SaaS referral program: ask activated users, reward both sides

Aug 22
4 min read

Updated: 3 days ago

Introduction

Referral programmes in software have a wide range of outcomes. Some are the largest acquisition channel a product has; most produce a handful of signups and a spreadsheet nobody maintains.

The difference is rarely the reward. It is whether the programme asks the right users at the right moment, and whether recommending the product is something they would do anyway.

1. A SaaS referral program should only ask activated users

The most common design error: prompting every new signup to invite colleagues.

A user who has not yet experienced the product working has nothing to recommend, and asking makes the product feel more interested in growth than in their problem.

Gate the ask on activation. Users who have reached the point where the product has demonstrably worked are the only ones who can refer credibly, and their referrals convert at a much higher rate.

2. Ask at a moment of demonstrated value

Timing beyond activation matters too.

Good moments: after a task completes successfully, after a positive support interaction, after a milestone in usage, or after a user takes an action indicating satisfaction.

Poor moments: on login, in a persistent banner, or immediately after an error. Context-free prompts train users to dismiss them, which also reduces the effectiveness of every future prompt.

3. Reward both sides, in product rather than cash

Two-sided rewards convert better because the referrer is offering their contact something rather than profiting from them.

Product credit, extended trial, additional usage or a free month works better than cash for most software products. It costs less, it keeps the value inside the product, and it avoids the awkwardness of appearing to pay people to recommend you.

Cash rewards suit affiliate arrangements with people who have audiences. They suit ordinary customers less well.

4. Size the reward against acquisition cost

The arithmetic that determines whether this is worth running.

Compare the total cost of both rewards against what you currently pay to acquire a customer through advertising. Referral acquisition should be cheaper, and referred users typically retain better, which increases what you can afford.

If the reward has to approach your paid acquisition cost to generate any activity, the underlying problem is that users are not enthusiastic enough to recommend the product — and that is a product question, not a programme question.

5. Make the mechanics take one action

Every additional step loses most of the remaining participants.

A personal link, copyable in one click, that works when pasted anywhere. Optionally an email invitation flow with prefilled text. Not a form requiring the referrer to enter names and addresses.

Show them the status too: who signed up, what they have earned, what is pending. Referrers who cannot see the result of the first referral rarely make a second.

6. Plan for abuse before launch

Referral programmes attract exploitation, and retrofitting controls is harder than including them.

The predictable patterns: self-referral with a second email address, referrals that sign up and never activate, credit farming through disposable accounts, and posting codes on discount sites when you intended personal sharing.

Controls that work: pay the reward on activation or first payment rather than signup, exclude matching payment details, cap rewards per account, and state the rules clearly so enforcement is not arbitrary.

7. Recognise where referrals genuinely fit

Not every product suits this, and forcing it produces the dead programmes described at the start.

It works best where users know others with the same problem — tools for a specific profession, collaborative products, and anything used by teams.

It works poorly where the product is a competitive advantage the user prefers not to share, where the buyer's network does not overlap with other buyers, or where usage is private. In those cases case studies and comparison content are a better use of the effort.

8. Measure referred-user quality, not referral volume

The metric that matters: how referred users behave compared with users from other channels.

Track activation rate, conversion to paid, retention and expansion separately for referred users. They usually outperform, which is the strongest argument for investing further in the channel.

Also track what proportion of referrals came from a small number of referrers. It is normally a small group, and they deserve recognition beyond the standard reward — those relationships are worth more than the programme mechanics.

Conclusion

Gate the ask on activation and place it after a moment of demonstrated value rather than on login or signup.

Reward both sides with product credit sized against your paid acquisition cost, reduce the mechanics to one action with visible status, design the anti-abuse rules before launch and pay on activation rather than signup, be honest about whether your product is one users would naturally recommend, and measure referred-user retention rather than referral volume.