Habits and routines in purchasing decisions

A habit is a link between a context and a response

The useful definition is narrow. A habit is not a preference, a loyalty or a strong feeling about a brand. It is an association between a situation and an action, learned by repetition, that runs without a decision being made.

That definition has a consequence most marketing writing skips. If the habit lives in the context, then the context is what you have to reach. Changing someone's mind about your product does very little when their behaviour is attached to a shelf position, a time of day or a saved payment method.

My position is that habit is the least useful frame for most small businesses selling considered services, and the most useful frame for understanding why your marketing is not working when you sell something people buy repeatedly. Those are different situations and the advice for them is not the same.

What the research supports

Wendy Wood and her colleagues ran diary studies asking people to report what they were doing through the day. Roughly 43% of everyday actions were performed almost daily in the same context, usually while the person was thinking about something unrelated. That is the strongest evidence for the scale of habitual behaviour, and it is about actions in general rather than purchases specifically.

Wood and David Neal's later theoretical work established the model behind the definition above: habits as context-response associations, which is why stable contexts predict repeated behaviour better than stated intentions do.

On how long habits take to form, the number worth knowing comes from Phillippa Lally and colleagues. In a study of 96 volunteers adopting a daily eating, drinking or activity behaviour, the median time to reach automaticity was 66 days, with a range from 18 to 254.

Two things follow. The 21-day rule that circulates in marketing has no support. And the range is so wide that any specific promise about how long it takes to build a customer habit is unsupportable.

Note what I am not saying. The claim that most purchasing decisions are subconscious, which an earlier version of this article made, is not a finding from this literature. The 43% figure describes daily actions in stable contexts. It is not a statistic about buying decisions, and it should not be quoted as one.

The habit loop is a simplification

Most marketing treatments of habit use the cue, routine, reward loop popularised by Charles Duhigg's 'The Power of Habit' in 2012. It is a reasonable teaching device and it misleads in one specific way.

The reward is doing less work than the loop implies. In the academic model, reward matters while a habit is forming and matters much less once it has formed. A formed habit runs off the cue whether or not the reward is still satisfying, which is why people keep buying products they no longer particularly like.

The practical error this produces is loyalty schemes designed as permanent reward machines. Rewards help while the behaviour is being learned. After that, the money is often paying for behaviour that would have happened anyway.

The cue also gets misread. Marketers reach for the advert as the cue. In the research, the cue is usually the ordinary context: the time, the place, the preceding action, the state of the cupboard. An advert can be a cue, but it competes with a context that is present every single time and costs nothing.

The transition window

Here is the part that is both well evidenced and rarely used, and it is where I would concentrate effort.

If habits are attached to contexts, then habits become changeable when the context breaks. Bas Verplanken's habit discontinuity hypothesis says behaviour change interventions work better during life changes. In a field experiment run with Deborah Roy in Peterborough, an intervention promoting sustainable behaviours produced a significant improvement only among households that had moved within the previous 3 months. Verplanken, Roy and Lorraine Whitmarsh summarise the hypothesis and the evidence in Habit discontinuities as vehicles for behaviour change.

For a business, that reframes targeting. You are not looking for people who want your category. You are looking for people whose context has just broken.

The transition	What breaks with it	Roughly how long the window lasts	What to do about it
Moving house or premises	Supplier lists, routes, local providers, delivery arrangements	Weeks to a few months	Be findable locally and be easy to start with, not cheapest
A change of role or a new hire in the buying seat	Inherited suppliers stop being defended by the person who chose them	The first few months in post	Address the new person's need to make an early visible improvement
A system, platform or tool migration	Everything attached to the old workflow, including your competitor	Until the new setup settles	Be present in the migration conversation, not in the category conversation
Growth past a threshold	Processes that worked at the previous size	Until something breaks badly enough to force a fix	Describe the stage, not the service
A regulatory or reporting change	The assumption that last year's approach still applies	The run-up to the deadline	Be early with a plain explanation of what changed
A bad experience with the incumbent	Trust in the current arrangement, temporarily	Days to weeks, and it closes fast	Be easy to find and quick to respond in that moment

The last row is worth dwelling on. It is the shortest window and the one small businesses are most likely to catch, because it does not require data about someone's life. It requires being findable and responsive when somebody has just been let down, which is a content and operations question rather than a targeting one.

When habit is the wrong frame

Most of this research concerns frequently repeated, low-cost, low-consequence behaviours. Toothpaste, coffee, a commuting route.

A professional service bought once a year, or once ever, is not a habit in this sense, and treating it as one leads to bad decisions. Nobody develops an automatic context-response association with your consultancy. What you are competing for is memory at the moment of need, which is mental availability rather than habit, and I cover that in how to make your brand top of mind.

The exception is your own operational relationship with existing clients. Retainers, reporting cycles and check-in rhythms are genuine habits, and they are formed the way the research describes: same context, same time, repeated until it runs without effort. That is why an unpredictable delivery rhythm costs more than it looks like it should.

What to do with this

If you sell something bought repeatedly, work on the context rather than the message. Be the default, be in the saved list, be the thing that requires no decision. That is a design and availability problem, close to the work in designing decision interfaces.

If you sell something bought occasionally, stop trying to build a habit and start identifying transitions. The table above is a targeting brief, not a theory.

And if you are currently paying for a loyalty scheme, check whether it is teaching a behaviour or subsidising one that already exists. That is a question your data can answer, and the answer is often uncomfortable.