Referrals

Referrals

Referral programs fail at small scale for one reason: they formalize word-of-mouth that doesn't exist yet. The order is: make the product worth mentioning → make mentioning effortless → then add incentives. This skill builds that ladder and stops you from shipping a rewards system nobody triggers.

Context

Read product-context.md if present. Otherwise ask for:

1. Product, pricing, and current user counts (free / paid)
2. Any organic word-of-mouth signal so far (users mentioning it publicly, "how did you hear about us" answers)
3. What a user could give a friend (extended trial, credits, discount) and what it costs you
4. Stack — where invite links and credit grants would live

Process

1. Readiness check (be honest)

A referral program amplifies existing word-of-mouth; it cannot create it. Prerequisites:

• ≥ 1 organic referral you know of (someone mentioned it unprompted)
• Retention holds — referring users into a leaky product burns the referrer's trust too
• < both → skip to step 2 and stop there; revisit incentives at ~100 active users

2. The zero-infrastructure referral loop (do this first)

• Make output shareable by default. Whatever the product produces should carry a tasteful attribution ("Made with [Product]") that links back — with the user's ?ref= code if they have one. This is the highest-converting referral surface most products already own.
• The founder ask. After a user's clear win (activation, a success moment, praise in support), send one line: "Glad it worked — if you know one other founder with this problem, here's a link that gives them [perk]." Timing beats incentive: ask at the peak, not in a drip.
• Track with ?ref= codes — a per-user code appended to the product URL, captured into a source column on signup. One column, no SaaS.

3. Choose the incentive (when step 1 passes)

Incentive	Use when	Watch out
Two-sided credit (both get a month/credits)	Paid SaaS with usage-based value	Cost caps: limit rewards per referrer/month
Extended trial for the friend only	Conversion bottleneck is trial length	Weak referrer motivation — pair with recognition
Founding-member status / early access	Pre-launch or waitlist stage	Only works while access is genuinely scarce
Cash / affiliate %	High LTV, B2B, promoters are creators	Attracts mercenaries; needs payout plumbing — last resort

Default for a solo SaaS: two-sided credit in your product's native unit (e.g. "you both get 20 credits"). It rewards in product value, costs marginal COGS, and reinforces usage.

4. Wire the minimum program

1. Per-user referral code on the account page: yourapp.com?ref=<code> + a pre-written share message they can copy
2. Signup captures ref → source; a credit grant to both sides on the referred user's activation (not signup — reward real users, not email addresses), idempotent by (referrer, referred) pair
3. Show the referrer their tally ("2 friends joined — 40 credits earned")
4. Announce once to existing users (email + changelog), then let the product surfaces carry it

5. Measure whether it's real

Monthly: referred signups, referred-user activation rate vs. baseline, cost of granted credits. Healthy loop: referred users activate better than average. If referred signups < 5% of total after 60 days, the loop isn't the constraint — go back to retention or acquisition.

Output

1. Readiness verdict — amplify or wait, based on this product's actual signals
2. Shareable-output spec — where "Made with" attribution + ref link goes in this product
3. The ask — founder message with the trigger moment defined for this product
4. Incentive recommendation — one incentive with cost math at current usage
5. Implementation spec — schema (ref code, source capture, idempotent grants), account-page UI copy, and the announcement email
6. 60-day scorecard — the three numbers that decide keep/kill

---

The share message needs something worth sharing. SoloMax gives every user a launch video made from their product URL — the kind of output people post publicly, attribution and all.