Should you spend $10,000 building your own loyalty app?

The pitch is seductive: your logo on the customer's home screen, push notifications you own, a direct channel no platform can take away. The problem is the first step. Everything in that pitch is downstream of a customer choosing to download an app for one restaurant, and that choice is made far less often than the pitch deck assumes.

The number to hold onto: 3–8%. That is the share of a single independent venue's customers who will install its app, even with a launch incentive. It is not a marketing-quality problem; it is a reasonable consumer decision. People install apps for things they use weekly across many contexts, not for one coffee shop. Meanwhile, the same customer at the same counter will scan a QR that opens a mini app 55–75% of the time, because scanning costs them nothing and installs nothing.

What $10,000 actually buys. It buys a white-label build: your branding on someone else's loyalty app skeleton, shipped in four to eight weeks. It does not buy downloads. Paid installs run roughly $1.20–2.00 each and get more expensive as you exhaust the easy demand, and cost per active member — the only number that matters — is typically three to five times cost per install. A $10,000 build with a $6,000 install campaign that produces 4,000 installs and 900 monthly-active members has cost you around $18 per active member, against roughly zero for the QR route.

Run the test before the build. Spend 30 days putting a QR mini-app flow on every table and at the counter, with a genuine first-scan reward. Count three things: scan rate as a share of transactions, completed enrolments, and repeat opens in the following month. If mini-app enrolment lands where it usually does, you now have a concrete comparison for any app quote you are considering, and the conversation with the agency changes completely.

The four conditions for a genuine yes: fifteen or more locations so the build amortises; a brand customers search by name, giving you organic installs you do not pay for; at least $12,000 a year earmarked for app marketing, because an unmarketed app is a $10,000 icon nobody has; and an internal owner for the roadmap. Miss any one and the app becomes a maintenance liability within eighteen months.

If you do build, do not replace the mini app with it. The app serves your top decile — the people who visit weekly and want the fastest path. The mini app remains the wide funnel that captures everyone else. Vendors who specialise in branded loyalty apps for regional chains are a reasonable choice when you genuinely meet the conditions; they are an expensive choice when you do not, and no build quality compensates for a 5% install ceiling.