How do you capture customer contact details at a cafe without slowing the queue?

Contact capture is the foundation of every retention tactic and the place most programmes quietly fail. The default design — the cashier asks for a phone number — has a fatal property: it competes with throughput. Twenty to forty seconds per transaction is survivable at 10am and impossible at 8:15, which means the mechanism switches itself off exactly during the hours when your most habitual, highest-value customers are in the queue.

The fix is architectural rather than motivational. No amount of staff training makes a thirty-second conversation fit into a rush. Move the action to the customer: a QR they scan while waiting, opening a one-field form in the browser or a mini app inside a messaging platform they already use. Placement matters more than design — the queue line, the counter edge at eye level, the table tent, the receipt footer. Venues that place it where people already stand idle typically convert 40–60% of transactions; venues that put it on a poster by the door convert almost nobody.

The incentive should be claimable on the next visit, not the current one. 'Scan to get a free pastry next visit' does two jobs with one ask: it identifies the customer and it creates the reason to return. Giving the reward immediately spends the same money and buys only the identification.

Keep the form to one field. Phone number, or a single platform-authorised tap. Every extra field — name, birthday, email, preferences — costs a measurable share of completions, and you can collect all of them later through the messaging channel once the relationship exists. Birthday in particular is better asked in a follow-up message a week later, where it converts well and costs nothing.

Then add a fallback for the people who did not scan: receipt capture. The customer photographs or forwards a receipt afterwards and claims the visit retroactively. It is slower and imperfect, but it recovers a meaningful share of the base — especially delivery and takeaway customers, who never stood in front of your QR at all.

Two failure modes worth naming. First, treating capture rate as a fixed property of your customers rather than of your placement: if you are at 12%, move the QR before concluding that your customers do not want to join. Second, capturing contacts and then not using them — an unused list decays, and a first message sent six months after sign-up performs poorly because the customer no longer remembers agreeing to anything.

On consent and record-keeping: state clearly what the number is used for, separate loyalty from marketing consent, and keep a record of when and how each consent was given. This is a legal requirement in most markets now and it is also a practical one — a clean, consented list is worth several times a large, doubtful one.