Market Basket & Co-Purchase Analysis Narrator

Example output

Basket co-purchase read — Bramblefield marinara, Heirloom Grocer, 8-week loyalty data

Headline: this basket is telling a "weeknight dinner build" story, not a "condiment" story. Four items appear across a huge share of your marinara baskets, and they stack into a coherent meal rather than four unrelated impulse adds — that is the strongest signal in this data.

Ranked pairings

1. Dry pasta — 61% of baskets
This is close to a default pairing, which is expected but still worth stating plainly: more than six in ten Bramblefield marinara trips already include pasta in the same basket. The shopper mission is straightforward meal assembly, not discovery. Because the rate is this high, the marginal upside of a chip-and-salsa-style co-display is lower than it looks — most of this pairing is already happening without your help. The bigger opportunity is basket size, not basket creation: a bundle here should aim to add a third item, not just confirm a pairing shoppers already complete on their own.

2. Garlic bread — 34% of baskets
A third of marinara baskets add garlic bread, which is a strong enough rate to treat as a real pairing rather than noise, but well below the pasta rate — meaning two-thirds of your shoppers are completing a marinara-and-pasta dinner without a bread side. That gap is the opportunity: a garlic bread cross-merchandising push isn't chasing an already-saturated pairing, it's recruiting the majority of baskets that currently skip it. Treat 34% as the baseline to beat, not the ceiling.

3. Parmesan — 29% of baskets
Close behind garlic bread and likely correlated with it in the same baskets — worth checking in your raw data whether parmesan and garlic bread tend to appear together or represent two distinct shopper segments (one finishing with cheese, one finishing with bread). This distinction changes whether you build one three-way bundle or two separate two-way pairings. Flagging this as a question the pasted summary can't answer on its own; it needs a basket-level cross-tab, not just individual pair rates.

4. Ground beef — 22% of baskets
The smallest of the four, but the most meal-transforming — beef converts a marinara-and-pasta side dish into a full meat-sauce dinner. Lower incidence likely reflects a real behavioral split between vegetarian-leaning marinara occasions and meat-sauce occasions rather than pure under-merchandising, so treat this pairing as a distinct occasion to target on its own rather than folding it into a single generic bundle. Assumption to validate: whether the 22% is stable across weeks or concentrated in one segment of shoppers.

Merchandising and bundling implications

- Lead cross-merchandising investment with garlic bread, not pasta. Pasta's pairing rate is already near its ceiling; garlic bread has the largest gap between current pairing rate and plausible upside.
- Consider a three-item "dinner build" secondary display — marinara, pasta, garlic bread — sized to the 34% rate as a realistic lift target, not the 61% pasta rate.
- Treat ground beef as a separate "hearty dinner" occasion, potentially a distinct end-cap or shelf-talker theme rather than folded into the default bundle, since it likely reflects a different shopper intent from the start of the trip.
- Hold parmesan for a follow-up cut of the data before building a display around it — its overlap with garlic bread needs to be confirmed first.

What this data can't tell us

Whether these pairings are incremental (the shopper wouldn't have bought pasta anyway) or simply confirm a trip that was already planned. Basket co-occurrence shows what sits in the cart together, not causality. A short in-store test — running the garlic bread cross-merchandise for four weeks and comparing the pairing rate against this baseline — would confirm whether display actually lifts the rate or just makes visible what shoppers were already doing.

Want me to draft the sell-in case for the garlic bread secondary display using this data?