TRIP

A handful of genuinely useful specific tips emerge - AI prompting for succinct alt text, question-format hierarchy, and the counterintuitive research-on-research finding that disabled respondents prefer direct accommodation questions - but the episode is heavily padded with origin stories, cheerleading, and generic exhortations to 'start small.'

“prompts had to be reinvented so that we didn't just say put all text against this image, but put succinct alt text against this image, keep it brief, in some cases even keep it down to less than 10 words”

“questions that don't give cues about the response options. If you say something like, how does this product meet your needs? And it's a scale of 1 to 5 and not everything is labeled. That is harder”

The universal-design-improves-everyone argument is well-worn, and the episode largely applies known accessibility concepts to market research without first-principles thinking; the lone counterintuitive finding - that disability communities prefer blunt, direct accommodation questions - offers a genuine flash of novelty.

“little wins are big wins when it comes to accessibility”

“it's not monolithic. It's far from monolithic. I mean we're dealing with neurodivergence, we're dealing with so many other disabilities”

Both guests are legitimate practitioners and volunteer consortium co-leads with real fieldwork experience, but neither commands a notable track record at scale - Steve references a single healthcare spreadsheet project and Kari a case-study with a colleague - placing them solidly in the mid-tier practitioner range rather than senior operator territory.

“I'm now doing my own thing at Strat360 and a client came to me wanting to populate a spreadsheet”

“I've spoken at probably in the last year and a half, maybe six conferences”

A few concrete anchors appear - cost reduced by a third, alt text capped at under ten words, a named upcoming speaker on a specific date - but the episode lacks systematic data, named client organisations, or quantified research outcomes; the 'trillions of dollars' market claim is asserted without a source.

“I used a combination of human beings, myself, Excel programming and some AI to do something for them that actually brought the cost down a third of the price”

“keep it down to less than 10 words. And then I was getting better results”

The hosts are clearly knowledgeable and one follow-up - asking whether the toolkit itself had been pressure-tested against AI outputs - produces a substantive response, but the overall tone is collegial and unchallenging, with no pushback on vague claims and predictable closing questions about ten-year visions.

“Have you pressure tested against the, uh, toolkit at all? I was just curious. Have you tried to use the toolkit and kind of fed the toolkit into some of your questionnaires”

“What is that vision or wish list that you would have for the future of their research”