One profile per person, known or anonymous.

Every visitor, lead and customer resolves into a single record in one derived feature store, the same one your crew of agents reads from. Five dashboards, eleven RFM segments, and a 360° view of anyone who’s ever shown up.

5 dashboards, one feature store11 RFM segmentsPII-free by construction
A representative day

Five dashboards. One feature store underneath.

Overview, Visitors & Acquisition, Customers & Segments, Catalog & Logistics, Growth & Automations — each a different lens on the same derived data, with a global 7/30/90-day control bar across all of them.

Overview

Revenue, orders, AOV and eight more KPIs, each with sparklines and deltas against the period before.

13 tracked KPIs
Visitors & Acquisition

Every session on your own domain, split into anonymous, identified-but-didn’t-buy, and customer.

3 identifiability tiers
Customers & Segments

The 11-segment RFM model, a cohort-retention heatmap, and the 360° profile on anyone.

11 RFM segments
Catalog & Logistics

ABC-classed products, basket affinity, and RTO scored down to the pincode.

12 cities scored
Growth & Automations

Abandoned carts with a channel, angle and incentive already picked, plus what agents did on their own.

%-likely on every play
See where this data comes from: the Data Flywheel

Eleven segments. A map, not a spreadsheet.

Recency, frequency and monetary value place every customer on the 2026 RFM model. Each segment carries a count, an average lifetime value, and one suggested action — hover to see why.

162 CUSTOMERS · 11 SEGMENTSA representative store
Champions10% of customers

Bought recently, buy often, spend the most.

Customers16
Avg lifetime value₹24,600
Suggested action

Reward · early access · ask for reviews

Grounded onrecency scorefrequency scoremonetary score

Click anyone. Get the whole story.

Identity where it exists, hashed where it doesn't, value tiles, RFM and behaviour, risk bars, and a recommended action that always shows its evidence. The drawer opens the same way whether the row is a customer or a nameless visitor.

Customers & Segments

Click a row to open the 360° profile — customer or anonymous, same drawer.

Representative
Priya Naircustomer
Anonymous visitoranonymous

Priya Nair

Bengaluru · 560095 · via meta

sha256: a92f18e3c1… · 5b7d0a44e9

Champions

Lifetime

₹34,200

Predicted

₹52,000

Margin

₹17,800

RFM & behaviour

RFM score5-5-5
Orders7
Last seen6d ago
Sessions24 · 4m avg
Pageviews96

Risk

Churn risk8%

consistently active — no action needed

Delivery risk12%

0 RTO · 0 NDR · 5 COD (100% success)

Jasmine Green TeaSignature Trio

Invite to subscription · ask for a review

Because

  • high recency + frequency
  • ₹52,000 predicted

Filter it. Watch the count settle. Push it out.

Combine segment, identity and region — the match count updates live, and the export path resolves itself: a hashed CRM list for known customers, a pixel audience for anyone who never gave a name.

New audience

Combine filters — the size and activation method update live.

New audience
SegmentAt RiskTypeCustomerRegionNorth
At Risk · Customers · North India0
hashed CRM listSHA-256 email + phone, ready for Meta / Google / TikTok

Add an anonymous visitor to the filter and the activation flips to a website / pixel audience automatically. There is no identity to hash, so none is fabricated. Push either kind to your ad networks or export CSV.

Derived, not guessing.

Every agent in the crew reads the same structured features this page shows you, never a raw session log or an unmasked field, and states its evidence every time it acts.

What agents read
RFM segment + predicted lifetime value
Churn-risk score
Delivery-risk band — COD/prepaid history
Behavioural audience membership
Product affinity
What agents never touch
Raw session or clickstream logs
Unmasked email, phone or address
Payment details or IDs
Anything before it is aggregated into a feature
RTO Shield

Nudged a COD order prepaid before the label printed.

Because2 past NDRs on this customerpincode area-risk indexoffer-engine coupon cap
Win-Back

Sized a reactivation offer to a quiet customer’s segment.

BecauseRFM: at-risk60-day frequency capCLV floor respected

One record. Every agent reads it the same way.

The feature store behind these five dashboards is the same one Cart Recovery, RTO Shield and Win-Back read from — derived, evidenced, and yours to export.

Every audience you build here travels with your storefront code and customer data — a written export guarantee, no lock-in.