A broad email blast sent to an entire list, and a version of the same email sent only to a specific behavioral segmentation of that list, rarely perform the same — segmented sends routinely see meaningfully higher open and click rates in the marketing platforms that report on this, which is exactly why segmentation tools ship in the free tier of nearly every email platform on the market. This article is about that kind of segmentation: splitting a list of actual customers or contacts you already have, for the purpose of running better campaigns — not the broader strategic exercise of segmenting an entire market before you've built anything.
The distinction matters because the two use completely different inputs. Market segmentation, done before launch, relies on research, surveys, and assumptions about a market you haven't sold into yet. The segmentation covered here relies on data you already own — purchase history, email opens, site behavior — which makes it more accurate but only usable once you actually have customers to analyze.
RFM scores every contact on three numbers pulled straight from purchase history:
Score each contact 1–5 on all three (often using simple quintiles of your own data) and the combination sorts a list into usable buckets without any guesswork: someone scoring high on all three is a Champion; high frequency and monetary but slipping recency is At Risk; low on all three but still on the list is effectively Lost. Each bucket gets a different message — Champions get early access and referral asks, At Risk gets a win-back offer, Lost gets one final re-engagement attempt before suppression.
RFM works well once someone has purchase history; lifecycle stage covers the period before and around that. Typical stages: lead (subscribed, never purchased), new customer (first purchase, under 30 days), repeat customer (2+ purchases), and lapsed (previously active, gone quiet for a defined window — often 90 or 180 days depending on typical purchase cycle). Each stage has a different job to do: leads need trust-building content, new customers need onboarding, repeat customers need loyalty and upsell, lapsed customers need a reason to come back before they're suppressed entirely. Sending a "10% off your first order" code to a repeat customer, or a loyalty-points reminder to someone who's never bought anything, are two of the most common lifecycle-segmentation mistakes — both waste the message on the wrong stage of the relationship.
| Segment | Approx. Share | Campaign |
|---|---|---|
| Champions (high RFM) | ~8% | Early access + referral program |
| Loyal but slipping | ~15% | "We miss you" + personalized recommendation |
| New customers (<30 days) | ~10% | Onboarding sequence, no discount needed |
| Browsers, never purchased | ~40% | Educational content, lower-commitment offer |
| Lapsed (90+ days inactive) | ~20% | One win-back offer, then suppress if no response |
Treat these percentages as illustrative — actual distribution depends heavily on list age, acquisition source, and purchase cycle length — but the shape is typical: a genuinely engaged core is usually a minority of any list, and the last segment (lapsed) matters because emailing them repeatedly with no response quietly damages sender reputation for everyone else on the list.
Klaviyo, Mailchimp, and HubSpot all support RFM-style and behavioral segmentation natively, using conditional logic on purchase and engagement data already flowing into the platform from an e-commerce or CRM integration. For most small-to-mid-size lists, this removes the need for a data analyst — the segmentation logic is drag-and-drop, and segments can be set to recompute automatically rather than needing to be manually rebuilt.
None of this means demographic and psychographic data are useless once you have behavioral data — they're just better suited to a different job. Behavioral segments (RFM, lifecycle stage, triggers) decide who gets which campaign and when. Demographic and psychographic detail — age range, stated interests from a preference center, geography — help shape the tone and content within that campaign once the recipient list is already set. A win-back email to a lapsed Champion in a cold-weather region can lead with a different product than the same email sent to one in a warm-weather region, even though both are in the identical RFM segment.
It's easy to keep splitting a list — by RFM, by lifecycle, by product category, by acquisition channel — until each "segment" contains a few dozen people and no campaign is worth the setup time to send to it. A practical ceiling: if a segment can't justify its own subject line and offer, it should be merged into an adjacent one. Most lists are well served by somewhere in the range of five to eight active segments; beyond that, the maintenance cost usually exceeds the performance gain.
One more habit worth building in: RFM scores decay. A Champion from eight months ago who hasn't purchased since is often now an At Risk contact wearing an outdated label — segments need to recompute on a schedule, not get set once and left alone. A monthly or quarterly refresh, depending on purchase frequency in your category, is usually enough to keep the buckets honest without turning it into a constant maintenance task.
Segmentation groups contacts into buckets that get a shared message; personalization tailors content to an individual, often using data like their name, browsing history, or specific past purchases. In practice they layer together — a segment determines which campaign someone receives, and personalization fills in the specific details within that campaign.
When a segment is too small or too similar to another to justify a distinct message, it's excess. Most lists perform well with five to eight active segments; needing to build and maintain twenty distinct campaigns for twenty micro-segments usually costs more in setup time than it returns in performance.
A spreadsheet works fine for a first pass on a small list — export purchase data, calculate basic RFM scores with formulas, and manually tag contacts. It stops scaling once the list grows past a few thousand contacts or the segments need to update automatically as behavior changes, at which point built-in segmentation in an email or CRM platform saves significant ongoing manual work.