Chapter 7
Five ways people buy
People buy in five recognisably different ways. The mode decides what marketing should do, and how much AI matters.
Three questions that sort every purchase
Three questions separate one purchase from another, each studied for more than fifty years.
| Question | The range | Where the idea comes from |
|---|---|---|
| How much is at stake? | A snack, up to a six-figure contract | Howard and Sheth, 1969 |
| Is it a first purchase or a repeat? | Something new, through to the usual re-order | Robinson, Faris and Wind, 1967 |
| How many people decide? | One person, up to a committee | Webster and Wind, 1972 |
Robinson, Faris and Wind's straight rebuy is what this book calls habitual buying. Their modified rebuy is the renew-or-switch moment, which cuts across every mode.
The five modes
Together, the three questions give five modes.
| Mode | Looks like | Who decides | What wins it |
|---|---|---|---|
| Habitual | Groceries, fuel, the usual re-order | One person, on autopilot, across a few familiar brands | Coming to mind and being within reach |
| Quick choice | Something cheap you have not bought before | One person, in minutes | Being the name in front of them at that moment |
| Considered personal | A car, a home loan, a holiday | One or two people, over days or weeks | Answering their questions and being believed |
| Owner-led business | A bookkeeper, a CRM, an agency | One to four people, over weeks | Being believed by the owner, plus a business case a second voice can check |
| Committee | Enterprise software, a major contract | A group, over months | Winning each person, and surviving scrutiny |
Habitual doesn't mean loyal to one brand. Ehrenberg-Bass researchers find buyers are loyal "to a repertoire of brands, not to a single brand" (Ehrenberg-Bass). The habit is a short list, and the job is to stay on it.
From mid-2017 to mid-2019 I was Head of Digital at CUA, the bank now called Great Southern Bank. Most weeks there, I bought in three completely different ways.
The coffee was habit. Every morning, sometimes twice a day, I went next door to the same café, to the same barista, and ordered the same thing with coconut milk. Once you've found your coffee, you don't switch. I didn't compare and I didn't think about it, and I liked it that way. It felt comfortable and reassuring, a bit like home.
Lunch was the opposite. I'd walk out into the city with no plan and duck into the first place that caught my eye. It felt exciting, and I came back to my desk with more energy than I left with.
And then there were the committees. Sigh. I hated them. Slow, drawn out, over-complicated and over-justified. The one I remember most is the company CRM. That process ran for more than a year, and it was still going when I left. It was a big, important decision, and procurement really did matter. But my gosh, the frustration was real. I'd walk out of those meetings tired and confused, and often not much caring which way it went.
Same person, same week, three completely different buyers.
To find your mode, picture your typical sale and follow the tree.
Not every purchase needs all seven
In all five modes, Known and Found are gates: if the buyer can't recall you or find you, nothing else you do is seen. What differs is what carries the weight once a buyer is through.
So there's no universal weakest link. A chocolate bar needs to be remembered, then within reach. A software contract has to win all seven, several times over. Knowing your mode, and which stages carry it, is the first act of strategy. Chapter 30 turns these weights into numbers in the Selectability Score.
Where the small business fits
An owner-led business doesn't buy like an enterprise. In G2's 2025 study, about 76% of businesses with up to 250 staff made software decisions with four or fewer people (G2). Among firms of 1,000 to 5,000 staff, about 37% did. Small firms turned first to Google (47%), ahead of review sites (38%) and AI search (35%).
In a 2026 Capterra study, about half of whose respondents were small businesses, successful software buyers shortlisted three suppliers and decided within three months (Capterra).
So an owner choosing a bookkeeper decides more like someone buying a car than like a committee. That's my reading, not a tested finding. The difference is business risk, and often a second voice: a partner, an accountant, a finance lead. So the owner needs proof from a business like theirs, and numbers the second voice can check.
What AI changes in each mode
| Mode | What the evidence says AI changes | How strong |
|---|---|---|
| Habitual | Little sign that it changes brand choice yet. My read is that it will reinforce habit before it disrupts it. | Thin |
| Quick choice | AI summaries cut the clicking, so the brand named in the answer probably gains. No purchase data yet. | Thin |
| Considered personal | The clearest effect. Complex categories draw the most AI referrals, and there those visitors convert better than most channels. | Moderate to strong |
| Owner-led business | Google still leads, with AI close behind. It behaves like a considered personal purchase with a business case. | Moderate, by inference |
| Committee | AI shapes and speeds the shortlist. The day-one list still wins, and evaluation takes longer. | Moderate |
AI's hand is strongest where the buyer has the most to work out. In the two low-effort modes, nobody yet knows.
In committee buying, AI has made suppliers easier to find and harder to buy from. In G2's 2026 survey, 40% of buyers named evaluation as the longest stage. And 49% had seen a finance chief veto a purchase already approved (G2).
What is established and what is mine
The three questions are established research. The five modes are my own synthesis of them, tested against the evidence on AI from 2025 and 2026. I found no accepted buying typology for the AI era. Renewal isn't a sixth mode: every mode has one, and the customer can decide differently.
In every mode, somebody advises the buyer. Increasingly, the first adviser isn't a person.
Whose evidence this is. G2 runs a software review marketplace. Capterra is a software review site. The Ehrenberg-Bass Institute sells research to sponsor brands. The three questions come from academic work by Howard and Sheth, Robinson, Faris and Wind, and Webster and Wind.
Take your three biggest products or services through the decision tree, and write each one's mode beside its name.