Most referral programs are set up once, on a good day, and then left alone. Someone picks round numbers because they sound fair, switches the program on, and moves to the next thing. The store grows, the average order shifts, segments pile up — and nobody goes back to check whether the original setup still makes sense.
That’s the gap Referral Expert is built to close. It’s an AI assistant inside ReferralCandy — a referral specialist you can talk to in plain language. Ask it a question about your program and it reads your setup the way an experienced operator would, explains what it sees, and helps you act on it. Starting today, every merchant on ReferralCandy can chat with it.
Referral Expert works from your actual store. Ask it to look at a campaign and it pulls the details that decide whether a program is healthy — the reward structure, the trigger and delay, the audience segments, your average order value, your order count — and reasons from those numbers instead of from generic best practice. You ask in plain language; it answers like a specialist who already knows your store.
Take a common, expensive mistake. A $20 advocate reward looks generous, but on a store with a $21 average order it isn’t generous — it’s an accounting problem. Each successful referral nearly cancels the order that triggered it, and a program meant to fund growth quietly funds losses instead. That’s exactly the kind of thing Referral Expert catches, because it knows your order size and checks the reward against it.
When it spots problems, it doesn’t hand you a wall of equal-weight suggestions. It prioritizes. Fix the reward economics first. Unify the messaging next — one offer, one tone, applied everywhere, instead of outdated numbers lingering in old copy. Trim overlapping segments after that. Then watch the right ratio to find where the funnel actually leaks. You get a sequence you can act on, not a report you have to interpret.
The same reasoning runs in reverse. Ask Referral Expert to launch a new program and it proposes a starting setup tuned to your store — rewards sized to your order value, a sensible audience to begin with, clear benefit-led copy — and explains why each choice fits.
Then it stops and asks for your approval. Nothing applies until you say yes, and it will offer a more conservative option if you’d rather see one first. When you approve, it sets the campaign up and tells you exactly what it left unchanged, so there are no silent decisions buried in the setup.
That pause is deliberate. Automation that just acts leaves you guessing about what changed. Referral Expert shows its reasoning, waits for your call, and shows its work. You stay the one deciding. It does the analysis you’d never quite get around to doing by hand.
Setting a program up is where most tools stop. Referral Expert keeps going. Once a campaign is live, it looks at who is in it and who could be — recent buyers already subscribed to your emails, sitting just outside the audience you picked — and offers to bring them in. You approve, and it enrolls them in the background while you do something else, with a running count you can check whenever you want.
In its first week, that quiet follow-through imported more than 10,000 customers into live programs that would otherwise have left them out. None of it depended on a merchant remembering the step. The expert noticed the gap, asked, and closed it.
The conversations have been more interesting than a setup wizard would predict. One merchant stopped to stress-test the reward structure before launching — was there real merit in giving the friend and the referrer the same amount, or was that symmetry for its own sake? Another pushed back on discounts altogether, wanting product rewards that fit a brand built on fair, fixed pricing. A third paused the expert mid-flight, not because anything was wrong, but because they wanted to understand the plan before it went live.
The most telling ones come after launch. A merchant arrived with a welcome email that was getting opened and almost never shared, and asked, plainly, what the problem was. The expert read the email, wrote three alternatives, and applied the one the merchant chose — all in the same conversation, without anyone leaving to go edit a template.
A referral program isn’t won at launch. It’s won in the dozens of unglamorous judgment calls that follow — whether the reward respects your margin, whether your segments overlap, whether your copy says one clear thing. Those calls are easy to skip and expensive to get wrong, and until now they mostly depended on someone remembering to revisit them.
Referrals grow a brand because customers trust each other more than they trust your ads. That trust only turns into profitable growth when the economics underneath it hold up — when the reward is worth sharing and the store can still afford it. Keeping that balance right, as the store changes, is the whole game.
Referral Expert is now in your hands to help you do exactly that — on your real numbers, one approval at a time. Open ReferralCandy and ask your new AI assistant to take a look at your program.
ReferralCandy is a customer acquisition tool used by thousands of ecommerce retailers and business owners around the world.
Grow your sales at a ridiculously
lower CAC.
