Hi, my name is
Johanna Evans

For the past two years, I’ve been deliberately handing parts of my job over to AI — from writing feature specs and running early customer discovery to keeping work on-brand across the company.

At Leading Design London, I’ll share what I’ve learned from actually doing it: what I gave away, what worked, what didn’t, and what I found myself doing with the time I got back.

My biggest takeaway? As execution gets cheaper, our value shifts. Less time producing, more time deciding what’s worth making in the first place — and applying the taste, judgment, and context that machines still can’t.

Session details →

Ask a language model for the evidence behind a treatment decision and it will hand you citations. Some of them will not exist. In one published audit, nearly half the references a chatbot produced were fabricated, and newer models have narrowed that gap without closing it.

In this hands-on hour with Colombian dermatologists, I work through the part nobody trains clinicians to do: how to search with AI, how to validate what comes back, and how to tell when a source is real. We look at the tools built specifically for evidence retrieval, including the ones that actually work for Spanish-language and Latin American literature.

The goal is never to convince anyone to trust these systems. It is to give physicians a repeatable way to decide what deserves their trust, case by case, source by source.

See the post →

At Leading Design London, I led a conversation about a question we’d been wrestling with at Guru: how do you design AI systems people can actually trust?

We talked about what builds confidence, and what just looks like it does. I shared what we’d learned from launching Guru’s AI Chat and Research features, including how we thought about trust debt, citations and verification, and the tricky balance between giving people enough transparency to understand what AI is doing without overwhelming them.

The best part was comparing notes with other design leaders facing many of the same questions in their own products.

Session details →

In March 2025, I joined a group of design leaders at Leading Design to talk about what AI might mean for our craft, our teams, and the role of design itself.

I hosted a table of leaders wrestling with many of the same questions I was: What skills would still matter? What would we need to let go of? And how do you prepare a team for changes that are moving faster than anyone can comfortably predict?

What I remember most was how candid people were. There was excitement, but also plenty of uncertainty — and no attempt to pretend we had the answers.

Looking back, that’s what makes the conversation interesting. Not whether we predicted what came next, but the questions we were asking at that particular moment in AI.

See the post →

In September 2020, I was a guest speaker for Stanford GSB’s Startup Garage, a course that has helped launch companies including DoorDash and SoFi.

I introduced students to customer interviewing and empathy mapping — using what they heard from real customers to get beyond assumptions, understand their needs, and make better decisions about the products they were building.

It was a chance to bring practical, user-centered design methods into the earliest stages of building a company.

See the post →

In 2016, I gave a long-form talk at True University, True Ventures’ two-day gathering at UC Berkeley’s Haas School of Business, on something I had become deeply interested in at SurveyMonkey: how to use data without losing judgment.

I walked through how we built two new product lines at SurveyMonkey from the ground up, using customer research, behavioral data, usability testing, and constant experimentation to decide what to build, what to change, and what to leave alone. I also shared plenty of the tests that failed — and why those were often just as useful as the ones that worked.

The core idea was simple: intuition gets you started, but data helps you find your way. The trick is knowing what to measure, when to trust the numbers, and when experience still has to make the call.

Watch the talk →

In 2016, I moderated an AIGA SF panel at Zendesk with design leaders from Google, IDEO, and InVision about what innovation actually looks like inside product teams.

I surveyed the audience ahead of the event and used their responses to help shape the conversation — from where good ideas come from to how teams make room for experimentation while still getting the everyday work done.

It was a lively, opinionated discussion about creativity, constraints, and whether innovation is really something you can plan for.

See the post →

In 2014, I spoke at SurveyMonkey’s Girl Geek Dinner, an evening bringing together women in technology from across the Bay Area.

I joined a great group of women from SurveyMonkey to share what we’d learned building products, growing teams, and navigating careers in tech. The evening ranged from lightning talks on product and engineering to a conversation about how to build a more metrics-driven organization — and brought together a community of women at a time when there were still far too few of us in the room.

See the post →

In 2014, I spoke at PayPal’s Masters of Quality Conference in San Jose, alongside an impressive group of technology leaders including Jez Humble, Danese Cooper, and Douglas Crockford.

At the time, I was leading design work at SurveyMonkey. My talk explored how we were using data throughout the product development process, from identifying the right problems and shaping an MVP to learning from users and deciding what to build next.

It was also a chance to bring a product and design perspective to a conference largely focused on engineering, quality, and developer practices.

See the post →