RFM Segmentation model
Understanding Recency, Frequency, and Monetary scores and the 8 RFM segments LiftSignals assigns.
What is RFM?
RFM is a classic customer analytics framework that scores each customer on three dimensions:
- Recency — How recently did they buy? (1–5, where 5 is most recent)
- Frequency — How often do they buy? (1–5, where 5 is most frequent)
- Monetary — How much do they spend? (1–5, where 5 is highest spender)
Each dimension is scored 1–5 based on where the customer falls in your store's distribution. A score of 5 means "top 20% of your customers".
The 8 RFM segments
| Segment | Profile | Recency | Frequency | Monetary |
|---|---|---|---|---|
| Champions | Best customers — bought recently, buy often, spend most | 4–5 | 4–5 | 4–5 |
| Loyal | Regular buyers who spend well | 2–5 | 3–5 | 3–5 |
| Potential Loyal | Recent buyers with growth potential | 3–5 | 1–3 | 1–3 |
| New Customers | Bought recently for the first time | 4–5 | 1 | 1–5 |
| Promising | Recent buyers but not yet consistent | 3–4 | 1–2 | 2–3 |
| At-Risk | Used to buy often but haven't recently | 1–2 | 2–5 | 2–5 |
| Hibernating | Low scores across the board — may be lost | 1–2 | 1–2 | 1–2 |
| Lost | Lowest scores — haven't bought in a very long time | 1 | 1–2 | 1–2 |
RFM vs. Churn score: A customer can be in the At-Risk RFM segment (Recency 1–2) but have a low churn score if the model predicts they typically take long gaps between purchases. The churn model accounts for each customer's personal rhythm, while RFM uses store-wide percentiles.
How RFM updates
RFM scores update daily after the nightly sync. A customer's segment can change as their behaviour evolves — for example, a Champion who stops buying will gradually move through Loyal → At-Risk over time.
Using RFM in audiences
RFM segment is one of the most powerful filters in the Audience builder. Common uses:
- Export "Champions" to a Klaviyo VIP list for early access campaigns
- Target "Former Champions now At-Risk" for personalised win-back outreach
- Filter "New Customers" (R=5, F=1) for second-purchase nurture sequences
Learn more: Customer segmentation
Related in Intelligence Models
All 12 articles →How the intelligence models workAn overview of LiftSignals' signal models — how they train, how accurate they are, and when to trust them.2 min readChurn Detection modelHow the churn score is calculated, what the risk bands mean, and how to act on them.2 min readCLV Forecast modelHow LiftSignals predicts each customer's revenue over the next 12 months and the CLV tier system.1 min readWin-back Timing modelHow LiftSignals identifies the optimal day to re-engage a lapsed customer — and why timing is everything.1 min read
Still stuck?
Write to support@liftsignals.com — a human answers within four business hours.