Ibotta Performance Network (ipn)

Scaling Enterprise Information Architecture

The task

The platform was originally designed for expert promotion marketers, but as usage expanded to broader digital and social marketing audiences, the existing information architecture began to show strain. Discoverability issues, inconsistent terminology, and unclear task pathways increased friction for self-service users while slowing internal decision-making around structure and naming.

The challenge was evolving the IA to scale beyond expert users without fragmenting the experience—balancing legacy concepts, mixed user mental models, and organizational complexity in a platform that needed to support both velocity and long-term growth.

Problem Statement

How do we evolve an information architecture built for experts into a scalable, intuitive system that supports diverse user needs—without increasing cognitive load, introducing structural debt, or stalling cross-functional alignment?

information architecture audit

The Before

Annotated navigation screens and an external sitemap surfaced systemic friction between user intent and platform structure. Overloaded navigation items mixed actions with destinations, terminology like “Programs” lacked clarity, global and local navigation patterns were duplicated, and account context (e.g., organization state, alerts, profile access) was ambiguous—revealing accumulated structural debt and reinforcing the need for a validated, lifecycle-driven IA before scaling further.

The Goal

The goal was to establish a validated, information architecture that aligned with real user mental models, improved task discoverability, and enabled confident self-service as the platform scaled.

By grounding structural decisions in research and shared evidence, we aimed to reduce internal debate, align UX, Product, BizOps, and Engineering, and create a durable IA foundation that could evolve with future features rather than constrain them.

Approach: IA Testing Overview

Phase 1.

Internal card sort pilot (qual-moderated | *AMs, ACs, CPs, COAs)
Approach: Moderated sessions with 5-7 *internal participants

Phase 1 focused on quickly identifying obvious labeling issues and internal bias before involving external users. Internal stakeholders had deep domain knowledge, making them well-suited to pressure-test terminology, surface ambiguous labels, and highlight areas where internal mental models might diverge from how the product was currently structured.

What this de-risked

• Shipping confusing or overloaded labels into external testing

• Anchoring later research on flawed terminology

• Wasting external research cycles on known internal issues

Phase 2.

External card sort (quant-unmoderated | *Marketers)
Approach: 40 external participants (*digital, social, and promotion marketers)

Phase 2 validated whether proposed groupings aligned with how real users conceptualize the product—at scale. An unmoderated quantitative card sort allowed us to detect strong patterns across multiple user segments and compare internal assumptions against external mental models.

What this de-risked

• Designing navigation around internal structure instead of user intent

• Overfitting the IA to a narrow audience

• Making structural decisions without statistical signal

External, New IA structure prototype testing
(qual-moderated | *SBGs, ATs, BBTs)
Approach: 8-10 external participants who use the IPN (*Small Brand Generalists, Agency Tacticians, Big Brand Tacticians)

Phase 3 tested whether the validated structure actually worked in practice. Task-based first-click and usability testing allowed us to observe real decision-making, identify hesitation points, and validate that users could complete high-value workflows within the proposed IA.

What this de-risked

• Implementing a theoretically sound but practically unusable structure

• Discovering critical usability issues post-launch

• Misinterpreting card sort success as task success

Phase 2 Testing

Key Quantative Insights

Through phased IA testing, we found that users—regardless of role—mentally organize the platform around the campaign lifecycle rather than internal product structures. Pre-launch, mid-flight optimization, and post-campaign reporting were consistently grouped together, while billing and account management were expected to live side-by-side.

Phase 3 Prototype Testing

Key Qualative Insights

Across first-click tasks, participants were consistently successful when navigation labels reflected what they wanted to do (e.g., create, manage, review) rather than internal structural concepts. This validated the shift toward action-oriented groupings within the proposed IA.

Users reliably located and completed the organization switch task when it was positioned alongside profile and account settings. This reinforced earlier findings that org context is perceived as an extension of the user’s identity—not a global system control.

Users consistently distinguished between Drafts, Active, and Completed campaigns when Drafts were given a dedicated location. This reduced uncertainty around campaign state and eliminated incorrect assumptions about whether work was live or in progress.

Rather than uncovering major structural flaws, Phase 3 testing validated the overall IA direction and helped refine edge cases. This shifted conversations from whether to proceed to how and in what order to implement.

Design Decisions

The key insights directly informed a restructured IA that prioritized task clarity over legacy hierarchy. Campaign actions were consolidated, a dedicated Drafts area was introduced to reduce ambiguity, and billing was aligned with account settings to better match user expectations.

Not all findings led to immediate decisions. Homepage expectations and the role of “Programs” surfaced as areas requiring deeper alignment, so we intentionally deferred those changes and planned follow-up workshops with BizOps and Product.

1. IA restructuring, renaming, and restyling

   1. Restyling update for the Primary CTA and Left Navigation

   2. ‘Overview’ renamed to ‘Home’

   3. ‘Programs’ renamed to ‘Campaigns’ (In-Progress)

   4. ‘Expired’ status renamed to ‘Completed’ status

   5. Move help from global nav to the left nav (bottom)

   6. Updating the support chat treatment


2. Profile Org Picker

   1. Profile and Org are combined within the top right of the global nav header.


3. Dedicated Drafts section

   1. Adding drafts to help distinguish between an incomplete campaigns vs. a pending, active, cancelled, or completed campaigns.

Outcomes

The study validated a lifecycle-based information architecture that aligned with user intent across campaign creation, management, and reporting. Structural changes—such as separating Drafts and clarifying navigation labels—reduced ambiguity and improved confidence across priority workflows.

Impact
By grounding IA decisions in research, the work shifted cross-functional conversations from opinion-based debate to evidence-led prioritization. The validated structure unblocked engineering planning, informed naming decisions, and created a scalable foundation for future feature development without compounding navigation debt.

Reflection
Treating IA as a living system—and intentionally deferring unresolved areas—proved more effective than forcing premature alignment, increasing confidence in both the solution and the process.