I used information architecture (IA) to realign a product org.

Pictured is a box model diagram of the company’s IA, which spanned seven independently structured domains.

Following their IPO, this company focused on driving subscriber revenue with innovative features.

Their previous startup mentality was not compatible with new large-scale enterprise aspirations.

I led senior stakeholders to redesign the IA and restructure the product org.

Team
UX architect (me)
Interaction designer
UX researcher
2 sprint teams

Constraints
All-remote, vendor participation, limited time and budget, no dedicated internal team members

Project type
Strategic design sprint

Platform
Figma prototype, report

My role
Principal UX architect

Timeline
2021


Symptoms of a bigger problem

An ask from our SVP of product to, “Fix the product’s navigation,” was one symptom of the bigger IA problem.

Customers couldn’t find product value.

Product teams didn’t know where to put new features.

Engineering leaders weren’t aligned on how to scale.

Photo by Jacob Bøtter


Stakeholder anxiety map

“Discoverability” had long been cited as a hinderance to adoption by many feature teams. Product leadership was comparing notes and were ready to invest in fixing navigation issues globally.


I led senior stakeholders to redesign an IA the company had long outgrown.

We followed the double-diamond methodology alternating between periods of divergent and convergent thinking.

1. Discover

Research the object model

I wanted to make our conversations more precise and informed, so I visualized the existing IA with box model diagrams. I also compared our model to competitor models.


Competitor models

Each includes one product with seven layers of hierarchy.



Our model

Included seven products—each had its own objects!



Misaligned object ownership

During stakeholder interviews, I learned that all object types (except Data) had multiple owners.

Product owners

Object types

Split ownership led to UI incoherence.

Similar objects grouped by type.


Problem framing

Leadership isn’t aligned with the object model.

Customers can’t find valuable features they need.

Teams can’t build scalable features without a stable IA.

2. Define

Stakeholder education

This diagram inspired by Nielsen Norman helped me explain the relationship between product content (e.g., objects), an informed IA, and a robust, future-proof product navigation.

Navigation types

I encouraged the group to consider different roles that navigation plays in user wayfinding.

Where am I?

Global

What’s nearby?

Local

What’s related?

Contextual


Solution hypothesis

If we organize features by type, customers will be able to find what they need.

Also, if we align business owners with the object model, then teams will have the structure they need.

Teams will know where they fit in both the organizational schema and in the product navigation.

This will allow us to build a more scalable platform.

A kind request

3. Design

Brainstorm

Internal resourcing was tight, so I interviewed and selected a vendor. We brainstormed solutions given the following conceptual direction.

Object hubs

Every feature is listed in the navigation—organized by object type.

A landing page for each object type displays alerts and calculates metrics.

Create from anywhere

A ubiquitous “+” icon lives in a globally accessible navigation panel.

More discoverable and contextual: add any object from anywhere.

Portfolio management

Adhere to project portfolio management (PPM) principles: opinionated nomenclature and hierarchy.

Customers will create their project files inside portfolios.


Result: Concepts intended to improve object understanding and discoverability in order to increase adoption.

However, the inclusion of all-inclusive feature lists was overwhelming.

We needed to simplify further and focus on helping customers navigate their own work (not our system objects).


Plan

Hypothesis update

If we focus navigation concepts on customer-built work structures and allow people to find more relevant features (in the context of their work), they will discover value more quickly, which will increase the likelihood of adoption.

Process update

While the vendor generated a final round of concepts in the new direction, I scheduled two back-to-back design sprints. My goal was to recruit sprint team members from all disciplines and product feature areas.

At the end of the sprints, I would recommend an IA strategy to leadership, which would inform product team organization for the next fiscal year.

4. Deliver

Design sprints!

I ran two design sprints back-to-back, working with the vendor and inviting participants from all teams and practices around the company.

Sprint 1

Do people want a rigid PPM framework (e.g., portfolios & projects)?

Q:

“Not really.”

A:

Insight 1

Three layers of hierarchy is fine, but “portfolio” means one thing to a PMO and something completely different to creative ops. Be deliberate about using project management nomenclature.


Sprint 2

Should we merge container types?

Q:

“Yes, but not like this.”

A:

Insight 2

The concept placed both types of objects at the same level within the core app, but people didn’t understand how they differed from each other.

This insight validated the hypothesis that the redundant containers in the object model were confusing.


IA recommendation

Given the sprint insights, I proposed the following strategies to product leadership.


Consolidate objects and align ownership with object types.

Outcomes

Redundant objects like DataFuse and Data connection were merged.

Product teams of the same object type were situated within the same reporting structure.


Orient around solution navigation with flexible container language.

Outcomes

A “solutions-forward experiences” team was formed immediately following the IA initiative.

Joining them were the Workspaces and Grid Spaces teams, which were ultimately consolidated.


Present relevant, action-oriented features in the context of work.

Outcomes

A “tools & controls” team was formed as a collaborative enabler of global action services.

Teams like Conversations, Content, and Reviews worked together to standardize actions up and down the object hierarchy.

Impact

Product development wins

Organizational

All product & engineering teams were reorganized around the solutions-forward product experience.

Business + product

Solution-oriented model with clear object hierarchy gained momentum and went full-speed-ahead in the business.

Personal

I launched a workshop design and facilitation practice and led over 60 workshops in 6 months to help drive coherent implementation.

Design lessons learned

  • No two IA projects are alike.

  • Designing a good IA for a web application is different from designing a good IA for a website.

  • A facilitator is a creator and maintainer of momentum.

Previous
Previous

Managing design through acquisitions

Next
Next

Mission-critical resource planning