Why What You Bought Last Month Predicts What You’ll Buy Next
Past purchase behaviour is the single strongest predictor of future buying. Understanding this principle changes how you target, segment, and research.
If you want to know what someone will buy next month, the single most reliable predictor is what they bought last month. Not what they say they value. Not their demographic profile. Not their stated intentions. Their actual, documented purchase history. This principle is so well established in behavioural science that it barely qualifies as a finding anymore. Yet most product teams still build their customer understanding on surveys, interviews, and demographic segments rather than on the transaction data that would tell them far more.
Why Past Behaviour Predicts Future Behaviour
The academic evidence is extensive and consistent. Past behaviour is the strongest predictor of future behaviour across domains, from exercise habits to brand loyalty to subscription renewals. The reason is straightforward: most purchasing decisions are not deliberate, conscious evaluations. They are habits. People develop routines around what they buy, where they buy it, how much they spend, and how often they repurchase. These patterns are stable over time because they are driven by deeply embedded preferences, lifestyle constraints, and cognitive shortcuts that change slowly if at all.
A consumer who has bought premium coffee beans every fortnight for the past year will, with high probability, buy premium coffee beans next fortnight. Not because they have thought carefully about it, but because the decision is no longer a decision. It is a default. Understanding these defaults tells you more about a consumer’s likely future behaviour than any attitude survey could.
What Purchase Data Reveals That Surveys Miss
Surveys ask people to introspect about their preferences, and people are unreliable introspectors. They overstate their price sensitivity because it seems rational to care about value. They understate their impulse buying because it seems irrational. They claim to prioritise quality while their purchase history shows consistent selection of the cheapest option. The gap between stated and revealed preference is not a minor methodological nuisance. It is a systematic distortion that affects nearly every category.
Purchase data sidesteps this entirely. It records what happened, not what someone thinks happened or wishes had happened. From transaction records, you can extract price tolerance (the range within which a consumer actually buys), category repertoire (which brands and products they switch between), purchase frequency (how often they buy in a category), and spend allocation (what share of their budget goes to different categories). Each of these dimensions is more predictive than its survey-based equivalent.
Consider price sensitivity. A survey might tell you that 70% of consumers say price is their most important consideration. Their purchase data might show that only 25% consistently choose the cheapest option. The other 45% say they care about price but repeatedly pay premiums for convenience, brand familiarity, or habit. The survey gives you a useless number. The purchase data gives you a segmentation you can act on.
Category Loyalty and Switching Patterns
One of the most valuable signals in purchase data is switching behaviour. Within any product category, consumers fall on a spectrum from fiercely loyal to highly promiscuous. Loyal buyers repurchase the same brand or product consistently. Promiscuous buyers rotate through a repertoire of options, often driven by deals, availability, or novelty.
This distinction matters enormously for product strategy. If you are launching into a category dominated by loyal buyers, your challenge is breaking established habits. That requires a significantly better product or a compelling switching incentive. If the category is characterised by repertoire buying, consumers already rotate between options and are more open to trying something new. Your challenge is getting into the consideration set, not displacing an incumbent.
Purchase data reveals these patterns directly. You can see how often consumers switch, what triggers switches (price promotions, stockouts, new product launches), and which direction switches flow. If consumers consistently switch from Brand A to Brand B but rarely the reverse, that tells you something about relative product quality, value perception, or distribution advantage that no survey could surface as clearly.
Revealed Preference in Practice
Economists call this concept “revealed preference.” The idea, first formalised by Paul Samuelson in the 1930s, is simple: when a consumer chooses one option over available alternatives, they reveal that they prefer it, given their constraints. The beauty of revealed preference is that it requires no self-report. The choice itself is the data.
Every purchase is a small experiment in revealed preference. When someone buys a £4 sandwich instead of a £2.50 sandwich from the same shop, they have revealed that the perceived quality difference is worth at least £1.50 to them in that moment. Aggregate these micro-decisions across thousands of consumers and you build a remarkably detailed picture of how a market actually values different product attributes, without ever asking anyone a question.
This is why companies with access to large-scale transaction data have a structural advantage in understanding their markets. Retailers, payment processors, and subscription platforms sit on vast datasets of revealed preference. The insights they can extract from this data are fundamentally different in kind from what survey-based research produces, because they are grounded in behaviour rather than intention.
How This Underpins Modern Consumer Research
The predictive power of purchase history is the foundation of modern targeting, segmentation, and increasingly, synthetic consumer research. When a synthetic persona is calibrated against real purchase data, its responses are anchored in revealed behaviour rather than in the general knowledge of a language model. This is what makes behaviour-grounded synthetic research different from simply asking an AI to role-play as a consumer.
Targeting in digital advertising has moved in the same direction. Behavioural targeting, showing ads to people based on what they have bought or browsed, consistently outperforms demographic targeting. A 28-year-old woman and a 55-year-old man who both regularly buy premium running shoes are more similar, for the purposes of selling running products, than two 28-year-old women where one runs marathons and the other has never owned trainers. Purchase behaviour cuts across demographics to reveal actual preferences.
The Practical Takeaway
If your customer research is built primarily on what people say, you are working with the weaker signal. Stated preferences, survey responses, and interview quotes all have value, but they should supplement behavioural data, not substitute for it. Wherever possible, ground your understanding of your market in what consumers actually buy: how much they spend, how often they buy, which products they switch between, and what price points they consistently accept.
The most expensive mistakes in product development come from believing what consumers say they want while ignoring what they demonstrably choose. Past purchases are not a perfect predictor. People do change their habits, try new things, and respond to genuinely innovative products. But the base rate of behavioural consistency is high enough that any research methodology ignoring it is leaving its most powerful input on the table.