When most of the market is between purchases
In a category bought often, such as coffee or printer ink, a large part of the market is buying in any given month and existing customers come back on their own rhythm. In a category bought once in many years, the arithmetic reverses. If each household replaces its kitchen rarely, then in any month only a small share of households is shopping for one. Everyone else is between purchases: not looking, not reading kitchen pages, not responding to offers, and not reachable through searches they are not making.
Three consequences follow. Demand capture reaches only a sliver. Paid search and comparison sites serve the few people in market now, and every competitor is bidding for the same few. The repeat engine barely runs. The lifecycle marketing guide describes onboarding, repeat purchase and win-back for businesses whose customers come back often; a business whose customers return once in a decade gets most of its value after the sale from what those customers say to others. Memory has to bridge the gap. When the need finally arrives, the businesses that come to mind and are easy to find are the ones that get considered.
| Row label | Frequently bought category | Rarely bought category |
|---|---|---|
| Share of the market buying now | Large | Small at any moment |
| Where growth comes from | Repeat purchase plus new buyers | Almost entirely new buyers |
| What existing customers contribute | Repeat orders | Referrals, reviews and adjacent services |
| The buyer's own experience | Recent and relevant | Old or none, so they lean on others |
| How long the journey lasts | Minutes to weeks | Weeks to months, sometimes longer |
| What attribution windows capture | Much of the journey | Often only the last few steps |
Being remembered when the need arrives
Research from the Ehrenberg-Bass Institute, listed in the sources, has argued that brands grow by being easy to think of and easy to buy, and the mental and physical availability framework built on that argument splits the job in two. Mental availability is about memory: links between the brand and the situations in which people need the category. Physical availability is about being present and easy to buy where people look. For a rarely bought category, both are tested at one unpredictable moment, years after most of the advertising was seen.
Mental availability for an infrequent purchase means reaching people who will buy the category, before they are in the market, with advertising that leaves a clear, branded memory linked to the situations that will start their search. The framework notes that ReachUnique people who saw an ad. sits at the centre of this and that continuous presence across the year usually matters more than a single burst, because memory fades between bursts. The reach and frequency calculator estimates how many people a given volume of impressions reaches and how often. How much budget goes to memory against capturing current demand is a separate decision, set out in the brand versus performance comparison.
Loyalty matters less here than in frequently bought categories. The double jeopardy framework describes a pattern documented mostly in repeat-purchase categories, in which smaller brands have fewer buyers who are also slightly less loyal, and its practical lesson is that growth comes mainly from reaching more buyers. In a category bought once a decade there is little repeat loyalty to manage at all, which pushes even harder towards reaching the people who will buy next.
Physical availability means being easy to find and easy to start with when the moment comes: present in search for the category and the local area, with a complete business listing, reviews, a clear price guide or starting price where the business can give one, and a simple first step such as a visit, a call or a design appointment.
Trigger events and the signals around them
A rare purchase usually starts with an event. Some events are planned and give the buyer time to research; others are sudden and leave almost none. The mental availability framework calls the cues memory has to be linked to buying situations; for infrequent purchases, trigger events are the most important of them.
| Trigger | Lead time | What the person needs first | What the business can do |
|---|---|---|---|
| Moving home | Planned, often months | Ideas, costs and a sense of what is possible | Content for new homeowners; visible in local searches |
| Renovation or extension | Planned, long | Inspiration, process and budget guidance | Project galleries, cost guides, planning advice |
| Breakdown or failure | Sudden | Someone reliable who can come quickly | Be easy to find and to contact; fast response |
| Life change: marriage, children, retirement | Planned, varies | Reassurance that this is the right time and choice | Content tied to the situation, in the buyer's words |
| Business: new site, growth, a failed supplier, end of a contract | Varies | Specifications, references, procurement readiness | Clear specifications, case studies, a short route to a quote |
The intent signals guide explains how to read where a person is from what they do: the searches they make, the pages they read, how recently and how often. Around a trigger, the signals shift from problem-level questions ("how long does a kitchen fitting take", "signs a roof needs replacing") to solution-level ones (types, materials, costs) and finally to supplier-level ones (reviews, quotes, showroom visits). Write content for each level, and make sure the business is findable at the moment the search turns from ideas to suppliers.
Serving the long research phase
Once a trigger has started the search, a rare and expensive purchase usually brings a long consideration stage, as the stages guide notes for considered purchases. The buyer has little or no experience of their own, a lot of money at stake and a fear of choosing wrong. They read, compare, visit, ask friends and come back many times. The Sales Cycle LengthAverage time from opportunity creation to a closed deal. from first contact to order is long, and much of it happens out of the business's sight.
Serve it with content that answers the questions in order. In the See-Think-Do-Care terms of the see-think-do-care framework, most of this audience is in See and Think: they need explanations, examples and costs before they need offers. Guides to the process, honest cost explainers, galleries of finished projects, comparisons of materials and approaches, planning tools, samples and showroom visits all help. Proof carries more weight than in a quick purchase: reviews, case studies, guarantees, accreditations and the chance to speak to past customers.
Keep a way to stay in touch through the long middle. An email list built with consent, offered in exchange for something genuinely useful such as a planning guide, lets the business follow a buyer through months of research. Retargeting helps less than it seems. The remarketing entry explains that platforms cap how long a person stays in an audience at a documented maximum, and that long windows pad the audience with people who behave like cold prospects, so RemarketingAdvertising to previous visitors or engagers. covers only part of a research phase that runs for months.
After the sale: referrals, reviews and service
A customer who will not buy again for years still has three things to give: a review, a recommendation and the adjacent work that follows the main purchase. Because buyers in these categories have little experience of their own, they lean on other people's, which makes every finished job a piece of marketing for the next one.
- Reviews. Ask every customer once the work is complete and they have had time to live with it, in a way the platforms and regulators allow. The local reviews and reputation guide sets out what is allowed and what is not, including why reviews should never be incentivised or asked for only from customers expected to be happy.
- Referrals. Ask, separately from the review, whether they know anyone planning similar work, and make it easy to pass on a name, a link or a brochure.
- Service and care. Care guides, maintenance reminders, warranty registration and seasonal checks are useful reasons to stay in touch, each with a basis in the consent the customer gave. Keep the frequency low enough that every message is welcome; a stream of offers for a product they bought last month teaches customers to ignore the name they should be recommending.
Measuring across long gaps
Standard measurement works against infrequent purchases. A platform credits a conversion only within its Lookback WindowThe time after an ad interaction in which a conversion can still be credited to it., and the lookback window entry explains that a shorter window concentrates credit on interactions close to the purchase and undercounts products with long decision cycles. The right length is the one that matches how long real customers take to decide, which the business can read from its own data. For a purchase whose journey started with an advert seen long before the trigger event, no window is long enough.
Fill the gap from other sources.
| Question | Measure | Source |
|---|---|---|
| Are we being remembered? | Branded SearchSearch queries that contain a company's or product's brand name. volume over time; unprompted awareness among category buyers | Search data; occasional surveys |
| Are we reaching future buyers? | Reach among the target households or businesses | Media reports and the plan |
| Where did customers first hear of us? | Self-reported first exposure | A how-did-you-hear question on every enquiry |
| Is the research phase working? | Enquiries, quote requests, showroom visits, quote-to-order rate | CRM and booking records |
| How long is the journey? | Time from first contact to order | CRM records joined to orders |
| Are past customers selling for us? | Share of new customers citing a recommendation; review volume and rating | Self-reported answers; review platforms |
The self-reported attribution guide covers wording the question, coding the answers and comparing them with tracked sources. For infrequent purchases it is often the only record of the advert, the van, the neighbour's kitchen or the friend's recommendation that started a journey months earlier. Read all of these over long periods: a quarter of rising branded searches and recommendations says more about a memory-building plan than a month of in-platform conversions.
A fitted-kitchen maker plans for the long wait (illustrative)
Stairwhistle Kitchens rebalances towards memory
Stairwhistle Kitchens, an invented fitted-kitchen maker, sells about 140 kitchens a year at an average of £18,000. Its customers replace their kitchens rarely, and its £8,000 monthly budget goes almost entirely on paid search for kitchen terms in its region, where the cost per enquiry has risen from £95 to £160 over two years as rivals bid on the same searches.
It makes four changes. It moves £2,500 a month from search into video and display reach among homeowners within its delivery area, with every advert showing a finished kitchen and the company name clearly. It writes a guide for people who have just moved house ("what to do about the kitchen in your first year") and promotes it to new homeowners in the area. It asks every customer for a review a month after installation and, separately, whether they know anyone planning a kitchen. And it adds a required how-did-you-hear question to its design appointment form.
| Measure | Year before | Year after |
|---|---|---|
| Searches for the company name per month | 610 | 840 |
| Design appointments | 520 | 560 |
| Appointments citing a recommendation | 18% | 27% |
| Appointments citing seeing an advert or a van, with no tracked click | Not asked | 21% |
| Kitchens sold | 140 | 151 |
The ad platform credits the reach campaign with almost no conversions, because most buyers saw the adverts months before they enquired. The self-reported answers and the rise in branded searches show the effect the attribution windows missed, and the company keeps the reach spend for a second year before judging it again.
Common mistakes
- Spending everything on the few people in market. Paid search for a rarely bought category competes for a small group with every rival; future buyers are never reached.
- Bursting and then going silent. Memory fades between bursts, and the need can arrive at any moment.
- Treating past customers as finished. They will not buy again for years, but their reviews and recommendations sell to the next buyers.
- Nagging customers after the sale. Offers for a product they just bought train them to ignore the name they should be recommending.
- Relying on retargeting for a months-long journey. Audience windows end long before many buyers decide.
- Judging memory-building by platform conversions. Lookback windows cannot see a journey that started months before the enquiry.
- Never asking how customers first heard. In a category bought rarely, the answer is often the only record of what started the journey.
