Reduced support volume by 25-35% (saving 1.5-2 FTE in support operations) by consolidating fragmented identity signals into a single decision layer.
25-35%
Reduction in support volume
Authentication state, KYC verification, user roles, account status, and risk signals each lived in separate systems with no shared model. Users experienced unpredictable access errors: verified in KYC but blocked by mismatched account status, MFA triggering inconsistently, permissions misaligned with roles. Every conflict became a support ticket.
iPROMOTEu operates a complex B2B2C platform serving a network of promotional products affiliates, each with distinct access patterns, compliance requirements, and operational roles. When I joined as Senior Product Manager in March 2024, the platform's identity layer was a patchwork of independent systems that had grown organically over years of incremental development.
The core problem was structural: five distinct systems -- authentication, KYC verification, role management, account status, and device/risk signals -- each maintained their own version of the user. These systems were not synchronized. When one system updated a user's state, the others did not automatically reflect the change. The result was a class of errors that were nearly impossible to diagnose: a user could be fully verified in the KYC system but blocked by an outdated account status flag; MFA could trigger on a trusted device because the risk signal system hadn't received the device recognition update; an affiliate with the correct role assignment could be denied access because the permission system was reading from a stale cache.
My support team spent significant time manually reconciling these inconsistencies: pulling data from multiple systems, cross-referencing states, and applying manual overrides. This was not a support problem. It was a systems architecture problem masquerading as a support problem.
The decision to build a unified identity decision layer was not just a technical choice -- it was a strategic one. By creating a single source of truth for user identity, the platform could eliminate an entire class of errors at the root rather than treating symptoms.
A centralized identity decision layer that aggregated and standardized all user signals -- authentication state, KYC status, roles, account status, and device/risk signals -- into a single, real-time propagating model that governed access behavior across the entire platform.
I defined a unified identity model with clear state definitions and interaction rules across all five signal types
I aligned Product, Engineering, and Operations on a single source of truth for user identity state
I enabled real-time propagation of identity state changes to eliminate cross-system inconsistencies
I reduced manual troubleshooting by making system behavior predictable and auditable across all touchpoints
I established a scalable identity foundation that supported MFA rollout, compliance requirements, and future platform growth
Reduced support volume by 25-35%. Eliminated the class of identity-related system conflicts that had been generating the majority of escalated tickets. Decreased operational overhead from manual troubleshooting. Built the foundation for 100% MFA adoption and platform-wide access control improvements.
Fragmented identity is the root cause of most platform trust failures. A unified decision layer doesn't just reduce tickets -- it makes the entire system more reliable, auditable, and scalable. When every system reads from the same identity model, the platform becomes predictable. And predictable platforms build trust.
Support volume would have continued growing as the platform scaled, with no root-cause resolution in sight.
MFA rollout would have been impossible without a reliable authentication state model -- the security mandate would have stalled.
Audit and compliance reviews would have exposed the fragmented identity layer as a systemic risk.
The key architectural insight was treating identity not as a collection of attributes, but as a decision system. Each signal -- authentication state, KYC status, role, account status, risk score -- is an input to a decision. When those inputs are fragmented, the decisions are inconsistent. Centralizing the inputs centralizes the decisions, and centralizing the decisions makes the system trustworthy.
"I had five systems each maintaining their own version of the user -- and they disagreed with each other constantly"
"The fix wasn't patching individual systems -- it was creating a layer that all systems read from"
"Support volume dropped 25-35% because I eliminated the root cause, not the symptoms"
What the data says
“If issues are solved during the first customer interaction, 67% of churn can be prevented.”
Identity conflicts that surface during onboarding or first-use are among the highest-churn triggers. Resolving them at the system level rather than through support escalation directly addresses this dynamic.
Source“74% of potential customers will switch to a competitor if the onboarding process is too complicated.”
Access errors caused by fragmented identity are a primary driver of perceived onboarding complexity. A unified identity layer removes the most common class of friction.
Source“88% of customers say the experience a company provides is as important as its products or services.”
Unpredictable access behavior -- the direct consequence of fragmented identity -- degrades the experience regardless of product quality. System reliability is a CX investment.
SourceWhite Paper Thread: The Decision Layer
This case study establishes the foundational argument of the white paper: that identity is not a data problem but a decision problem. The unified identity layer is the first instantiation of the core thesis -- that building a decision layer above fragmented data sources is the architectural pattern that enables scale, reliability, and trust across complex platforms.
Read the White Paper →Connective Tissue
The unified identity layer was the prerequisite for the MFA rollout. Without a single source of truth for authentication state and device recognition, achieving 100% adoption without friction would have been impossible.
Read case study
The identity decision system and the Xebra rationalization were parallel workstreams that reinforced each other. Consolidating identity signals required resolving the legacy system conflicts that were generating inconsistent state.
Read case study
Both cases address identity as a system design problem. At Tend, the challenge was cross-border compliance state; at iPROMOTEu, it was multi-system signal fragmentation. The architectural pattern -- state-driven, centralized, real-time -- is the same.
Read case study
See the prototype built from this experience
This work informed a real product
The Operating System
ibuildsystems.io
Four frameworks. One repeatable system. Applied across banking, fintech, government, and B2B SaaS to turn broken workflows into scalable revenue engines.