NewWeather Demand Modelling is live. Forecast demand before it arrivesWeather Demand Modelling is live

See what each customer is worth, and what state they're in.Ranked by true margin, segmented to act on.

Customer Value & Segmentation ranks your whole base by lifetime contribution margin and segments it on margin, not revenue: who carries the profit, who is quietly slipping, and what each customer is likely to be worth next. Identity tells you who they are; this tells you what state they are in and what they are worth. We tell you who matters; we do not run your store.

Customer Value & SegmentationDemonstrative data
Profit concentration · cumulative contribution margin vs customers
0%25%50%75%100%Gini 0.620% of customers50% of customers100% of customers
The most valuable 10% of customers carry 55% of all lifetime contribution margin, and the bottom band actually destroys it. Revenue charts cannot show this; only a margin-true view exposes who really funds the business.
Total customers
19,217
lifetime base
Lifetime CM margin
$4.2M
true contribution
RFM segments
12
scored on margin
Fake VIPs
847
drop a tier on margin
CM at risk
$145k
slipping or dormant
Top-decile margin share
55%
held by top 10%
Margin intelligence from the award-winning team behind 100+ brands
PanasonicRainCoCheapest LiquorKing CoolingAuto ComfortiHeat & CoolAACAEInsider Experience SportsInterosPeter JacksonLa TrobeToy WorldForesightOncoreMindshopSurface SpectrumNAWTeafyEZI TagPanasonicRainCoCheapest LiquorKing CoolingAuto ComfortiHeat & CoolAACAEInsider Experience SportsInterosPeter JacksonLa TrobeToy WorldForesightOncoreMindshopSurface SpectrumNAWTeafyEZI Tag
Purest SolutionsCuratedVoir VodkaSwing & ServeNewLeafNavigataFirst EnergyHaldatecSirius GreenElite ElectricalMobile SkipsBrowedAsenoMGIDCRebateM8TMJ TutoringAcademic TutorsUCSPurest SolutionsCuratedVoir VodkaSwing & ServeNewLeafNavigataFirst EnergyHaldatecSirius GreenElite ElectricalMobile SkipsBrowedAsenoMGIDCRebateM8TMJ TutoringAcademic TutorsUCS
What it answers

The questions a revenue dashboard can't.

Not how many customers you have. What each is worth in margin, which segment they sit in, who is slipping, and who to act on.

01
How concentrated is our profit, and who destroys it?
Lifetime contribution spread across the base, value tiers from VIP to negative, so the few who fund you and the customers who erode margin are both obvious.
02
Who looks like a VIP on revenue but isn't on margin?
The RFM cube re-scored on contribution surfaces the discount-funded fake VIPs, customers who drop a segment the moment returns and discounts are costed.
03
What state is every customer in right now?
A seven-state lifecycle board, cadence-anchored to your own reorder gap, weights each state by the contribution it holds, so health is measured in margin, not headcount.
04
Who is slipping, and where is the margin leaking?
State-to-state migration shows who slid down a state over the period and the contribution that moved with them, so the biggest leak is a list you can act on.
05
What will each customer be worth next?
A backtested predicted-value model gives each customer a probability of being alive and an expected residual contribution, so you save the high-value ones before they go.
06
Who should we protect, grow, win back or suppress?
Every tier, segment and state is a ready audience, with a suppress move no revenue tool ships, so retention effort lands on the customers that actually fund you.
Margin-true value tiers

Not every customer is worth the same.

Ranking customers by revenue rewards the wrong people. A high spender who returns half their orders and only buys on discount can contribute less than a quiet full-price regular. Tiering the base on true lifetime contribution puts each customer where they really sit, from VIP to outright drag, and turns each tier into an audience you can defend, grow or quietly let go.

  • Tiered on margin. Value measured in contribution, not gross spend
  • Customers vs profit. Each tier's share of the base against its share of margin
  • Every tier an audience. Exportable lists to protect, grow or suppress
See your value tiers
Customer Value & Segmentation · value tiersDemonstrative data
Margin-true value tiers · share of customers vs share of contribution
VIP8% cust41% marginHigh15% cust30% marginMid28% cust21% marginLow35% cust12% marginNegative14% cust-4% margin
VIPs are just 8% of customers but 41% of contribution; the negative tier is 14% of customers and an outright drag. Tiering on margin rather than spend puts a discount-eroded big spender where they actually belong.
RFM segmentation

Twelve segments, scored on margin not revenue.

Every customer is placed in a twelve-segment RFM cube, from Champions to Lost, on recency, frequency and value. The difference is the value axis: we score it on true contribution, not revenue. Flip the toggle from revenue to margin and the discount-funded fake VIPs hollow out, customers who look like Champions until their returns and discounts are costed, then drop a tier. It is the segmentation your email tool can't build.

  • Margin as the M. The monetary axis is contribution, not gross spend
  • Fake VIPs exposed. Who reclassifies down the moment margin is counted
  • Every segment an audience. Champions to win-back, ready to export
See your RFM cube
Customer Value & Segmentation · RFM cubeDemonstrative data
12-segment RFM · scored by contribution margin
RevenueCM1 ✓
Champions18%3,247 cust · of CM1
Loyal14%5,832 cust · of CM1
Potential Loyalist6%2,156 cust · of CM1
Promising4%1,840 cust · of CM1
New2%1,823 cust · of CM1
Need Attention5%2,410 cust · of CM1
Can’t Lose Them9%1,180 cust · of CM1
At Risk7%4,387 cust · of CM1
About to Sleep3%2,940 cust · of CM1
Hibernating1%6,521 cust · of CM1
One & Done1%8,943 cust · of CM1
Lost-1% cust · of CM1
847 fake VIPs. Customers who segment as Champions on revenue but drop a tier the moment discounts and returns are costed, a $267k revenue-to-margin gap. Revenue segmentation rewards them; margin segmentation tells the truth.
Lifecycle & migration

Know what state every customer is in, and who's slipping.

Each customer sits in one of seven lifecycle states, anchored to your own reorder cadence rather than a generic 30-day rule, and each state is weighted by the contribution it holds. The flagship is migration: who slid from Loyal to At-risk over the period, and the margin that moved with them. It turns base health from a headcount into the single biggest leak you can act on this week.

  • Cadence-anchored. States tuned to how often your customers actually reorder
  • Weighted by margin. The equity slipping, not just a count of quiet accounts
  • Migration as a list. The Loyal-to-At-risk leak, ready to win back
See your lifecycle board
Customer Value & Segmentation · lifecycle boardDemonstrative data
Seven-state lifecycle · customers and contribution by state
New1,823$48k · 8%
Champions3,247$1.7M · 89%
Loyal5,832$1.2M · 76%
Potential2,156$320k · 38%
At risk4,387$260k · 12%
Hibernating6,521$110k · 2%
Lost18,221-$60k · <1%
Biggest leak: Loyal → At-risk. 1,247 customers and $134k of contribution slid down a state in the last 90 days. The migration view hands you exactly who, so you win them back before they hibernate.
Predicted value

See what each customer will be worth, and who to save first.

Backtested on your own history, a predicted-value model gives every customer a probability of still being alive and an expected residual contribution. Plot the two and the urgent quadrant appears: high future margin, low probability of return. That is your win-back queue ranked by recoverable contribution, not a guess, and the model's accuracy is shown openly so you can trust the number.

  • Forward, not backward. Expected residual contribution per customer, in margin
  • The save-first quadrant. High value, low probability of return
  • Backtested. Accuracy reported against your own held-out history
See your predicted value
Customer Value & Segmentation · predicted valueDemonstrative data
Predicted value · probability of being alive vs residual contribution
at-risk <40%median CMSave firstprobability alive →residual contribution →
$95k of recoverable margin sits in the save-first quadrant, 8,754 high-value customers whose probability of returning has dropped below 40%. They are the win-back queue, ranked by the contribution you stand to lose.
Where the numbers come from

Built on your store, reconciled to true profit.

Every value, segment and prediction is built from real orders and real costs, so the ranking is margin, never gross spend.

Orders & revenue

Every order, refund and discount per customer, so lifetime value and every segment are net of returns and the discounts a customer actually used.

ShopifyWooCommerceBigCommerce

True cost

COGS, shipping and fees on every order, so the monetary axis of the RFM cube and every lifetime value is contribution, not revenue passed through.

Cost feeds3PL / freightPayment fees

Owned audiences

Every tier, segment and lifecycle state syncs to your email tool as a ready audience, so a segment becomes a campaign your team controls.

KlaviyoEmail / SMS
Every figure is reconcilable to its source, a customer-value number a CFO signs off.
The team behind it

Built by an award-winning analytics team.

Margin OS comes from Blufire, trusted by 100+ mid-market and enterprise brands and recognised across the APAC and Global Search Awards. The same people now model your margin.

100+
Brands served
$5M-$1B
Turnover served
4
Industry awards
100 Fast StartersAPAC Search Awards 2025 WinnerGlobal Search Awards 2025 FinalistGlobal Agency Awards 2025 Finalist
Questions

The things buyers ask.

Those measure value in revenue or gross spend and segment on rules like last purchase or total spend. We measure value in true contribution margin and make margin the monetary axis of the RFM cube, so a big spender who erodes margin ranks below a quiet full-price regular, and the discount-funded fake VIPs reclassify down where they belong.
When the RFM cube is scored on revenue, discount-funded customers look like Champions. Flip the monetary axis to contribution margin and they drop a tier, because once returns and discounts are costed they aren't actually your best. We surface exactly how many reclassify and the revenue-to-margin gap behind them.
This is the behavioural and financial lens: what state a customer is in (lifecycle), what they're worth (margin and predicted CLV) and which segment they sit in. Persona Identification is the identity lens: who they are in life and where they live. They are orthogonal and pair, an RFM segment cuts across every lifestyle audience.
A purchase-and-survival model, fitted on your own order history, gives each customer a probability of being alive and an expected residual contribution. We backtest it against held-out history and report the accuracy openly, and where a customer has too little history to model we show an honest null rather than a fabricated number.
No ad connection is required, the whole read is built from your store orders and costs, live on day one. Every value tier, RFM segment and lifecycle state is a ready audience: export the VIP list to protect, the lapsing high-value list to win back, or the negative-margin list to stop subsidising, and sync it to your email tool. You approve and send; we never touch your store.

See what your customers are really worth.

Connect your store and your costs. We rank every customer on true lifetime contribution, segment them on margin, predict what they'll be worth next, and hand you the audiences to act on.

01

Connect your store

Orders, refunds and discounts. No data team.

02

Load your costs

COGS, shipping and fees, mapped with us.

03

Rank & segment on margin

Value tiers and a margin-true RFM cube.

04

See state & predicted value

Lifecycle, migration and predicted CLV.

05

Get the audiences

Protect, win back and suppress, exported to send.

Book a demo Talk to an analyst