Nine steps to choose loyalty software and providers

What to take away

Why supplier-led buying fails

A demo shows the product at its best. It does not show how the platform behaves when a member earns points in a shop, redeems in an app and contacts support about a missing reward. Those three events touch different systems, and the seams between them are where programmes break.

Buying from the demo backwards produces two familiar problems. The first is a platform that fits the pitch but not the data model. The second is a contract that prices growth as a penalty, so a successful launch becomes an unbudgeted cost.

A better sequence starts with the problem and ends with the contract. The middle steps establish what you need, who can supply it and how the pieces connect. That order also gives you something to compare suppliers against, rather than comparing them with each other.

Step 1: define the commercial problem

Write down what the programme must change. A retailer with flat repeat purchase rates needs different tooling from a subscription business trying to reduce churn, even though both may buy a points engine.

Name the metric, the current value and the target. If nobody can state the current value, the programme has no baseline and no way to prove itself later. Baseline work belongs before procurement, not after.

Keep the problem statement to one page. It becomes the reference point for every later decision, including which suppliers get a second meeting.

Step 2: map the customer journeys you must support

This map exposes requirements that demos hide. Expiry rules, for example, vary by market and affect both the liability on your balance sheet and the wording on every statement.

A journey map also shows where a loyalty platform must hand off to other tools. Referral tracking might sit in the loyalty product, in a separate referral platform or in your ecommerce stack. Decide which, and write it down.

Step 3: classify the data and the roles

Under the UK GDPR, you need to know whether you are a controller, a processor or both for each processing activity. Loyalty programmes often involve all three positions across different data sets.

The ICO's guidance on controllers and processors sets out how to determine each role and what belongs in a written contract between the parties. Read it before you draft requirements, because the answers shape what you can ask a supplier to do.

Practical consequences follow. If the supplier determines why personal data is used, it may be a controller for that activity. If it only acts on your instructions, it is a processor and needs a data processing agreement.

Step 4: write requirements a supplier can price

Split requirements into must-haves, should-haves and exclusions. Must-haves are the features without which the programme cannot launch. Exclusions matter just as much: they stop suppliers padding responses with modules you will never switch on.

Typical must-haves include a points engine with configurable earning rules, tier management, expiry and liability reporting, referral tracking, an API or webhooks for your ecommerce stack, and integration with your points of sale. Write each one as a testable statement, so a supplier's answer can be marked yes, no or partial.

Attach volumes to each requirement. Number of members, transactions per month, peak earning events and expected referral volume all change the price. A supplier quoting on a vague brief will quote conservatively.

Our loyalty software selection guide shows how to turn these requirements into a scoring model, which is the fairest way to compare responses from suppliers of different sizes.

Step 5: build the longlist from several sources

Analyst research from firms such as Gartner and Forrester, peer recommendations, partner directories and the marketplaces around your existing stack all produce candidates. If you trade on Shopify or Adobe Commerce, the app marketplaces there list loyalty and referral apps with reviews and integration notes attached, which is a fast way to see what fits your stack.

Aim for eight to twelve names. Include at least one specialist in referral mechanics and one platform with strong retail point-of-sale experience, even if neither seems a perfect fit at this stage.

Candidate lists in this market usually include loyalty platforms such as Eagle Eye, LoyaltyLion, Antavo, Open Loyalty, Annex Cloud and Comarch, and referral specialists such as Mention Me, Buyapowa, Talkable and ReferralCandy. Judge each name against your must-haves rather than against the other suppliers' marketing, and add any platform your ecommerce or point-of-sale provider already integrates with.

Record why each name is on the list. A longlist with reasons is easier to defend than one assembled from whoever replied to an email.

Step 6: run a structured evaluation

Send the same brief to every supplier and ask for the same response format. Score answers against the weighted requirements from step four. Keep the scoring with the responses so the decision trail survives staff changes.

Ask each supplier for two references in your sector and one that left. Departures are more informative than successes, though few buyers ask. If a supplier cannot provide a former customer, treat that as a data point.

Our supplier comparison guide covers how to normalise pricing across different commercial models, which is usually the hardest part of the comparison.

Step 7: test the integration before you commit

Integration is where timelines slip. Ask for a sandbox, a test member account and access to the API documentation before signing. Then get your own engineers to attempt one real journey, such as earning points on a test order.

The HubSpot API reference is a useful benchmark for what good documentation looks like. If a loyalty supplier's developer materials are thinner than a mainstream CRM's, your build will cost more than the quote suggests.

Decide early which system owns the customer record. Duplicate records across a loyalty platform and a CRM create reporting disputes that outlast the implementation project.

Step 8: plan the acquisition channels

A loyalty programme needs members before it can retain them. Paid search, paid social and partner placements all feed enrolment, and each has different cost per sign-up characteristics.

Enrolment prompts at checkout, referral offers and bonus-point campaigns are run from inside the tooling itself, so ask each supplier to show how those campaigns are set up, measured and capped.

Set a target cost per enrolled member and a payback window. Without both, acquisition spend is judged on volume alone, which flatters cheap sign-ups that never earn or redeem.

Step 9: negotiate the contract and the exit

Pricing models differ: per member, per active member, per transaction, platform fee plus usage, or a revenue share. Model each against your three-year forecast, not just year one.

Check the clauses that cost money later. Overage rates, data export fees, notice periods and auto-renewal terms deserve as much attention as the headline price. For example, a team paying £400 a month on a per-active-member model could see that double if inactive members are reclassified at renewal.

Our vendor due diligence checklist covers security questionnaires, financial checks and the questions to ask about ownership of programme data.

Implementation and the first year

Go-live is the start of the operational workload, not the end of the project. Points liability needs reconciling monthly. Fraud rules need tuning as soon as real members test them. Statement queries arrive in the first week.

Staff the programme before launch. A common failure is a well-built platform with nobody assigned to run campaigns, manage the reward catalogue or answer member queries. The technology is rarely the constraint after the first quarter.

Our implementation guide sets out a phased rollout that keeps the first release small enough to fix quickly.

Choosing between a platform and a specialist

A full loyalty platform handles enrolment, earning, redemption, tiers and reporting in one product. It suits teams that want one contract and one data model, and it usually costs more at low volumes.

A specialist referral or rewards tool does one job well and connects to the rest of your stack. It suits teams with existing CRM and ecommerce systems, but you carry the integration work and the joins between systems.

Judging total cost over three years

Add licence fees, implementation, integration development, ongoing support and internal time. Internal time is the line most often omitted, and it is often the largest.

Ask suppliers to price two scenarios: your current volumes and a plausible growth case. The gap between them tells you how the contract behaves if the programme succeeds, which is the outcome you are buying the software to achieve.

Common questions

How long should a selection project take?

For a mid-market team, six to twelve weeks from problem statement to signed contract is realistic. Integration testing adds time. Compressing the process usually shifts cost into implementation rather than removing it.

Should we buy loyalty and referral software from one supplier?

Only if one product does both well. Many platforms offer referral as a minor module, and a weak referral engine undermines the acquisition channel that feeds the whole programme. Compare the referral mechanics specifically before deciding.

What should we check before signing a data processing agreement?

Confirm each party's role for every processing activity, the sub-processors involved, where data is stored and how it is returned or deleted at exit. The ICO guidance in step three is the reference point for these terms.

Is a points programme suitable for a small business?

Sometimes, but the operational overhead is fixed while the member base is not. A small retailer may get more from a simple referral offer or a tiered discount than from a points engine with liability reporting and expiry rules attached.