Betting platforms

Bonus abuse prevention, multi-accounting & account integrity

Two hundred signups. Twelve devices. Keverd links every account back to the hardware — so promotional spend reaches real bettors, not coordinated farming rings.

Trusted by forward-thinking teams

3Payd logo
StartinEV logo
IBM logo
Duniafrika logo
Nailinq logo
Vabu logo
Grupchat logo
Lumeka logo
3Payd logo
StartinEV logo
IBM logo
Duniafrika logo
Nailinq logo
Vabu logo
Grupchat logo
Lumeka logo
3Payd logo
StartinEV logo
IBM logo
Duniafrika logo
Nailinq logo
Vabu logo
Grupchat logo
Lumeka logo

Overview

Kenya has one of the most active sports betting markets in Africa — millions of users and billions of shillings in annual handle. Scale makes it a large target for organised fraud.

The dominant problem is bonus abuse. Welcome bonuses, free bets, deposit matches, and cashback are the primary acquisition channel — and the primary fraud target. One device cluster can register hundreds of accounts, each looking like a new customer.

Beyond promotions: high-balance account takeovers, arbitrage bots, withdrawal fraud with stolen deposit credentials, and coordinated multi-account betting. Keverd addresses all of it at the device layer — registration through withdrawal.

The problem

What betting platform fraud looks like

Bonus abuse & promotion farming

Register account 1, claim Ksh 1,000 welcome bonus, drain it. Repeat with a new phone and ID from the same device — 50, 100, or 200 times. Each account passes new-user checks; only the device fingerprint connects them.

Multi-accounting for advantage

Hedge across accounts on the same event, bypass per-account limits, or coordinate clusters to move odds. Invisible per account — clear at the device cluster level.

High-balance account takeover

Winners and heavy depositors are prime targets. Phished credentials, login from a new device, immediate withdrawal. Keverd flags the device change at login before funds leave.

Arbitrage bot operations

Scripts monitor odds and place bets algorithmically. Machine timing and navigation — BOT and AUTOMATION flags let you limit or review systematic margin drain.

Withdrawal fraud

Stolen mobile money or card deposits, minimum bets to clear wagering, withdraw elsewhere. Deposit device vs. withdrawal device tells the story — flagged before payout processes.

Self-exclusion & age bypass

Excluded users re-register under new identities. Underage users borrow adult IDs. Device history links the new signup to the excluded or flagged account at registration.

How Keverd solves it

Five layers across the bettor journey

Registration, promotions, login, bets, and compliance — Keverd captures:

  • Device fingerprint at registration and every promotion claim
  • Account count per device — farm detection before SMS verification
  • Promotion claim history per device across all accounts
  • Login continuity and new-device alerts on high-balance accounts
  • Deposit and withdrawal device pairing — mismatched devices flagged
  • BOT and AUTOMATION on bet placement — timing and navigation patterns
1

Registration farms

Fingerprint every signup. Flag or block when one device exceeds your account threshold — before the account is created and before welcome bonus SMS goes out.

2

Promotion claim defence

Check every welcome bonus, free bet, deposit match, and cashback claim. Device already claimed this promo type on another account → hold before credit.

3

Login & high-value actions

Continuity on login. Extra check on withdrawals, account changes, and large bets — new device requires re-verification before completion.

4

Bet placement bots

BOT and AUTOMATION flags on bet submission. Your risk team decides: restrict stakes, require verification, or manual review.

5

Responsible gambling links

Connect new registrations to self-excluded or previously flagged devices — same hardware, different identity does not bypass exclusion.

Integration

Four touchpoints — registration & promos first

Activate registration and promotion claim first for immediate value. Web uses the Keverd script; mobile apps call the same signals server-side from your backend.

Account registration

Placement
Registration form page
Trigger
Registration form submission
Response
device_id, account_count_for_device, suspect_score, risk_tier, action_taken

Block or flag before account creation and before SMS verification — farm stopped at the door.

Promotion claim

Placement
Promotion claim or activation flow
Trigger
Bonus claim or promotion activation
Response
device_id, previous_claim_count, suspect_score, risk_tier, action_taken

Device that already claimed this promo type on any account → flagged before bonus is credited.

Login & account actions

Placement
Login, withdrawal, settings, large bet confirmation
Trigger
Login, withdrawal, account detail change
Response
device continuity flag, is_new_device, suspect_score, risk_tier, action_taken

High-balance threshold triggers mandatory re-verification on new devices.

Bet placement

Placement
Bet slip submission
Trigger
Bet submission
Response
bot_flag, automation_flag, suspect_score, risk_tier, action_taken

Surfaces arbitrage and automated betting for your risk team to act on.

Workflow

Bonus abuse farming

  1. 1

    Operator

    Registers dozens of accounts from one device cluster — different phone numbers and IDs each time.

  2. 2

    Keverd

    Fingerprints at registration and at every promotion claim. Links accounts to the same device.

  3. 3

    Your platform

    Farm threshold breached → registration blocked or promotion claim held before credit.

  4. 4

    Finance

    Promotional spend reaches real players. Blocked claims × average bonus = Ksh protected.

High-balance account takeover

  1. 1

    Attacker

    Obtains credentials for a high-balance account and logs in from a new device.

  2. 2

    Keverd

    Flags no continuity with the account’s known device history.

  3. 3

    Your platform

    Withdrawal or large bet held until re-verification — before processing layer.

  4. 4

    Bettor

    Legitimate user protected; attacker blocked before balance is drained.

Field guide

Reading Keverd flags on betting flows

FlagWhat it meansHow to use it
BOTSession behaviour matches non-human patterns.Review or restrict on bet placement. Strong signal with AUTOMATION.
AUTOMATIONBet or form flow driven by a script.Flag for risk review, stake limits, or account restriction.
USER_AGENT_SPOOFEDDevice misrepresents browser or OS.Common on farming stacks. Review at registration and claim.
TIMEZONE_IP_MISMATCHTimezone does not match IP location.Context only — VPN use is common. Do not block on this flag alone.
AD_BLOCKERAd blocker detected on device.Informational. Low weight in bonus abuse scoring.

First 30 days

What success looks like

  • Week 1bonus abuse blocks visible — farm and duplicate promotion claims
  • 100%of registrations and promotion claims fingerprinted
  • Day 3initial device cluster report after go-live
  • 30 daysKsh value of bonuses protected — blocked claims × average bonus
  • Day 30threshold review: account farm limit, promo rules, high-balance device policy

Lead the 30-day review with Ksh protected: blocked promotion claims × average bonus value. For most platforms that number exceeds subscription cost in the first month.

Default configuration

Tuned for betting platforms

Account farm>3–5 accounts per device before flagStart conservative — tune for shared-family-device false positives
Promotion claimBlock repeat claim per device per promo typeCatalogue promo types at onboarding
High-balance loginNew device → re-verify above agreed balanceThreshold set with your risk team
AD_BLOCKERIgnored in farm scoreLow correlation with bonus abuse

For platform teams

Starting the conversation

Angles that surface bonus abuse and multi-accounting quickly with risk and commercial teams.

The opening question

How much did you spend on welcome bonuses and free bets last month — and how much of that went to accounts that were actually the same operator on the same device? Most platforms know the second number is uncomfortable.

Device cluster analysis

Offer a passive cluster analysis on existing registrations — no live integration required. Show how many accounts map to how few devices. The data sells itself in a day.

Regulatory posture

Multi-accounting and bonus abuse are financial fraud and responsible-gambling exposure. Proactive device-level controls strengthen your position with regulators.

Lead with Ksh saved

At 30 days: blocked promotion claims × average bonus value = promotional budget protected. For most platforms this exceeds subscription cost in month one.

Onboarding

5–7 working days — registration & promos first

  1. 01Share URLs for registration, login, promotion claim, bet placement, and withdrawal
  2. 02Confirm web, mobile app, or both — app uses server-side API; web uses script
  3. 03Add JavaScript snippet to web touchpoints; configure API for mobile backend
  4. 04Configure webhook at each touchpoint on your backend
  5. 05Agree account farm threshold and promotion types for claim checks
  6. 06Set high-balance new-device re-verification threshold
  7. 07Confirm self-exclusion list format for device cross-reference
  8. 08Test run: register multiple accounts from same device; verify farm and promo blocking
  9. 09Brief risk team using workflows and field guide below
  10. 10Go live — Keverd monitors first 72 hours and delivers day-three cluster report

Known limitations

Interpret signals correctly

  • Keverd works at device and session layer — not betting patterns, odds movements, or stake analysis. Complements your own risk and trading systems.
  • Mobile app integrations use server-side API calls rather than browser script — slightly higher engineering lift; scoped at onboarding.
  • Shared devices (e.g. family on one phone) can trigger farm false positives — start with conservative thresholds and tune from live data.
  • One physical device per account farming is expensive and harder to cluster — less common locally but it exists.
  • Promotion claim history builds from integration go-live unless you supply historical registration data to bootstrap fingerprints.
  • Pre-sales cluster analysis on historical data can bootstrap device history before live integration — making week one significantly stronger.