RFM Analysis for Ecommerce: Why Your Segments Are Lying

The quintile scoring, the four codes worth building, and the mutual-exclusivity fix Klaviyo tutorials skip

What Is RFM Analysis in Ecommerce?

Most RFM reports lie a little.

RFM analysis is a customer segmentation method that scores every buyer on recency, frequency, and monetary value, then groups them into named cohorts such as Champions or At-Risk. Klaviyo's built-in RFM report requires at least 500 customers who have placed an order; below that, build it manually.

Not because the three numbers are wrong. Because the segments most stores build on top of those numbers overlap, and almost nobody checks for the overlap before trusting the report.

Klaviyo's own RFM report lives under Advanced KDP or Marketing Analytics, and it needs 180 days of order history plus some customers with three or more orders before it calculates anything. Most Shopify stores we work with in Malaysia and Singapore sit below that bar, or don't want the advanced-tier price tag just to read a dashboard. So they build the same logic by hand, in the standard segment builder.

That manual build is where the mistake creeps in. Get the quintile scoring right and the codes worth building are simple. Get the recency ladder wrong, and every tier below the top double-counts customers who already live in the tier above.

How Do You Score Customers Into RFM Quintiles?

Break your list into fifths, not into guesses.

RFM quintile scoring means ranking every customer on recency, frequency, and monetary value separately, splitting each ranking into five equal groups, and scoring 5 (best 20%) down to 1 (worst 20%) on each axis. Combine the three scores into a 3-digit code, for example 555, out of 125 possible combinations. Focus on the edges, not all 125.

Export your order history and rank customers three separate times: most recent order first, most orders first, most total spend first. Cut each ranking into five even bands.

Every customer now carries three single-digit scores, read together as one code: recency-frequency-monetary. A 555 customer bought recently, buys often, and spends the most. A 111 customer is the opposite on all three.

Klaviyo tutorials commonly use an illustrative recency scale, a 5 for an order in the last 7 days, a 4 for 8 to 30 days, a 1 for anything past 150 days, but the middle scores are rarely defined and the whole scale is meant to be recalibrated against your own order cadence, not copied. A skincare brand with a 45-day repurchase cycle needs tighter bands than a furniture store where a repeat order every 8 months is normal.

Five times five times five is 125 possible codes. You do not need a segment for each one. Build for the edges: your best customers, and the ones about to disappear. Skip the murky middle.

Which RFM Codes Are Worth Building First?

Four codes carry almost all of the decision-making value.

The four RFM codes worth building as live segments are 555 (Champions), 155 (Champions you've lost), 515 (a single very large recent order, possibly bulk or B2B), and 511 (recent but very low spend, low future value). Everything else in the 125-cell grid blends these patterns or is too small a group to campaign against.

Code	Meaning	What to send
555	Champions, best on every axis	VIP early access, loyalty rewards, referral asks
155	High-value customer gone quiet	A targeted win-back with a real reason to return
515	One order, unusually large spend	Route to a sales or wholesale review
511	Recent but very low spend	Low-cost nurture only, or exclude from paid incentives

Source: standard Klaviyo RFM segment-building practice.

The buildable Champions segment, written as Klaviyo segment conditions:

Placed Order at least 7 times (over all time)
AND Placed Order at least 1 time in the last 7 days
AND total revenue is at least $500

That $500 monetary floor is an example, not a rule. One widely used Klaviyo walkthrough originally demoed this recipe at $121, then flagged its own number as too low. Set yours from your actual average order value and margin.

For consumable catalogues specifically, our beauty and personal care email marketing guide breaks down how the RFM model shifts for shorter repurchase cycles; this post covers the general-purpose build.

Why Do Your Recency Segments Overlap, and How Do You Fix It?

Here's the part every RFM article on the internet skips.

Recency segments overlap by default because "engaged in the last N days" conditions are cumulative, not exclusive: a customer active 10 days ago also qualifies for your 30-, 60-, and 90-day windows at once. The fix is a negative condition on every tier below the top, each one stating "0 times in the last [prior tier's bound]" to exclude the tier above it.

This is the payload of this post. Everything else is background.

Build a tier ladder for prospects (0 orders ever) and a matching ladder for customers (at least 1 order ever). Every tier in both ladders needs both consent conditions added separately, because Shopify's Accepts Marketing is true and Klaviyo's own email consent do not perfectly overlap.

The hottest tier, A+, covers 0 to 30 days and includes on-site behaviour:

Tier A+ (0 to 30 days)
  Placed Order = 0 times over all time        [use "at least 1" for the customer ladder]
  AND Accepts Marketing is true
  AND Klaviyo email consent is true
  AND ( Created in the last 15 days
        OR Clicked Email in the last 30 days     [Bot Click = false]
        OR Opened Email in the last 30 days      [Apple Privacy Open = false]
        OR Added to Cart in the last 30 days
        OR Started Checkout in the last 30 days )

Tier A, the 31 to 60 day window, repeats the base and consent conditions, then adds the negative condition that keeps it exclusive from A+:

Tier A (31 to 60 days)
  Base + consent
  AND ( Opened Email in the last 60 days OR Clicked Email in the last 60 days )
  AND Opened Email 0 times in the last 30 days
  AND Clicked Email 0 times in the last 30 days

Tier B, the 61 to 90 day window, follows the same pattern one step out, and so on through lapsing, lapsed, and a Hail Mary tier for the coldest contacts, saved only for your biggest promotions of the year. Skip the trailing negative condition and a customer who opened an email 12 days ago sits in three or four tiers at once, and every report built on those tiers is inflated by the overlap.

Does this sound like your store? Find out where you're leaking revenue. Take the free Revenue Score. 3 minutes. Free. No pitch.

What Consent and Filter Conditions Does Every RFM Tier Need?

Two consent boxes, two engagement filters, every tier.

Every RFM recency tier needs both Shopify's Accepts Marketing is true and Klaviyo's own email consent, added as two separate conditions, because the two systems do not perfectly overlap. Every open condition also needs Apple Privacy Open = false, and every click condition needs Bot Click = false, or the engagement isn't real.

Skip the double consent check and you message contacts who are opted into one system but not the other, which is a deliverability and compliance risk, not just a targeting miss.

Apple Mail pre-fetches images on delivery, which inflates open rates for anyone using Apple Mail regardless of whether they actually read the email. Filtering Apple Privacy Open = false on every open condition removes that noise. Bot Click = false does the same job for security scanners that click every link in an inbox before a human sees it.

Website-engagement events carry an expiry date, too. Drop Added to Cart and Started Checkout from Tier B (61 to 90 days) and every tier older than that. Past roughly 90 days, cart and checkout activity stop correlating with what a contact will actually do next, so keeping them in older tiers just adds false positives to a "warm" bucket that isn't warm anymore.

How Do You Validate That Your RFM Tiers Are Actually Exclusive?

Add up the tiers before you trust them.

The validation check for mutually exclusive RFM tiers is a sum check: add every tier's member count together, and the total should approximate your total consented pool with no double-counting. If the sum runs noticeably higher, at least one tier is missing its negative "0 times in the last N" exclusion.

Run this in Klaviyo's Audience Breakdown, or by exporting each segment's count and adding manually. Prospects and customers validate separately since they're two different base pools.

A sum that overshoots your consented list by 15 to 20% or more almost always traces back to one mistake: a tier built with only the positive engagement window and no negative exclusion added. Check the newest tier first. It's the one most likely to have skipped the exclusion, since it's easy to copy an existing tier's positive logic and forget the negative half.

If you're layering RFM on top of an existing setup, our ecommerce email segmentation guide covers the six broader revenue segments this tier ladder slots into.

How Much Incentive Does Each RFM Segment Actually Need?

Not every tier deserves a discount code.

The RFM incentive rule: the higher a customer's combined score, the less incentive they need to buy again, and the lower the score, the more incentive is required to bring them back. Below a monetary floor, commonly set somewhere in the $15 to $25 lifetime-spend range and calibrated to your margins, stop incentivising entirely.

A 555 Champion rarely needs a discount to buy again; they're already sold. Save the incentive budget for mid-tier customers who are recent and reasonably frequent but not yet high-spend, where a discount can genuinely tip a marginal decision.

The exclusion floor matters as much as the incentive scale. If a customer's lifetime spend sits below what it costs you to acquire and serve them, a 20% off code turns a low-value customer into a money-losing one. Build this as a suppression condition on your discount-send segments, not as a list you remember to check by hand.

Route the lapsing and lapsed tiers into a dedicated win-back email campaign rather than your standard promotional calendar. The incentive size and the message both need to differ from what an active customer sees.

Does RFM Still Matter Once You Have Predicted CLV?

Eventually, the manual build retires itself.

RFM stays useful up to the point where order volume is high enough for Klaviyo's Predicted CLV to do the same job automatically. Klaviyo requires 500+ customers with completed orders and 180 days of history before Predicted CLV activates. Past that, practitioners commonly treat roughly 10,000 customers as where it starts replicating manual RFM, though that figure is a rule of thumb, not a published Klaviyo cutoff.

Predicted CLV forecasts how much a customer will spend over the next year, using a model Klaviyo retrains weekly against your full order history. It answers a version of the same question RFM answers, how valuable is this customer going forward, without a hand-built quintile spreadsheet.

RFM still tells you something Predicted CLV doesn't: the behaviour pattern behind the value (recent-but-cheap versus infrequent-but-huge). For the machine-learning layer on top of this, our guide to AI segmentation for Shopify stores covers what tools like Klaviyo automate once you're past the manual-build stage.

Malaysia and Singapore complicate the recency math further. Consumable categories run shorter repurchase cycles than tutorial defaults assume, and a 30-day recency window means something different for a brand whose buyers concentrate spending around 11.11, 12.12, and the Raya or CNY calendar. A customer who bought during 11.11 and goes quiet for 45 days isn't cold; they're waiting for the next mega-sale. Build recency bands around your own promotional calendar, not a generic default.

Once your tiers exist, the real work is wiring them into flows. Our ecommerce email automation flows breakdown covers where each RFM tier plugs into a triggered sequence versus a manual campaign.

Frequently Asked Questions

What is RFM analysis in ecommerce?

RFM analysis scores customers on recency (days since last order), frequency (total orders), and monetary value (total spent), then groups them into cohorts like Champions or At-Risk. Klaviyo's built-in version requires 500+ customers with completed orders and 180 days of order history to activate.

How many RFM segments should a Shopify store build?

Most stores only need 3 to 4 live RFM segments: Champions (555), lost Champions (155), a bulk/B2B flag (515), and a low-value exclusion (511). Building all 125 possible RFM code combinations wastes setup time on cells too small to justify a dedicated campaign.

Why do my Klaviyo recency segments overlap?

Klaviyo's "engaged in the last N days" conditions are cumulative by default, so a customer active 10 days ago also qualifies for 30-, 60-, and 90-day windows at once. Fixing this means adding a negative "0 times in the last [prior window]" condition to every tier below the top one.

What's a good monetary threshold for a Champions segment?

There's no universal number; it depends on your average order value and margins. A commonly cited Klaviyo tutorial example uses $500 lifetime spend for a Champions segment, after the same source flagged its earlier $121 demo threshold as too low for most brands.

When should I switch from manual RFM to Predicted CLV?

Klaviyo's Predicted CLV activates once you have 500+ customers with completed orders and 180 days of order history. Many practitioners treat roughly 10,000 customers as the point where Predicted CLV closely replicates manual RFM scoring, though Klaviyo hasn't published that specific cutoff.

Keep Reading

• Ecommerce Email Segmentation: Stop One-Size-Fits-All
• AI Segmentation for Shopify: How It Works
• Win-Back Email Campaign: Bringing Lapsed Customers Back