Ashley

Does this context affect search summarizes or just chat? What about searches made via the API or in Ask Dovetail via Slack?

I was just thinking this same thing when coding some interviews this morning! I can get much more value out of it now!

Hey Anahita! I've been using User Testing for web message testing + market research for a few years now, but their packaging model + how its changed over the past year have me looking elsewhere. It should technically do everything you need, but the user experience is not the best compared to other options + theres a lot lacking. For alts, I've been looking at and have had good trial experiences with Dscout as well as Outset.ai. The UX research team at my company has Maze. It seems promising and like it'd work well for integrating with Figma, however the way it functions doesn't make it good for testing stimuli that aren't a prototype. Displaying videos or image stimuli and trying to collect verbal responses to questions about them just doesn't work easily. It's not the best for marketing research, but seems great for user research. Wynter might be a good option for you if you want to pay as you go vs. doing a larger annual contract. Its been a bit since I've used them, so unsure about Figma, but I was happy with their output the last time we used them. Also unsure on video/audio recording, but still wanna throw this one out there as an option. Qualtrics has their own unmoderated tool they launched recently, but its still super basic so I wouldn't recommend it till its more developed.

Thanks Rod! Happy to chat more on formatting!

I wish this was possible too! It would make my reports so much nicer looking.

I like it so far! During reasoning, I've seen an "enriching data" label pop up. Does that mean its pulling context from tags/fields/project overview/etc to enrich the relevant datapoints + give them context? One thing I am a bit sad about is losing the ability to prompt it to format the output. I don't think I need full blown formatting for this to be useful, but I do think it needs to be able to add some white space, chunk the information it shares, or just use basic formatting to make it easier for our stakeholders to skim.

No worries, thanks for looking into it!

How many tags do you have in your project? In one of the training sessions at the Dovetail conference this year they described how users with larger tag boards tend to have worse results from suggested tags compared to those with a smaller # of defined tags. I also wonder if "training it" by highlighting a few good examples of what does fit under this tag first, then applying suggestions would work. It's unclear if existing highlights under a tag are used to help predict other potential highlights. 🤷🏻‍♀️ I know these two things aren't full solutions, I'd like to see improvements to this feature as well, but figured I'd throw them out there!

But i just found a different org installed blocker (ublock origin) that was messing with it! Seems to have resolved the issue to turn filtering all the way off for that. I guess they just changed their settings or something cause I've never had this issue. All resolved now though, thanks!