The subscription price is easy to put in a spreadsheet. The cost of three delayed integrations, a failed redemption at checkout and a campaign that takes six weeks to launch is usually less visible. Those costs still belong in the decision.
Why SaaS can be the stronger commercial choice
An enterprise SaaS loyalty platform can give a brand an operating foundation it would otherwise have to build and maintain: reward rules, transaction history, permissions, integrations and tools for the people running the program. Where those capabilities fit the business, teams can spend more time improving offers and less time maintaining infrastructure.
Speed matters because a useful experiment delivered this month has a chance to teach you something this month. But SaaS is not automatically cheaper, and the word enterprise is not evidence of quality. A product that requires extensive custom work or charges heavily for every interaction may be a poor fit. Compare the complete operating model, not the labels.
Count the costs on both sides
For SaaS, include implementation, migration, connectors, staff training, support, message delivery and any usage-based fees. Ask what happens to the bill when membership, transactions or branches double. Confirm which integrations exist today, which need configuration and which are a separate development project.
For a custom build, include engineers, product management, quality assurance, hosting, monitoring, security maintenance and the ongoing cost of changing commerce and POS integrations. Include the work that does not make an attractive demo: failed-event recovery, refunds, staff permissions and reconciliation.
Use the same time horizon and scope for both options. A three-year SaaS estimate should not be compared with only the first release of a custom build. Also account for exit costs: data export, migration support, contract notice and the ability to reconstruct customer balances elsewhere.
| Cost area | Question to put in the proposal |
|---|---|
| Implementation | Who owns each connector and the acceptance tests? |
| Ongoing operation | Which tasks need our team, and how much time? |
| Usage | What changes the bill as the program grows? |
| Support | Who responds when redemption fails during trading? |
| Exit | Can we export transaction history and migrate balances? |
Measure lift, not the member halo
Your best customers may be the first people to join. If they already spent more than other customers, comparing member revenue with non-member revenue will overstate what the program caused. Membership and loyalty are related, but the difference is not a clean measure of incremental growth.
Use a pilot with a suitable comparison group. Where practical, randomly assign eligible customers to an invitation or offer group and a holdout. Measure over a buying cycle long enough to observe repeat behaviour. Watch whether a promotion simply brings a planned purchase forward or creates an additional profitable purchase.
Agree the calculation before seeing the result. Use net sales after returns and discounts, a consistent contribution-margin definition, and program costs that are counted once. Report a range when the evidence is uncertain rather than presenting a forecast as a guaranteed return.
A simple business case, with the assumptions visible
Consider a purely illustrative pilot: 200,000 in incremental net sales, a 35% contribution margin before program costs, and 40,000 in combined reward, software, implementation allocation and operating costs for the same period. Incremental contribution before those costs is 70,000. Net benefit is 30,000, and ROI on the 40,000 cost base is 75%.
That is arithmetic, not a LoyOne result or a promised return. If incremental sales are only 100,000 under the same assumptions, net benefit becomes negative 5,000. The point of a model is to show which assumptions matter. If reward expense is already included in the margin figure, do not subtract it again.
Make the pilot a buying condition
Ask the vendor to demonstrate a purchase, a partial return, a reward redemption and a reconciliation report using a realistic journey. Then define pilot acceptance criteria with marketing, operations, finance and IT. The criteria should cover the customer experience and the people who have to support it.
SaaS earns its premium when it reduces work, improves execution and helps the business run profitable experiments. A custom build can be justified when requirements are genuinely unusual and the business has the team to own them. Choose the option with the strongest credible operating case, then keep measuring after the contract is signed.
- Set a baseline and comparison method before launch.
- Budget implementation and operation, not just the licence.
- Test returns, failure recovery and data export.
- Review incremental contribution alongside customer experience.
Original editorial imagery created with AI for this series. Images are illustrative; people shown are not customer endorsements.
Put the thinking to work.
Bring one customer journey and your current systems. Let’s explore what a connected loyalty program would need to do for your brand.
Book a demo
