Giselle · Segmentation and KPIs

Segments built on how players play, and what each one earned.

690+ behavioural indicators over 19 timeframes, segments that recalculate in real time, hourly or daily, every player clustered overnight and banded by churn risk, and what each segment's members did after they joined. The CRM tool you are shopping around for is already in it, in the same back office as the wallet.

Drive the back office ↓Send us a segment

Keeping your platform? Giselle Cloud computes the same indicators from your events →

01 · The back office

The screens your CRM team reads before it builds anything.

Five screens from the product, with demo data, and the controls work. Save a churn band as a segment, sort segment health, change a rule and count who matches, open a player's KPIs. On a call we open the real one, on your brand.

Back office · Demo Casino

Player Behaviour

How players play, read from ten behavioural features, clustered overnight and banded by churn risk.

Players clustered48,210Fit last updated 31/10/2026 03:40
Behavioural clusters6Named by your team
Players scored51,930Scored 31/10/2026 04:10
High churn risk3,486Banded, not a probability
Nothing selected. Choose a cluster, a churn band or a scatter region.

Cluster distribution

Players per cluster. Select one to build a cohort.

Churn risk

Banded high, medium and low. Select a band to build a cohort.
Top drivers among high-risk players

Sessions 7d1,920 players

Deposit gap, median 30d1,310 players

Withdrawal to deposit ratio 7d640 players

Behavioural scatter

Any two features, one dot per sampled player, coloured by churn band.
XY
HighMediumLowNot scored140 of 51,930 shown

Segment Intelligence

Where players moved this month, and what each segment's members have done since they joined it.

Month
Segments reported6compared with September 2026
Segment memberships9,224Memberships, not distinct players
Memberships gained3,846New entries this month
Nothing selected. Choose a segment's health card to build a cohort.

Segment health

October 2026 against September 2026. Select a segment to build a cohort.Sort

Lift since entry

Counted from each member's own entry date, not over their lifetime.Metric
High value, casino1,031 of 1,048 members measurable€3,912,400
Reactivated depositors3,962 of 4,106 members measurable€1,284,600
Sport first bet, 7 days2,187 of 2,210 members measurable€412,800
Dormant, 30 to 60 days5,711 of 5,930 members measurable€228,900
Registered, no deposit9,406 of 9,812 members measurable€96,300
Bonus-only players, watchlist380 of 388 members measurable€61,200

Create New Segment

The details, the rules on the behavioural indicators, then a review before it goes live.

Lapsing casino players
Played casino this month, no deposit in the last 7 days.
Group 1Every rule in a group must be true
Casino Place Turnoveris greater than or equal to500Last 30 Days
ANDTotal Deposit Countequal0Last 7 Days
ANDAccept PromotionsequalYesNo period

Eligibility Filters

Reusable rules that decide who a CLOE flow may reach. Build one once, point any flow at it.

IDFilter NameRulesCreatedLast Updated
214Verified, promotions acceptedKYC Status is VerifiedAND Accept Promotions equal Yes12/03/202628/10/2026
209Active real-money depositorsAccount Status is ActiveAND Total Deposit Count, Last 30 Days, is greater than or equal to 102/04/202621/10/2026
203SMS reachableMobile Phone Verified equal YesAND Accept Communication SMS equal Yes18/05/202618/05/2026
198Hand-picked VIP listUsername is in csv (42 values)09/06/202630/09/2026
191Casino-led playersLead Product is Casino, Live Casino27/06/202614/10/2026
187Bonuses acceptedAccept Bonuses equal Yes03/07/202603/07/2026
Show 50 entriesShowing 6 entries

Customers

Every player's KPIs by the hour, the day, the month or the year, or over their lifetime.

External IDUsernameFirst Name
Advanced Filters
IDExternal IDUsernameFirst NameLast NameEmailCountryAccount Status
100318DC-7318demo.player.318LenaDemolena.demo@example.comDEACTIVE
100316DC-7316demo.player.316MilaDemomila.demo@example.comBGACTIVE
100315DC-7315demo.player.315NoahDemonoah.demo@example.comCAACTIVE
100314DC-7314demo.player.314EvaDemoeva.demo@example.comFITIMEOUT
100313DC-7313demo.player.313RaviDemoravi.demo@example.comDEACTIVE
100296DC-7296demo.player.296AinoDemoaino.demo@example.comFIACTIVE
Show 50 entriesShowing 1 to 6 of 48,210 entries

Read players by how they play. Ten behavioural features per player, from the rhythm of their deposits to their stake after a loss, clustered overnight and banded by churn risk. Pick a cluster, a band or a region of the scatter and save exactly those players as a segment.

What each segment did this month, and since. One card per segment with its members, new entries and the change against last month. Lift since entry counts the turnover, GGR, deposits and bonus released each member generated after joining.

A segment is rules on indicators the platform already computes. Pick an indicator, an operator, a value and a period, combine them with AND and OR, and count who matches before you save. It recalculates in real time, hourly or daily, as you choose.

Who a journey may reach, written once. An eligibility filter is a named set of rules on the same indicators as a segment, and CLOE flows apply it. Clone one to start the next.

One player’s KPIs, over any period. Logins, sessions, deposits, withdrawals, turnover, revenue and margin, hourly, daily, monthly, yearly or lifetime. Every figure on it is also an indicator a segment rule can use.

Demo data. Every label is a control in the shipped product.

Giselle V2, the whole suite →

02 · What is in it

Segmentation and KPIs, in three parts.

  1. 01SegmentationSegment health and lift since entry on the overview, the segments list with its builder, and the eligibility filters CLOE flows apply.Overview · Segments · Eligibility Filters
  2. 02Platform AnalyticsBehaviour across the brand: clusters, churn bands, the behavioural scatter and month-over-month cohort retention. Then any one player’s KPIs.Overview · Customers KPIs
  3. 03Segments ReportsEach segment over time, its KPIs, and segments side by side.Segment Overview · Segment KPIs · Segment Comparison

03 · Standard, and ours

Every CRM has segments. Here is what is under these.

Standard. Here too.

  • Segments built from rules on player data
  • Scheduled and real-time recalculation
  • Static lists and dynamic segments
  • Filters reused across campaigns
  • A KPI view per player
  • Segment size, entries and exits over time
  • Exports

Different here.

  • The indicators are already computed.

    690+ behavioural indicators across 19 timeframes, from the current hour to lifetime, in one catalogue that segments, campaigns and eligibility filters all read. Nobody writes a query, and the rule guide in the builder says how each one is calculated.

  • Behaviour, not only totals.

    Ten features per player describe how they play: the rhythm and size of their deposits, how long they stay, how wide they play, how their stake moves after a loss, how much of what goes in comes back out, how many transactions each euro of NGR takes. Players are clustered on them overnight, and churn risk is banded high, medium and low.

  • Any selection becomes a segment.

    A cluster, a churn band, a region of the scatter or a segment’s health card saves as a segment in one step: a snapshot of exactly those players that never recomputes. Launch a CLOE flow on it from the same bar.

  • A segment’s worth, from the day each player joined.

    Lift since entry counts what members generated after they joined the segment, not over their lifetime: turnover, GGR, deposits or bonus released. Members, new entries and the change against last month sit on one card per segment.

  • Margin counts what a player costs.

    Margin per player counts gaming tax and payment costs, because on Dench Core the suite reads the same wallet as everything else. The player whose revenue looks healthy and whose payment bill eats it shows up as what they are.

04 · The record

The catalogue under every segment.

690+behavioural indicators, one catalogue for segments, campaigns and eligibility
19timeframes per indicator, from the current hour to lifetime
10behavioural features per player, recomputed overnight
30+regulated brands have run Giselle on the platform since 2017

Ran on it

  • 138.comUnited Kingdom, 2017 to 2020, under our UK Gambling Commission licence.
  • Bet99Canada from 2020, licensed in Ontario when the regulated market opened.
  • KlondaikaLatvia, the full platform including CRM, until Entain acquired it in 2022.
  • GGPokerA global poker brand, on the platform in 2019.

Running on it now

  • Spins.lvLatvia
  • EverbetBulgaria
  • SlotinoBulgaria, the brand we built and ran ourselves
  • AladinsBulgaria
  • QbetBulgaria
  • Slots99Canada, Europe and Latin America
  • ISO/IEC 27001 since 2017
  • UKGC licensed software supplier
  • GLI certified in three jurisdictions
  • Bulgarian manufacturer licence

The licence register, market by market →

05 · How it is bought

Four ways to take it. One price list.

The usual

Platform

You hold the licence and run the operation: marketing, support, payments relationships, compliance. We deliver and run the technology: the platform with its CRM and gamification suite, integrations, front end, hosting, releases, support. Your provider contracts or ours.

Custom

Enterprise

The platform built out to your market and your specification: dedicated environment, custom front end, features against your requirements. The way the largest accounts have been delivered for ten years.

Our licence

White label

Our infrastructure end to end, run under the group’s Curaçao setup: licensing, provider contracts, payments. You bring the brand and the marketing.

Without the platform

Giselle Cloud

The engagement suite, journeys, loyalty, jackpots, tournaments, bonusing and the analyst, on the platform you already run.

You pay for the modules you take, on the platform you already run.

Take the whole suite or the part you are missing. IN-GRID Loyalty and the Jackpot Server are licensed on their own as well. Connecting your platform is typically a few engineer-days on your side, and the mapping is ours.

To quote, we need four things

  1. The platform you run on today.
  2. The modules you want: the whole suite or a part of it.
  3. The brands and markets it would run on.
  4. How your platform emits events today: a transaction feed, webhooks, or neither.

06 · Asked first

The three questions that arrive before the demo.

How fresh is a segment?

You choose, segment by segment: real time, hourly or daily. A real-time segment is re-evaluated with every customer transaction; hourly or daily suits most lists.

A segment saved from a selection on the behaviour or intelligence screens is the other kind: a snapshot of exactly those players, which never recomputes.

Does the churn band predict who will leave?

It reads players as they go quiet, while a win-back journey can still reach them, and bands them high, medium and low. A raw score means different things in different markets, so the screen shows the band and the top drivers behind the high band, never a probability.

Does it run on our platform?

Yes, as Giselle Cloud: your platform posts its events to one endpoint, and we compute the same indicators from them. For margin, you give us the cost inputs.

Giselle Cloud →

The six questions we get asked →

Send us the segment you cannot build today.

The players you want to find, in your own words, and what you run on today. We build it as rules on demo data and show you who it catches, on a call.

Send us a segment