Decisions · Technology

Should you build or buy the loyalty platform

Buy the loyalty platform, not build it, unless your programme's only defensible edge is a points or rewards mechanic no vendor can configure and you can staff a product team for at least five years. For everyone else, buying transfers the cost of maintenance and compliance to a vendor.

The decision

The choice is not between a vendor contract and an empty codebase. It is between renting a shared operational backbone and owning the system that records member liabilities, promotion rules, and points expiry. The mechanism underneath is who absorbs future change. A vendor spreads the cost of new features, security patches, and regional compliance across many programmes, while an in-house build concentrates every future change on your own engineers. For most loyalty programmes, that pooled maintenance is cheaper and lower risk than building a private platform.

What it costs you

Buying costs you a percentage of member lifetime value in vendor fees, some loss of exclusivity, and the discipline to configure rather than code. As an illustrative calculation, you might pay a 25,000 dollar annual licence plus 0.03 dollars per active member per month. With 200,000 active members, the monthly per-member fee is 200,000 times 0.03, which is 6,000 dollars. Annualised, that is 72,000 dollars, so the total vendor bill is 97,000 dollars. A minimal internal build would need at least two engineers at 140,000 dollars each fully loaded, or 280,000 dollars a year before hosting, security, or audit costs. The buy path also leaves you with less direct control over member data flows, but it avoids the deeper liability of holding points balances in a system you must secure and audit alone.

How to decide

Apply this test. Write down the single promotion mechanic that creates the most member value in your programme. If you cannot name one that no vendor can implement after a paid proof of concept, buy. If you can name one, then check whether you can hire and retain at least three product engineers and one loyalty operations analyst for five years. Only when both conditions hold should you build. Most teams fail the second test even when they pass the first, because a loyalty platform is not a one-off project but a permanent operating system.

What would change our mind

Our recommendation flips to build when a vendor's annual total cost exceeds the fully loaded cost of four internal engineers for two consecutive years and the programme requires a points expiration, transfer, or redemption rule that no vendor can configure after a documented trial. The cost threshold is not enough on its own; a high vendor bill may still be cheaper than the cost of owning member liabilities during a self-build failure. But when a specific, revenue-critical mechanic is impossible to buy and the internal team can be funded from the savings, building a thin layer on a generic ledger becomes the better path.

Related