Basket Analysis: Which Products Are Bought Together and How to Use That Data
Basket analysis (also called market basket analysis or affinity analysis) identifies which products customers buy in the same transaction. For SMB retailers, it reveals bundling opportunities, cross-sell prompts, and store layout decisions that directly increase average transaction value.
- The Upsell Your Team Is Missing Every Day
- Three Metrics Every Basket Analysis Generates
- Five Ways to Act on Basket Analysis Findings
- Running Basket Analysis From POS Data
- Seasonal Basket Patterns and How to Use Them
The Upsell Your Team Is Missing Every Day#
A garden centre in the UK was selling compost at high volume but had no systematic process for suggesting what to buy alongside it. Their POS data showed that 68% of customers who bought compost in a given basket also bought plant feed — but only when plant feed was displayed within two metres of the compost. In sections of the store where plant feed was shelved separately in the fertiliser aisle, the co-purchase rate dropped to 22%. This proximity effect was worth £3.40 per compost transaction in additional revenue. The garden centre had been leaving this money on the table across 800 compost transactions per month because nobody had pulled the basket data and connected it to the floor plan. This is the commercial reality of basket analysis: it surfaces profitable product relationships that are invisible to the naked eye but unmistakable in the transaction data.
Three Metrics Every Basket Analysis Generates#
Basket analysis generates three key metrics for every product pair or group. Support: the percentage of all transactions that contain both products — if 15% of all your transactions include both wine and cheese, the support is 15%. Lift: how much more likely customers are to buy product B given they bought product A, compared to the base rate of buying product B at all — a lift of 2.5 means customers who buy product A are 2.5 times more likely to also buy product B than the average customer. Confidence: the percentage of transactions containing product A that also contain product B — if 60% of all wine purchases include cheese, the confidence is 60%. High-lift, high-confidence pairs are your prime bundling and cross-sell opportunities. High-support pairs tell you what to put next to each other on the shelf. Low-lift pairs that occur frequently are simply popular products that happen to both be in many baskets — no special relationship worth acting on.
Five Ways to Act on Basket Analysis Findings#
The five highest-value applications of basket analysis for SMB retailers are: product placement optimisation (place high-lift pairs adjacent to each other on the shelf or in the same category in your online store); bundle creation (package high-confidence pairs as bundles with a small price saving — customers value convenience and you increase transaction value); staff cross-sell scripts (if your team knows that 70% of customers who buy X also buy Y, they can confidently recommend Y without guessing); promotional pairing (discount product A to drive volume and let the basket data prove that it pulls product B sales without a discount); and email product recommendations (send post-purchase emails recommending the high-lift complement to whatever the customer just bought). Each of these applications requires the same underlying data — the basket analysis output — and can be implemented without any technical expertise beyond reading a product-pair table.
Running Basket Analysis From POS Data#
The raw material is your POS transaction history at line-item level — each row is one product in one transaction, with a transaction ID that groups items bought together. You need at least three to six months of data to identify reliable patterns; anything less may reflect seasonal anomalies rather than genuine affinities. The calculation itself is straightforward in concept but tedious in practice in a spreadsheet. AskBiz runs basket analysis automatically on your POS transaction data and outputs a ranked product-pair table showing support, lift, and confidence for every pair that appears in more than 1% of transactions. The most interesting output is usually the high-lift pairs with moderate support — these are non-obvious product relationships that your team would never have guessed but that appear consistently in your customer behaviour data.
Seasonal Basket Patterns and How to Use Them#
Basket patterns change by season and your analysis should account for this. A home goods retailer in Singapore found that their Q4 basket data showed completely different product affinities from Q2 — gift-wrapping supplies became the highest-lift partner for almost every product category during November and December, whereas in Q2 they barely appeared in combined transactions. Running basket analysis quarterly rather than annually reveals these seasonal shifts and allows you to adjust shelf placement, staff training, and promotional bundles accordingly. The most actionable approach is to identify your top ten high-lift pairs for each calendar quarter, brief your team on the cross-sell script for each quarter's top pairs, and update your online recommendation engine (if you have one) seasonally. This quarterly refresh ensures your cross-sell strategy reflects current customer behaviour rather than a stale annual analysis.
Basket Analysis for Service Businesses#
Basket analysis is usually discussed in the context of product retail, but it applies equally to service businesses. A beauty salon in Edinburgh analysed which treatments were most commonly booked in the same appointment or the same month. They found that clients who got gel nails were highly likely to also book a gel removal appointment 3-4 weeks later — 78% confidence. This sounds obvious in retrospect but they had never systematically offered the removal booking at the time of the gel application appointment. When they started doing so, their same-client repeat booking rate increased 25% and their monthly revenue per client grew significantly. The same logic applies to any service business: which services cluster together? Which first-service triggers a predictable second-service need? Answering these questions from booking data generates the same cross-sell insights as product basket analysis.
Avoiding the Basket Analysis Traps#
Two common traps undermine basket analysis findings. First, confusing correlation with causation. If customers who buy luxury items frequently buy other luxury items in the same transaction, this reflects a customer type (high-spending) rather than a genuine product affinity. Cross-selling in this context adds no value — these customers would have bought both items anyway. Second, optimising for existing patterns rather than identifying complementary gaps. Basket analysis shows what customers are already buying together; it does not show what they might buy together if you stocked the right complementary product. If 60% of customers buying running shoes also buy socks, you should definitely cross-sell socks aggressively. But you should also ask: are 40% of those customers not buying socks because you are out of stock, because your selection is too limited, or because they bought socks last month? The data tells you what is happening; your judgement is still needed to interpret why.
People also ask
What is basket analysis in retail?
Basket analysis generates three key metrics for every product pair or group. Support: the percentage of all transactions that contain both products — if 15% of all your transactions include both wine and cheese, the support is 15%.
How do I find which products are bought together in my store?
The five highest-value applications of basket analysis for SMB retailers are: product placement optimisation (place high-lift pairs adjacent to each other on the shelf or in the same category in your online store); bundle creation (package high-confidence pairs as bundles with a…
What is a good lift score in basket analysis?
The raw material is your POS transaction history at line-item level — each row is one product in one transaction, with a transaction ID that groups items bought together.
How does basket analysis help with product placement?
Basket patterns change by season and your analysis should account for this. A home goods retailer in Singapore found that their Q4 basket data showed completely different product affinities from Q2 — gift-wrapping supplies became the highest-lift partner for almost every product…
Can basket analysis work for a service business?
Basket analysis is usually discussed in the context of product retail, but it applies equally to service businesses. A beauty salon in Edinburgh analysed which treatments were most commonly booked in the same appointment or the same month.
Our team combines expertise in data analytics, SME strategy, and AI tools to produce practical guides that help founders and operators make better business decisions.
AskBiz runs basket analysis automatically from your POS transaction history. Try free at askbiz.co
AskBiz connects to your existing tools and surfaces insights like these automatically — no spreadsheets, no analysts, no waiting.
Connects to Shopify, Xero, Amazon, QuickBooks, Stripe & more in minutes