How Kombatix Works

Identity scoring at every stage of the fraud lifecycle.

Kombatix runs two complementary products on a shared identity infrastructure — PreAuth screens before the transaction; Defense scores when a dispute is filed. Both feed the Kombatix Network, which grows stronger with every customer.

The Cost of a Friendly Fraud Claim

Every claim costs $25 – $45. Where Kombatix intervenes changes the math.

A single $25 friendly fraud claim can cost a merchant anywhere from $25 to $45 depending on how it's handled. Defense intervenes before the chargeback — recovering the transaction and saving the downstream fees.

Cost decision tree showing four outcome paths for a single $25 friendly fraud claim without Kombatix, plus the Kombatix Defense intervention that preserves the transaction entirely.
Customer makes $25 purchase

Customer later disputes the charge

Net: −$25

Direct refund at first contact

Customer called, claimed fraud, got refund

Full transaction lost. No chargeback, no evidence, no deterrent.

−$25

Net: −$35

Alert-based resolution

Ethoca / Verifi alert fires post-dispute, pre-chargeback

$25 refund + $10 alert fee. Chargeback avoided but money gone.

−$35

Net: −$5

Chargeback filed — representment wins

You contest the chargeback successfully

Transaction retained but $5 chargeback fee assessed. Best case without Kombatix.

−$5

Net: −$45

Chargeback filed — representment fails

Contest rejected, money returned to cardholder

$25 refund + $20 chargeback penalty. Worst case. Merchant monitoring program exposure.

−$45

With KombatixNet: $0

Kombatix Defense deflection

Defense confirms friendly fraud at first contact; agent deflects

Transaction retained. Customer declines to escalate after identity-match evidence is presented. No chargeback filed.

$0

Loss pathsBest case without KombatixKombatix intervention

Pre-Authorization Screening

Is this transaction about to be trouble?

Before authorization completes, PreAuth checks the transaction's identity against the Kombatix Network. The check is lightweight — one POST, millisecond response, sub-cent cost on no-hits. When a hit returns, you know the identity has been confirmed as friendly fraud by at least one other Kombatix merchant. What you do next is your call.

Inputs

Name · email · phone · billing address · optional shipping · optional card token

Output

Hit / no-hit, severity level on hits, optional match metadata

Kombatix Defense — The Composite Score

Three engines, one composite score, one clear answer.

When a customer files a dispute — directly with you at support, or via their bank as a chargeback — Defense scores the disputing identity against the original transaction.

Engine 01 — Identity Match

Did the person who bought it just say they didn't?

The Identity Match engine cross-references the disputing party's submitted contact data — name, email, phone, address — against the data provided at the time of the original transaction, and against a verified identity network. Match strength determines how likely it is that the person claiming "I didn't buy this" is the same person who did.

Score range: 0–100. 100 = all dispute and transaction fields match the same verified identity. 0 = no identity match (possible stolen identity, synthetic fraud, or thin-file target).

Engine 02 — Identity Risk

Is this a trustworthy identity or one with a history?

The Identity Risk engine evaluates the identity itself, independent of the transaction match. It weighs identity age, fraud history connections, verified data point volume, activity consistency across data streams, and flags for suspicious patterns.

Score range: 0–100. Higher scores indicate a cleaner identity. Lower scores indicate fraud history, narrow verified footprint, or inconsistent data.

Engine 03 — Behavioral Risk

Does this person behave like a legitimate customer?

The Behavioral Risk engine analyzes activity patterns over time — transaction velocity, payment method stability, IP address and location changes, identity element modification frequency, chargeback and refund history. Legitimate customers show stable, consistent behavior. Fraud actors show sudden changes clustered around the disputed transaction.

Score range: 0–100. High scores indicate stable behavior. Low scores indicate velocity spikes, chargeback history, or multiple simultaneous risk indicators.

The Composite Defense Score

One number. Clear signal. Decide fast.

All three engines feed their scores into the composite Defense scoring model, which weights signals based on data completeness and cross-signal consistency. The result is a single 0–100 score your team — or your automated workflow — can act on without interpretation.

Score band interpretation and recommended action
Defense ScoreSignalRecommended Action
80–100Strong friendly fraud indicatorDeny refund. Provide identity-match evidence to customer and/or issuing bank. Close account if policy allows.
50–79Mixed signals — review warrantedManual review. Request additional customer verification before refund or representment decision.
0–49Low match / possible true fraudTreat as potential true fraud. Consider refund or escalation to fraud team. Defense billing is weighted toward actionable results — scores in this band contribute minimally to your monthly usage.

The AI Defense Narrative

Every report comes with a ready-to-use narrative.

Along with the score, every Defense report includes a plain-English narrative summarizing the result. Your agents copy it directly into support tickets, customer responses, or bank representment packages. No writing required, no interpretation needed.

Sample AI Defense Narrative

The disputing party submitted the name "John Smith" and phone "555-123-4567," both of which match the verified identity associated with the original transaction dated 2026-02-24. The transaction email "johnsmith@gmail.com" and the billing address "123 Main St, Anytown, CA 12345" also match this identity exactly.

Behavioral signals identified include use of 8+ distinct payment methods and transactions originating from 5+ locations over the past 90 days, both of which are consistent with identity churn rather than theft.

The composite Defense Score of 94.3 indicates strong evidence of friendly fraud. The dispute claim that the cardholder did not make the purchase is not supported by the identity data submitted.

Access Methods

Two products. Three ways to subscribe.

PreAuth is API-only. Defense scoring is available through two channels — API for engineering integrations and Web Portal for no-code support teams. Pick the products and channels your team actually uses; each is billed independently.

Defense API

From $75/month

REST endpoint for engineering integrations. Embed Defense scoring in your CRM, helpdesk, or chargeback workflow.

PreAuth API

From $49/month

REST endpoint for pre-authorization screening against the Kombatix Network. API-only — not available in the portal.

Web Portal Access

From $35/month

No-code Defense scoring in your browser. Built for support and CX teams without engineering capacity. PDF export, AI insights, searchable history.

The gap in your stack has a product. Three of them.

Get Instant Access