Most e-commerce brands using Klaviyo segment their lists by product interest, acquisition source, or engagement level. Those are useful starting points. But they miss something that matters more for long-term revenue: how your customers actually behave over time. That's where RFM analysis comes in.

RFM stands for Recency, Frequency, and Monetary value. It's a framework from the 1990s that still works because customer behavior hasn't fundamentally changed. People who bought recently are more likely to buy again. People who buy often are worth more than one-time buyers. And people who spend more deserve different treatment than those who don't.

This guide walks through how to build RFM segments in Klaviyo, what to send each one, and how to maintain them over time.

What RFM analysis actually measures

Each dimension tells you something different about a customer's relationship with your brand:

Recency: when did they last buy? A customer who ordered last week is in a completely different headspace than one who ordered 8 months ago. Recency is the strongest predictor of future purchases.
Frequency: how many times have they bought? One order could be an accident. Three orders is a pattern. This separates trial customers from repeat buyers.
Monetary: what's their total spend? This helps you identify your highest-value customers, but it's most useful when combined with the other two dimensions.

The real value of RFM isn't in any single score. It's in the combinations. A customer who bought recently, buys often, and spends a lot needs a completely different email strategy than someone who spent a lot once, two years ago. Treating them the same is leaving money on the table.

In practice: RFM gives you a structured way to prioritize where your CRM effort goes. Not every customer deserves the same level of attention.

Why most Klaviyo accounts skip this

I've audited over 20 Klaviyo accounts in the past year. Fewer than a quarter had any form of RFM segmentation in place. The reasons are usually the same:

The team knows about RFM but hasn't had time to implement it
They tried building segments manually but the thresholds felt arbitrary
Klaviyo's predictive analytics features seem like they should handle this automatically

Klaviyo does offer predictive analytics that calculate expected LTV and churn risk per profile. These are useful, but they're black-box scores. You can't see the logic behind them or adjust the thresholds for your specific business. Building your own RFM segments gives you full control over the definitions and the ability to tailor your messaging per group.

What gets lost without RFM: you end up sending the same campaigns to everyone, discounting to people who would have bought anyway, and missing the window to re-engage customers before they fully lapse.

The five segments to build first

You can get very granular with RFM scoring, but for most brands, five segments cover the decisions that matter.

1. Champions

Bought recently, buy frequently, spend the most. These are your best customers. In a typical e-commerce account, they represent 5-10% of your list but 30-50% of your revenue.

The most common mistake with Champions: sending them the same discount-heavy campaigns as everyone else. They don't need 20% off to buy. They need recognition. Early access, VIP treatment, personal touches.

2. Loyal customers

Good frequency and decent spend, but their last purchase was 60-90 days ago. They haven't left yet, but they're drifting. This is the easiest group to retain because they already like your brand. A gentle nudge works better here than a hard sell.

3. At-risk customers

Previously solid buyers who haven't purchased in 90-180 days. This is where your retention budget has the highest ROI. They know you, they've bought before, and re-engaging them costs a fraction of acquiring someone new. But the window closes quickly. After 180 days, conversion rates drop sharply.

4. New customers

One purchase, within the last 30-60 days. The critical metric here is time to second purchase. According to data from multiple Shopify stores, customers who make a second purchase within 45 days of their first have 3x higher lifetime value than those who wait longer. Your post-purchase flow for this segment should be your highest-priority automation.

5. Lapsed customers

No purchase in 180+ days, regardless of what they spent before. Be honest about this group: not all of them are worth reactivating. Run a short reactivation sequence (2-3 emails), offer something meaningful, and if there's no response, move them to suppression. Continuing to email unresponsive contacts hurts your deliverability.

The pattern here: each segment needs a different message, a different cadence, and a different offer strategy. Uniform campaigns ignore this reality.

How to build RFM segments in Klaviyo, step by step

Step 1: Find your recency thresholds

Pull your order data and calculate the median time between purchases for repeat customers. If that number is 45 days, then 'active' means purchased within the last 60-90 days. 'At-risk' starts at 90-120 days. 'Lapsed' is anything beyond 180 days.

These numbers vary by category. A supplement brand selling 30-day supplies has a shorter cycle than a clothing brand. Use your actual data, not industry benchmarks.

Step 2: Set your frequency thresholds

For most e-commerce brands:

1 order = new/trial customer
2-3 orders = repeat customer
4+ orders = loyal customer

Check your own distribution. If 80% of your customers have only 1 order, your frequency thresholds will look different from a subscription brand where most have 6+.

Step 3: Define monetary tiers

Calculate your average customer lifetime value, then set thresholds at roughly 2x and 0.5x that number. If your average CLV is 150 EUR:

High value: 300+ EUR total spend
Medium value: 75-300 EUR
Low value: under 75 EUR

Step 4: Build the segments in Klaviyo

In Klaviyo, go to Audience > Lists & Segments and create a new segment for each group. Use these condition types:

'What someone has done' > 'Placed Order' > count (for frequency)
'What someone has done' > 'Placed Order' > date (for recency)
'Properties about someone' > total CLV or custom calculated property (for monetary)

Example for Champions: Placed Order at least 4 times over all time AND Placed Order at least once in the last 60 days AND total CLV greater than 300.

Adjust the numbers to get reasonable segment sizes. If your Champions segment has 8 people in it, your thresholds are too tight. If it has 40% of your list, they're too loose. Aim for 5-15% of your active list per segment.

Step 5: Verify and adjust

After building all five segments, check for overlap and gaps. In Klaviyo, a profile can belong to multiple segments. Make sure your segment conditions are mutually exclusive where they need to be, or at least that your flow filters handle overlap correctly.

The practical test: pull 10 random profiles from each segment and manually check if they feel right. If your 'Champions' segment includes someone who bought once for 15 EUR eight months ago, your conditions need work.

What to send each segment

The segmentation only matters if you actually change what you send. Here's what works:

Champions: new product launches before anyone else, loyalty program invitations, personalized recommendations based on purchase history. No discounts unless strategically necessary.
Loyal (drifting): a 'we miss you' email that acknowledges their history, paired with a product they haven't tried yet. Light incentive if needed (free shipping, small gift).
At-risk: a direct reactivation sequence. Lead with value (what's new, what they're missing), then offer an incentive in email 2 or 3. Keep it short. 3 emails max.
New customers: post-purchase education. Help them get the most out of what they bought, introduce your range, build the habit of opening your emails. This is relationship-building, not selling.
Lapsed: one strong reactivation attempt. If no response after 2-3 emails, suppress. Protecting your deliverability is more important than chasing cold contacts.

What changes: your discount strategy, your tone, your cadence, and your CTA all shift based on the segment. That's the whole point.

A real example: what RFM revealed for a fashion brand

A mid-size fashion brand I worked with had been running the same promotional calendar for two years: weekly campaigns, seasonal sales, and a birthday flow. Open rates were decent but revenue per email had been declining for six months. They couldn't figure out why.

We ran an RFM analysis on their Klaviyo data. The results explained the revenue decline immediately. Their Champions segment (top 8% of customers) had been receiving the same 20% discount emails as everyone else. Over time, these customers had learned to wait for promotions instead of buying at full price. Their average order value dropped 35% in 12 months while their purchase frequency stayed the same.

Meanwhile, their At-risk segment (about 22% of the list) was getting zero differentiated communication. The same weekly blast as everyone else. No acknowledgment that they'd been quiet for 4 months.

The fix was straightforward. Champions got moved to an exclusive early-access flow with no discounts. At-risk customers got a dedicated 3-email re-engagement sequence with a personal touch from the founder. Within 90 days, revenue per email was back up 18% and the At-risk segment had a 12% reactivation rate.

What this shows: the problem wasn't their email creative or their offer calendar. It was the lack of differentiation between customer groups. RFM made the diagnosis obvious.

Common pitfalls to avoid

After implementing RFM for multiple brands, a few mistakes come up repeatedly:

Setting thresholds based on gut feel instead of actual purchase data. Your intuition about what 'recent' means is probably wrong. Pull the numbers.
Building segments but not connecting them to different flows or campaigns. Segmentation without differentiated messaging is just extra work for no payoff.
Ignoring the monetary dimension entirely. Recency and frequency are easier to implement, so teams often skip monetary value. But it changes who your Champions are: a customer with 10 orders of 5 EUR each is different from one with 4 orders of 200 EUR.
Treating RFM as a one-time project. Customer behavior shifts. Seasons change. New products change purchase patterns. If you set it up once and never revisit, your segments will drift out of alignment with reality within 6 months.

The common thread: RFM works when it's connected to action and maintained over time. It fails when it's treated as a reporting exercise.

Maintaining your segments: a quarterly checklist

RFM segments are dynamic. Customers move between them as their behavior changes. Here's what to review every quarter:

Are segment sizes still reasonable? If Champions is shrinking, you might have a retention problem. If Lapsed is growing, your re-engagement flows need attention.
Do your thresholds still match your business? If your average order frequency has changed (seasonal shift, new product line), adjust accordingly.
Are your flows actually using these segments? It's easy to build segments and forget to connect them to different messaging. Check that each segment receives differentiated content.
How are the transitions? Track how many customers move from New to Loyal vs New to Lapsed. This tells you whether your post-purchase flow is doing its job.

This review takes about an hour per quarter. That's a small investment for the clarity it gives you about where your customer base is headed.

💡 SPARKCRM includes RFM analysis as part of its Klaviyo audit. It flags accounts where segmentation is missing or outdated, and shows which segments are underutilized in your current flows and campaigns. Request a free audit at sparkcrm.cc.