You’re running a promotion. Everybody on your list gets the same 15% off email. Your loyal regulars — the ones who’d buy anyway — take the discount and you eat the margin. The people who haven’t shopped with you in months ignore it completely because 15% isn’t enough to bring them back.
That’s not a promotion. That’s a donation to people who didn’t need it.
The fix is straightforward: different customers get different offers based on how they actually behave. Artificial intelligence (AI) tools can now do this sorting and sending automatically, even on a small business budget.
The Logic Is Simple, Even If You’ve Never Done It
You probably already know this intuitively. The customer who buys from you every week doesn’t need a 25% discount to come back. A small perk — free shipping, a bonus item, a loyalty point — keeps them happy without wrecking your margins.
But the customer who hasn’t purchased in two months? They’ve half-forgotten you exist. They need a real reason to come back. A bigger discount, a specific offer tied to something they bought before.
The problem was never understanding this. The problem was doing it manually when you’ve got hundreds or thousands of customers. That’s where AI-powered tools come in.
Tools That Do the Sorting for You
**For online sellers:** Klaviyo uses machine learning (ML) — basically, software that learns patterns from your sales data — to automatically group customers by purchase history, browsing behavior, and churn risk. You set rules like “send 20% off to people who browsed but didn’t buy in the last 7 days” and “send free shipping to customers with 3+ purchases.” No coding. Klaviyo offers a free tier for small contact lists, with paid plans starting in the low-to-mid tens of dollars per month depending on your contact count — check klaviyo.com/pricing for the current free tier limits and paid plan costs before you commit, since they adjust these periodically. Mailchimp offers similar predictive segmentation on its Standard plan, also in the roughly comparable monthly price range — again, verify the current price at mailchimp.com/pricing, since both platforms adjust pricing regularly.
**For brick-and-mortar:** Square’s loyalty add-on automatically groups buyers by visit frequency and spend. A coffee shop could send a “We miss you — here’s $2 off” text to customers who haven’t visited in 30 days, while weekly regulars earn points toward a free drink. The system decides who gets what based on transaction history. Square adjusts loyalty pricing periodically, so check squareup.com for the current monthly cost before signing up.
**The budget option:** If you already have customer data — even a spreadsheet export from Shopify or your point-of-sale (POS) system — you can paste *anonymized* purchase data into ChatGPT Plus or Claude and ask it to group customers into segments, then draft tailored discount copy for each one. ChatGPT Plus runs around $20/month, though OpenAI has adjusted this price before — check openai.com/pricing for the current number.
**Critical privacy note:** Never paste personally identifiable information — names, emails, phone numbers — into any AI chat tool. Aggregate or anonymize your data first. Various federal and state privacy laws may apply depending on your business size, location, and where your customers are. Some, like California’s consumer privacy law, have specific revenue and data-volume thresholds that determine whether they cover you. The safe move: treat customer data carefully regardless of whether a specific law technically requires you to.
Why This Matters for Your Margins
Blanket discounts train your best customers to wait for sales and underwhelm the customers you’ve already lost. Segmented offers let you spend your discount dollars where they’ll actually change behavior.
We can’t promise you a specific return-on-investment (ROI) percentage — vendor marketing pages love to throw those numbers around, but they’re cherry-picked. What we can tell you is this: sending the right offer to the right customer segment is common-sense math. You stop giving away margin where you don’t need to, and you invest it where it might actually bring someone back.
Your One Takeaway
Pick one tool from above that fits your setup — Klaviyo or Mailchimp if you sell online, Square Loyalty if you’re brick-and-mortar, or even ChatGPT with a spreadsheet if you want to test the concept for the cost of a monthly subscription. Create just two segments: