RFM Analysis for Ecommerce: Segment Customers With AI

Your best customer got the same email as a stranger.

The RFM analysis ecommerce founders actually need fits on one screen. Score every customer on three things: recency (how many days since their last order), frequency (how many orders they've placed) and monetary value (how much they've spent). Give each a score from 1 to 5. The combination tells you who your best customers are, who's slipping and who's gone. Then each group gets a different message instead of the whole list getting one blast.

And here's the part most guides skip. You don't have to build it anymore. Shopify already scores your customers this way and sorts them into 11 groups. AI writes the segment for you from a sentence. Your only real job is deciding what to say to each group, and making sure the page they land on can close.

What is RFM segmentation, in plain numbers?

RFM segmentation is a ranking. It looks backward at what each customer already did and sorts them by it.

Here's how the three scores work on a store with a few thousand customers. The cutoffs below are a hypothetical example, because every store's cutoffs come from its own data.

Score	Recency (days since last order)	Frequency (orders)	Monetary (total spend)
5	0 to 30	6+	$400+
4	31 to 90	4 to 5	$250 to $399
3	91 to 180	3	$150 to $249
2	181 to 365	2	$80 to $149
1	366+	1	under $80

A customer who ordered 12 days ago, has 7 orders and has spent $520 scores 5-5-5. That's a Champion. A customer who spent $480 across 5 orders but hasn't bought in 200 days scores 2-4-5. Same money, very different conversation.

Shopify's own guide explains that it scores each measure from 1 to 5, then combines the recency score with the average of the frequency and monetary scores to assign the group. So recency carries the most weight. That's correct. A big spender who stopped buying is a problem. Treat them like one.

Recency is the only RFM score that changes while you sleep. Every day you don't act, your best customers quietly slide a group down.

Who are my best customers, and how do I find them in Shopify?

Open Analytics, then Reports, then the RFM customer analysis report. Shopify has already done the scoring. Pick a group, click Preview segment, and you have a list you can email. Whether that email belongs in a flow or a campaign depends on what the group just did.

The groups come straight from Shopify's customer RFM group definitions. There are 11 of them. Champions are customers "with very recent purchases, many orders, and the most spend." At risk are customers "without recent purchases, but with a strong history of orders and spend." Dormant is the bottom of the barrel: infrequent, low spend, not recent.

Here's where most founders stop. They see the Champions number, feel good, and go back to running ads.

I think the most valuable row in that report is At risk, not Champions. Here's the math.

Run it on a store like this, a hypothetical: 6,000 customers and $588,560 in revenue over the last 24 months.

Champions and Loyal (900) $243,000

At risk and Previously loyal (420) $130,200

Almost lost and Dormant (2,800) $117,600

Active, New, Promising, Needs attention (1,880) $97,760

6,000 customers. Champions and Loyal are 15% of customers. At risk and Previously loyal are 7%.

Champions and Loyal: 900 customers, $270 each on average, $243,000. That's 15% of customers producing about 41% of revenue.

At risk and Previously loyal: 420 customers, $310 each, $130,200. Seven percent of the list. About 22% of the revenue. And every one of them has stopped buying.

That's the group bleeding out while you're celebrating the Champions.

What should you send each RFM group?

This is where RFM stops being a report and starts making money. One offer per group. Never the same email to everyone.

RFM group	Who they are	Send this	Don't send this
Champions, Loyal	Recent, frequent, top spend	Early access to a new product, at full price	A sitewide discount
Active, Promising	Recent, a few orders	The accessory or refill that fits what they bought	Your full catalog
New	First order, very recent	How to get results from what they bought	A second offer on day 2
Needs attention	Recent, moderate	A bundle that raises the next order	A generic newsletter
At risk, Previously loyal	Big history, gone quiet	What's new since they left, then a reason to return	Nothing (the usual mistake)
Almost lost, Dormant	Low spend, long gone	One clean winback, then suppress	Weekly blasts that hurt deliverability

Shopify's guide lines up with this: it suggests inviting Champions to a product preview, recommending an accessory to Active customers and sending New customers product-use guidance.

Now watch what happens with the "20% off everything" blast, still on our hypothetical store.

Say 135 of your 900 Champions and Loyal customers order from that email at a $110 list price. They were going to buy anyway. That's what makes them Champions. The 20% discount hands each of them $22 they never asked for. That's $2,970 of margin given away to the people who'd happily pay full price.

The discount wasn't a reward. It was a tax on your best customers' loyalty.

Flip it to the At risk group. A winback that brings back 1 in 10 of those 420 customers at a $96 order is 42 orders and $4,032 from a single email. People who already trust you, coming home. If you want the sequence, the winback flow timing and discount rules are laid out step by step.

Treat the whole list the same and you're bribing the people who'd pay full price while ignoring the ones walking out the door.

How is AI changing customer segmentation in Shopify?

Two years ago, RFM meant exporting a CSV, sorting three columns, writing formulas for quintiles and redoing it every quarter. By the time you finished, the recency scores were stale.

Here's what's true today, checked against the vendors' own pages:

• Shopify assigns the groups for you. The RFM report sorts every customer into one of the 11 groups and refreshes as your store data syncs.
• Sidekick writes the segment. Shopify's help center says each segment is a ShopifyQL query, and you can ask Sidekick to write one from a plain-language description. Type "customers who last ordered more than 90 days ago and are subscribed to marketing emails" and it builds the filter and shows how many people match.
• Shopify predicts future spend. Predicted spend tier sorts customers into High, Medium and Low based on how often they buy, their average order compared with your store, their order count and how recently they ordered. It switches on once your store passes 100 sales.
• Klaviyo keeps the group live on every profile. Its RFM report writes a current and previous RFM group onto each profile and refreshes them every night, so a flow can fire the day a Champion slides to At risk. The catch: you need at least 500 customers who've ordered, 180 days of order history and a paid add-on.

So the work moved. Scoring is automated. Segment building takes a sentence. What AI can't decide for you is the offer, and whether the group is actually profitable. That last part matters more than people think. Polar's RFM guide points out that RFM ignores margin and discount dependence, so a discount-driven Champion can look like a top customer while losing money. Check each group against your contribution margin after ads before you crown anyone.

Why does RFM matter for profit, not just email?

Because keeping customers is where the profit sits. Bain's research on retention found that increasing retention by as little as 5% can boost profits by as much as 95%. That's the ceiling, not a promise, but the direction is the point. Every At risk customer you win back skips the acquisition cost you already paid once.

That's also why RFM belongs next to your LTV to CAC ratio. Lifetime value tells you what a customer is worth over time. RFM tells you who to talk to this week.

But here's the thing. Every segment you build sends people to the same place: a product page.

Your Champions click the early-access email and land on the new product's page. Your At risk customers click the winback and land on the page they bought from two years ago. If that page is a spec sheet, the segmentation did its job and the page dropped the ball.

I've watched it happen. Perfect list, perfect timing, and the page didn't answer the one question the returning buyer had: "What's different now, and why should I come back?"

Here's what a page that closes looks like. A bedding brand we worked with went from a conversion rate of 1.0% and an average order value of $125, which is $1.25 in revenue per visitor, to a conversion rate of 3.5% and an average order value of $231. That's $8.10 per visitor. On 10,000 visitors, $12,500 became $81,000. See the full case study numbers. Real client numbers, not typical results, and not a promise of what your store will do.

Same visitors. The page did the selling. Now send your segmented traffic into a page like that and watch what each group is worth.

What should you do next with your RFM report?

One move. Open Shopify's RFM customer analysis report today and preview your At risk group. Write down how many customers are in it and their total spend. Then send them one email this week: what's new since they last ordered, linked to a product page that answers why they should come back.

If you want to see how this ties into the rest of your numbers, the ecommerce profit analytics system shows where RFM fits next to profit per product, and our guide to asking your store data questions with AI covers the reporting side.

---

Book Your Profit Audit

Segments decide who you talk to. The product page decides whether they buy when they get there, and that's where most stores leak. Book a free profit audit and we'll show you where your page is losing returning customers, then show you how to rebuild a high-converting product sales page in less than 15 minutes.

Book Your Profit Audit →

Or go here to check it out → revenueflows.ai

P.S. Your Champions don't need a discount. Your At risk customers need a reason. Stop bribing the first group and start talking to the second.