How Much to Offer for a Referral (and Why It Works)

A landscaper we talked to last spring had a referral problem hiding as a marketing problem. His customers loved him, told their neighbors, and still almost nothing came in through word of mouth. When we asked what he offered someone for sending a referral, he shrugged and said whatever felt fair, which is another way of saying he had no answer at all. That's the answer most small business owners give, and it's why so many referral programs quietly die. The real question, how much to offer for a referral, has a math answer, not a gut-feeling one.

Owners either go too small, a $10 gift card nobody bothers to redeem, or too big, a flat $100 that eats the whole job's profit. Both mistakes come from the same place: nobody sat down and worked out what a referral is actually worth to the business before deciding what to pay for one. Fix that one number and the rest of the program, the asking, the tracking, the follow-up, gets a lot easier.

Why the guess usually fails

A reward that's too small insults the person doing you a favor. Somebody who just sent you a paying customer, at real risk to their own reputation, notices when the thank-you is a $10 coupon. A reward that's too big does the opposite damage: it either gets ignored by an owner who's quietly panicking about margin, or it gets paid out and turns what should be a profitable acquisition channel into a break-even one.

There's a second failure that has nothing to do with the dollar amount: nobody asks. We've seen owners build a whole referral page on their site and never once say the word "referral" out loud to a happy customer. A Google review is public and passive. A referral is private and requires someone to actually ask for it, in person or in a text, which is exactly why most businesses never get one going.

How much to offer for a referral, in three steps

The number should come from your margin, not your gut. Here's the version we walk clients through.

Step one: find the real gross margin on a typical job or sale, not the revenue. A $500 HVAC repair with $300 in labor, parts, and drive time nets about $200. That $200 is what a referral is actually worth to you if it leads to one job like it.

Step two: set the reward at 10 to 20 percent of that margin. Below 10 percent, most people won't bother. Above 20 percent, you're giving away the reason you did the work. For the $200 example, that's a $20 to $40 reward, which most businesses round up to a clean $25 or $50 because round numbers get remembered and repeated.

Step three: pick a form that costs you less than its face value. A $50 credit toward a future service costs a service business roughly its cost of goods, often $15 to $20, while a $50 cash reward costs exactly $50. Credit toward a future purchase does double duty: it rewards the referral and books a second sale.

What that looks like for four different businesses

A contractor running a service van, average ticket $500, margin around 40 percent, nets roughly $200 a job. At 15 percent of margin that's a $30 reward, which most contractors round to a flat $50 gift card, still comfortably under 20 percent and easy to say out loud on a job site.

A bakery doing custom and wedding orders, average ticket $300, margin closer to 35 percent, nets about $105. Fifteen percent lands near $16, which a bakery we've worked with rounds to a $20 credit toward the referring customer's next order rather than cash, since it also guarantees a repeat visit.

A dental office, where a new patient is worth roughly $700 in year-one billing at about 45 percent margin, nets around $315 per referral. Even a conservative 10 percent, about $31, rounds cleanly to a $50 credit toward the referring patient's next cleaning, which patients notice and dentists can afford.

A boutique with a lower average sale, say $90 at 50 percent margin, only nets about $45 per transaction, so a straight percentage reward feels small no matter how it's framed. Boutiques do better with a tiered structure instead: refer one friend, get $10 off; refer three in a season, get $40 off, which keeps the per-referral math sound while still feeling generous.

In practice: what USBN does

When a client asks us to build a referral program, the first thing the team behind USBN does is pull 90 days of sales or invoice data, not run a brainstorm. That data gives an actual average ticket and, once the client tells us their cost side, a real margin number, not an estimate.

From there we set one reward, test it for 60 days with existing customers before ever putting it on the website, and track two things: how many people redeem it and whether the referred customer becomes a repeat one. If redemption is under 5 percent after two months, the number was probably too low or nobody actually asked. That work usually sits inside a broader marketing engagement, because a referral program that isn't promoted anywhere customers actually see is just a page nobody visits.

We've also told clients not to launch a referral program at all, more than once, usually because their current customer base is too small or their margins are too thin to support any reward that would motivate someone. Better to fix pricing first and come back to referrals once there's room in the number.

The one thing to do this week

Pull your last 20 sales. Find the average ticket, estimate the margin as honestly as you can, and multiply by 15 percent. That's your number. Then ask five customers you already know are happy, by text or in person, if they know anyone who could use what you do, and tell them what they'll get for it.

If you want help pulling the real numbers or building the program around them, that's a short conversation, not a long project: reach out and the team behind USBN will walk through it with you.