $4 apart on one line.

Eggs behaved the same way. On several recipes the downtown store's cheapest match was a 12-count of pasture-raised eggs at $7.99, while the suburban store had an 18-count store-brand carton at $2.19. That is not a price difference on one product - it is a different product, chosen because it was the cheapest one on that shelf.

The worst case: $6.03 a serving

The biggest gap in the whole set was a four-serving recipe of fiesta lettuce wraps and pepper boats: $18.78 per serving at one store, $12.75 at the other. Nine of its twelve ingredient lines priced identically to the cent. Two lines did nearly all the damage.

The recipe needs a lot of lettuce, so the package math calls for eight units. At the suburban store that was wrapped iceberg at $2.29, or $18.32. At the downtown store the cheapest suitable pick was romaine hearts at $4.79, or $38.32. Instant brown rice added another $3.10, and tilapia fillets differed by $0.49. Nothing else moved at all.

Note what this means for a real shopper: a person standing in that produce section would probably just buy whatever lettuce was there. The gap is real if you shop cheapest-suitable at each store, but it is also the most substitutable kind of gap there is.

Expensive recipes swing more than cheap ones

Store sensitivity scales with recipe cost. Cheap recipes are built from the categories that barely vary; expensive recipes lean on the ones that do. Splitting the 727 recipes into cost tiers makes the pattern obvious.

Source: CentsPerServing corpus pricing, July 2026. Gap is the absolute per-serving difference between the two stores.

The budget tier is essentially store-proof. If you cook from the under-$2-per-serving tier of our data, the median difference between these two stores is a nickel a plate. Cooking gets store-sensitive at the same point it gets expensive.

So does it matter where you shop?

Within one chain in one metro, for ordinary cooking: not much. At the median gap of $0.15 per serving, five dinners for four people - twenty servings - differ by about $3 between these two stores. Across all 727 baskets, the suburban store totaled $20,496 against the downtown store's $20,929, a 2.1% edge.

Direction was not consistent either. The suburban store was cheaper on 298 recipes, the downtown store on 181, and 248 tied exactly. There is no single "cheap store" here, just a slight tilt and a lot of noise.

What does move money is the recipe you pick, which our data shows swings dollars per serving rather than cents. Store choice is a rounding adjustment on top of that decision. Understanding cost per serving is worth far more than picking a banner.

When store choice genuinely matters

Three situations flip the conclusion. First, specialty ingredients: 21% of recipes differed by more than 10% of their own cost, and those are concentrated in dishes needing an item only one store stocks cheaply. Second, high-cost recipes, where the 90th percentile gap reaches $1.31 per serving.

Third, and most important, this comparison is within one chain. Two banners with different private-label programs and different sourcing would not share a centralized price file, and the identical-price finding above would not hold. This post measures store-to-store variation, not chain-to-chain variation.

The practical rule: for staple cooking, shop wherever is convenient. For a recipe with an unusual ingredient, price the basket before you drive.

Check the gap at your own stores

Everything above came from running the same pricing engine against two store IDs. You can do the same thing with any recipe: paste it into the pricer, change the store, and read the two numbers. The line-by-line breakdown shows exactly which ingredient moved, which is the part that actually tells you something. The same goes for a full week of dinners priced against your store.

More posts in the Compare series dig into the other half of this question - store brands, delivery markups, and fulfillment costs, which in our data move a basket more than location does. The whole blog publishes from priced baskets, never estimates.

FAQ

Data: CentsPerServing corpus pricing runs, July 2026 - 727 recipes fully priced at both Kroger store 01400513 (downtown, ZIP 45202) and store 01400421 (suburban, ZIP 45227), 6,470 comparable ingredient lines, item-level shelf prices with package-count math. All figures in this article are our own first-party measurements; no external estimates are used. Store prices change without notice and baskets re-price over time. This article reports our own measurements and is not financial advice.