What is RFM segmentation, and how to use it without a data team
Recency, frequency and monetary value, sorted into groups a marketer can actually write an email to.
RFM stands for recency, frequency and monetary value. It sorts a customer list by behaviour, not by who somebody is, and it's older than email. It still works, as long as what comes out of it is a handful of groups and not a grid.
What RFM actually measures
Recency is how long it's been since somebody last bought. Frequency is how many times they've bought. Monetary value is what they've spent with you in total. Three columns, and they're already sitting in your order data.
The method comes out of catalogue mail order, where postage cost real money and somebody had to decide who was worth a stamp. Email took the postage away. It didn't take the decision away, because attention is the thing you're spending now.
What each letter is good for, and what it isn't
| Letter | The question it answers | What it changes |
|---|---|---|
| Recency | Are they still with us? | Whether to sell, remind or try to win back |
| Frequency | Is this a habit or an accident? | Whether a subscription or a reorder prompt makes sense |
| Monetary | How much is this relationship worth? | How much effort, discount and patience it justifies |
Frequency and monetary value tell you who to look after. Recency tells you when to act. That's why they're worth keeping apart instead of blending them into one score.
Recency does most of the work
If you only ever cut the list one way, cut it by recency. Somebody who bought last month and somebody who bought two years ago need different sends, and no amount of lifetime value changes that. A big spender who's gone quiet is a winback problem. Not a VIP.
- Bought in the last month, and still opening
- Bought this quarter, quiet since
- Bought within the year, no recent engagement
- Older than a year, still opening
- Older than a year, opening nothing
Name four or five groups, not a grid of a hundred
The classic method scores each letter one to five and multiplies out into a grid of cells. Tidy on a slide. Useless on a Tuesday, because nobody writes a hundred emails. Collapse it into groups you can name out loud: best customers, regulars, one-time buyers, lapsing, gone.
Two ways RFM goes wrong in practice
The first is scoring people against each other. Quintiles move every time the list moves, so a customer can drop a grade without changing a thing. Fixed thresholds in months and orders are less elegant and far easier to act on.
The second is treating the segment as the campaign. A group's only worth building if a different email goes to it. Send lapsing customers the same thing as everybody else and you've built a report, not a segment. That group belongs in the winback flow, and the rest of the sorting belongs in lifecycle strategy.
Theo Tziapouras
Founder and strategy at Engage Commerce, the ecommerce agency for 7 and 8 figure DTC brands.
FAQs
What does RFM stand for?
Recency, frequency and monetary value. Recency is time since the last order, frequency is how many orders they've placed, and monetary value is total spend. Your shop platform already holds all three, so there's nothing extra to track before you can build the segments.
Do I need a data team to run RFM segmentation?
No. Klaviyo, Shopify and most email platforms will build recency, frequency and spend conditions out of your order data with no code at all. What you need is a decision about thresholds and the discipline to write a different email for each group. The analysis is the easy half.
Is RFM better than lifetime value?
They answer different questions. Lifetime value estimates what somebody will be worth over time, which is what you need to decide what you can pay to acquire them. RFM describes where they are today, and that's what decides this week's send. Keep both.
Sort by behaviour, then send something different
RFM survives because it asks the only three things a shop can answer about a stranger: when, how often, how much. It stops working the moment the output is a chart instead of a send. Cut the list into groups you could describe to a colleague in one sentence, write a different email for each, and leave the scoring grid to the slide deck. The glossary has the neighbouring terms if you want them.
Terms, defined properly
One term, answered in full, then linked into the glossary entry.
What is repeat purchase rate, and what actually moves it
One number, measured over a window you fix, that says whether the first order was the start of anything.
Read it
What is list segmentation, and when does it start paying?
Splitting one audience into groups you would write to differently. The definition is easy. Knowing which groups earn their keep is not.
Read it
What is customer lifetime value in ecommerce, and what should you do with it
The number that decides what you can afford to pay for a customer, and the version of it most brands quote wrongly.
Read itWould you rather this was just handled?
Bring your Klaviyo account and the thing annoying you most. We will tell you what we would fix first, on the call, before you spend anything.
Book a call with our founder.
We're all about relationships built on trust, mutual respect and a shared vision for success. If that sounds like your vibe, let's make some waves together 🌊
