Thinking about purchasing information architecture services and wondering how to measure their effectiveness? Here’s my take on what makes a good information architecture.

During our first meeting, a free 30-minute discovery call, I’ll ask questions about your desired results and assess if Little Language Models’ services are the right choice for your business.

• What does success look like for this project?
• How might we measure success?
• Why do you want to do this project now?
• Are there any initiatives that may affect or depend on this work?
• Do we have leadership support for this initiative?

Not everything is measurable, but it’s generally easy to think of and track important metrics in the context of websites and apps.

AI Policy: I personally write each draft and final copy on this website. All content reflects my own thinking, ideas, style, and craft.  I do not use AI such as ChatGPT or other LLMs to generate articles. Occasionally, I ask AI (such as Formalizer or Equativ) to summarize or re-state my own ideas and may restructure sections based on the response.

Information architecture outcomes, metrics, and activities

Outcomes include more conversions, increased user retention, and decreased customer service costs.
Metrics include task difficulty, time on task, site visits, search terms, return users, GitHub issues, rage clicks, and Customer Satisfaction Scores.
Activities include audits, workshops, interviews, tests, and content modeling, supported by tools like Figma, Hotjar, Google Analytics, Mixpanel, readability checks, accessibility checks, AI keyword clustering, heatmaps, and site crawlers.

Desired client outcomes

1. A client wanted their site to bring in more leads. They had received user feedback that the site was overdesigned and frustrating.
2. A client aimed to expand their pipeline and improve success rates from the first time AI engineers landed on the platform to importing sample models, verifying inference, and tuning a model.
3. Another client wanted to reduce customer support requests and the volume of unqualified scholarship applicants.
4. A client was migrating 30 local sites to a new architecture and wanted to ensure they were considering content implications.
5. A client wanted standardization across the university’s departments, but also templatized content that would make publishing easier and faster, and provide some space to express their uniqueness.
6. An open-source software team wanted to improve an experience they kept getting complaints about on GitHub.
7. I’ve been brought in by clients to design a minimum viable product (MVP) that would help them define the Request for Proposal (RFP) scope and find the right technical partner for platform implementation.
8. I’ve been brought in to secure leadership buy-in.

Every goal is different; on a high level, the first and the third client had opposing goals. One wanted more leads, the other wanted fewer.

A clear definition of success sets us up for success and ensures we’re not spending time or energy on initiatives that won’t move the needle.

I’ll provide a walkthrough of key user flows, what’s going well, and what could be improved. You can apply these guidelines to future content initiatives. You’ll also see the before and after for the new experience strategy. I’ll showcase content elements in a hierarchy that matches key user needs and business goals, and explain how I’ve assessed content effectiveness (and how we can test it).

What can’t be measured

Earlier today, I went to MA 35, the City of Vienna’s Immigration and Citizenship office, to pick up a document. I arrived early, which always gives me irrational joy. They took my passport and told me to wait a bit.

I sat in the waiting room and noticed that everyone around me had a queue number in hand. People who arrived after me left before I did. Still, I believed I was in the right place, even though the exchange with administrative staff took place in German, which isn’t exactly my strong suit.

Another employee came to the room and called my name, handed me an invoice, showed me the queue, told me to pay, and then go to her office, which she pointed to. This conversation also took place in German.

Two clerks. Two queues. They seemed to serve the same purpose, so I assumed I could join either… until I saw the signs at each window.

Photos weren’t allowed, so enjoy this drawing. Only one of the clerks could process payments, the one on the left with the sign that said Kassa (Pay).

I’m sure the clerk on the right offered many necessary services, but whoever put the signs up wasn’t focusing on them. Instead of listing the right booth’s services, they told people upfront that they couldn’t pay there. They used only two short, simple words recognizable even to non-German speakers.

Waiting isn’t generally enjoyable, and even less so when you should stay alert to hear your number or name, avoiding reading, listening to music, or talking to prevent missing your turn. Public service offices often feel unwelcoming, especially to immigrants. Waiting in the wrong line in a government office can ruin your mood.

Thankfully, the window signs were very clear and helped me avoid waiting in the wrong line. After the clerk processed my payment, I told them about this sliver of gratitude found in an unexpected place. She gave me a big smile with teeth. She restated what I had told her with joy and seemed to appreciate that I shared this feedback with her.

These interactions are difficult to quantify, but that’s what makes them beautifully, irreplacably special.

Unicorns and bucket lists

In her book Conversational Design, Erika Hall shares a food order confirmation email and how even unicorns can get old fast if you don’t focus on the humans.

Rather than thinking of interfaces as abstract events, consider the actual people who will use them, the customer interactions that happen over and over and over in real life.

The smiles or sighs.

Kudos in Slack from customer service reps or a weird tension when you ran across them at the office.

Smooth processes or looong forms with strange questions resulting in stories like the one below.

Cancer patient Mandie Stevenson had to postpone a bucket list trip to New York after she accidentally labelled herself a terrorist on her visa waiver form.

---

Take time to celebrate intangible wins like positive client remarks, a sense of relief, or how pleasant interactions with colleagues from a particular team now feel.

Not everything is measurable, but focusing on desired outcomes ensures I can plan the right information architecture activities, track the right metrics, and use the right tools to help you achieve them.