AI Prompts for Referral Programs

Write a referral program announcement email for [company/product] to send to existing customers who have [purchased at least once / been active for X days / reached milestone Y].

Referral offer: [describe the incentive for both the referrer and the new customer]
Product/service: [brief description]
Best customer outcome: [specific result customers achieve—the one you're proudest of]

Email requirements:
- Subject line: Don't mention "referral" or "discount"—find a hook that leads with what the referred friend gets, not the mechanics of the program
- Opening: Start with a genuine compliment framed as a specific observation about customers who use this product ("The customers who've been with us longest tend to be the ones who...")—NOT "We're excited to introduce our referral program"
- Core message: Frame the referral as sharing something valuable, not earning a discount
- Incentive presentation: Describe what both parties get, but lead with what the friend gets—referrers are more motivated by giving value than receiving it
- CTA: One specific action with the referral link or mechanism clearly described
- P.S.: A one-line reminder of the offer for email scanners

Under 200 words total. Tone: warm, like a brand they actually like, not a promotion they have to act on.

A DTC skincare brand launched a referral program with a 20% discount for both the referrer and the new customer. They promoted it via email to 40,000 subscribers. In the first month, they generated 23 referrals. At $180 average order value, they'd spent more on email platform costs and designer time building the program than they earned from it.

The program wasn't bad. The copy was. Nobody knew about the program because the emails explaining it were so generic they felt like every other discount email in the inbox—and the in-product placements asking customers to refer were buried and uninspiring.

That's the specific problem ai prompts referral program content workflows exist to fix.

Quick-Start (Copy This Right Now)

This is the referral program email prompt that outperforms generic discount announcements. Fill in the brackets:

Write a referral program announcement email for [company/product] to send to existing customers who have [purchased at least once / been active for X days / reached milestone Y]. Referral offer: [describe the incentive for both the referrer and the new customer] Product/service: [brief description] Best customer outcome: [specific result customers achieve—the one you're proudest of] Email requirements: - Subject line: Don't mention "referral" or "discount"—find a hook that leads with what the referred friend gets, not the mechanics of the program - Opening: Start with a genuine compliment framed as a specific observation about customers who use this product ("The customers who've been with us longest tend to be the ones who...")—NOT "We're excited to introduce our referral program" - Core message: Frame the referral as sharing something valuable, not earning a discount - Incentive presentation: Describe what both parties get, but lead with what the friend gets—referrers are more motivated by giving value than receiving it - CTA: One specific action with the referral link or mechanism clearly described - P.S.: A one-line reminder of the offer for email scanners Under 200 words total. Tone: warm, like a brand they actually like, not a promotion they have to act on.

What this does: The "frame the referral as sharing something valuable, not earning a discount" instruction is the most important line in the entire prompt. Customers who love your product will share it—but they need to feel like they're doing a friend a favor, not gaming a discount system. This framing produces dramatically different (and better-performing) email copy.

⚡ Pro tip: Test two subject line approaches for referral emails. Approach A leads with the offer: "Give $20, Get $20." Approach B leads with the value to the friend: "Know someone who would love [specific benefit]?" In most B2C tests I've seen, Approach B outperforms Approach A by 15-30% in open rate—because it activates the referrer's desire to be helpful rather than their desire for a discount.

Understanding the Variables

[Best customer outcome]: This is the most important input in the entire prompt. "Our skincare products improve your complexion" is too generic. "Customers who use our vitamin C serum consistently for 6 weeks typically see visible improvement in skin brightness" is specific and shareable. The referral email should be built around the outcome that makes existing customers most enthusiastic—not around the discount mechanics.

[Incentive presentation—friend first]: Most referral programs present the incentive as "give X, get X." Psychologically, referrers are actually more motivated by the idea of giving a friend something good than by what they'll receive. Leading with "Your friend gets 20% off their first order" before "and you'll get 20% off your next purchase" consistently outperforms the reverse framing.

[Tone: warm, like a brand they actually like]: This is the instruction most referral programs ignore. If your email sounds like a generic loyalty program, customers don't feel the personal warmth that makes sharing feel natural. The email should sound like it came from the brand they chose to buy from—with the voice that made them fans in the first place.

⚡ Pro tip: For any referral program, generate the in-product copy separately from the email copy: "Write 3 in-product referral prompts (for display inside the app/account dashboard) for a customer who just completed [milestone action]. Each prompt should be under 25 words, feel like a natural suggestion rather than a marketing message, and lead with what the referred friend gets. Variant A: immediately after purchase. Variant B: after achieving first result. Variant C: at day 30 of active use." Timing matters as much as copy in referral programs—asking for referrals at the moment of highest enthusiasm produces 2-3x more sharing.

Step-by-Step: Building All Your Referral Program Assets

Referral landing page:

Write a landing page for customers who click a referral link and land on our [product] page. This visitor was referred by a friend who uses our product. They're warm but skeptical. The page should: (1) acknowledge the referral in the headline ("Your friend [Name] thinks you'd love this"), (2) lead with the specific outcome customers achieve rather than features, (3) display the referral discount prominently, (4) include one brief testimonial from a customer with a similar profile to the referred visitor, (5) CTA: [primary action]. Under 300 words total.

Follow-up email to people who haven't referred yet:

Write a follow-up email to customers who received our referral program announcement 2 weeks ago but haven't referred anyone yet. This email should: NOT mention that they haven't referred yet, offer one additional reason to refer (new information or a limited-time incentive boost), and make the act of referring feel even easier than the first email did. One-click referral link or pre-filled sharing text should be the only CTA. Under 150 words.

Social share copy:

Write 3 social share copy options that referred customers can use to share on [platform]. Each should: sound like something a real person would actually post (not marketing copy), include a specific benefit or outcome the poster has experienced personally, and include a referral link placeholder. Format: one casual (for friends and family), one professional (for LinkedIn), one enthusiastic (for super-fans). Max 280 characters for Twitter/X variants, longer for Facebook/LinkedIn.

What this does: Pre-written social share copy dramatically increases the number of customers who actually share. The friction of writing their own post is the #1 reason customers who want to refer a friend don't. Giving them ready-made options—especially options that sound personal rather than corporate—removes that friction.

⚠️ Common mistake: Building a referral program without a clear "moment of ask." The best time to ask a customer to refer is when their enthusiasm is highest—immediately after a purchase, immediately after achieving a meaningful result with your product, or after a positive customer service interaction. Generic monthly referral program emails sent to all customers regardless of their recent experience with your product consistently underperform targeted, trigger-based referral asks.

Pro-Level Variations

Referral program for B2B:

Write a referral program email for [B2B product] targeting existing customers who are [power users / long-term accounts / recently achieved ROI milestone]. B2B referral framing is different from B2C: emphasize professional credibility (referring peers trust each other's professional judgment), make the incentive meaningful at the business level (account credits, premium features, or revenue share), and frame the ask as "who else in your network is dealing with [specific problem you solve]" rather than "refer a friend."

⚡ Pro tip: Six months after launching any referral program, run this retrospective prompt: "Here are the results of our referral program: [shares, conversions, revenue]. The emails that generated the most referrals were [describe]. The touchpoints that converted the most referred visitors were [describe]. Based on this pattern, write a hypothesis about what's actually driving referral behavior in our customer base—and suggest the one change that would most likely improve our referral rate in the next quarter." Building this retrospective habit turns your referral program into a continuously improving asset rather than a set-and-forget campaign.

Your Turn

The skincare brand from the opening story rebuilt their referral program emails using the Quick-Start prompt above, reframing from "here's a discount" to "share something your friends will love." Referrals in month two: 187. Same incentive, same program mechanics—different copy. Save your referral program prompts in PromptABCD so your next campaign launches from a proven foundation.

Optimizing Your Referral Program Over Time

Most referral programs are launched and then left alone. The brands with the highest referral rates treat their programs as continuous experiments—testing incentive structures, copy variations, and timing triggers on an ongoing basis.

My referral program has been running for [X months]. Here are the key metrics: - Total customers invited to participate: [X] - % who shared at least once: [X]% - % of referred visitors who converted: [X]% - Average order value of referred customers vs. organic: [X] Diagnose: (1) Which metric represents the biggest opportunity—are we losing people at the sharing step or the conversion step? (2) What's the most likely explanation for the lower-performing metric? (3) Suggest 3 specific tests we could run in the next 30 days to improve the weakest metric. For each test, describe what we'd change, how we'd measure success, and how long we'd need to run it before drawing a conclusion.

What this does: Framing the optimization as a diagnostic first—before jumping to solutions—produces better tests because you're fixing the identified problem rather than randomly trying things. The "how long before drawing a conclusion" instruction forces realistic testing timelines rather than making decisions after 3 days of data.

⚡ Pro tip: One referral program optimization most teams overlook: the thank-you experience for successful referrers. When a referral converts, the referrer usually gets an automatic notification (or nothing). Use this moment: "Write a personal-feeling thank-you email to send to a customer whose referral just made their first purchase. The email should: acknowledge the specific friend they referred (if we have the name), tell them their friend's purchase is confirmed, remind them what they've earned, and include one unexpected extra—either a handwritten-style note from the founder or early access to something coming soon. Keep it under 100 words and make it feel like a genuine thank-you, not a loyalty points notification." Customers who feel genuinely appreciated for referring are significantly more likely to refer again.