Feature upvoting lacks context and requires qualitative validation
Lindsay Allard · She Thought Her Product Was Obvious - Then Watched 10 Strangers Struggle · Oct 12, 2021 · at 16:58
Lindsay Allard, co-founder and CEO of PlaybookUX, discusses with host Sean Weisbrot why product teams cannot rely solely on community upvoting boards like Canny to determine what features to build.
0:00 / 0:53exact quote · 53.8s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“One of the challenges I do feel like with the upvoting is that you sometimes don't understand the why. Someone might get some user feedback. Oh, you know, I would love this feature, but what, why, why do you want it? What about it do you like? What value is it going to give to you? So one thing I always recommend if you do have those upvoting systems, take that data, which is amazing data, and then see what are the top five things. Okay, now let's look at those things. Let's actually interview people and talk to them. What would you do with this feature? How would you use it? What value would you get from it? And make sure that you're really understanding the full totality of that feature. Cause sometimes it can be like, oh, someone just said, oh yeah, that would be nice. Right? There's a difference between, oh, that would be nice. And hey, I would promote this to every single person. I know this would be a game changing feature for your team. And then with the qualitative research, you can understand that difference between, oh yeah, that's something I just suggested. And oh no, no, no, this would change the game for our organization.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Lindsay Allard
Asking users if they like a product invites polite false positives
“You're not like, hey, here's a product. Do you like it? Because then when you ask that question, it leads people to say, yeah, okay. They don't want to hurt your feelings, right? They don't want to say, no, I hate it. So you want to understand, like, is this a…”
Founders should have detached third parties conduct user interviews
“I always recommend if, you know, maybe have someone who isn't involved, If it's your baby, you're like, oh my gosh, like, I really want them to like this. So having someone who's a little bit detached is always good”
Allard: Open-ended questions alone do not remove user interview bias
“You could ask an open-ended question that's like, you know, what do you like about this product? All right, that's leading, right? Because you're already leading them to say that they like something about the product. They might not like anything about it. So …”
Product discounts fail as incentives for B2B user research participants
“If you're a company that's like a B to B software, and you're like, oh, you get like one month free, and they're a person who has no control of the billing, they don't, it really is irrelevant to them, and that their company is going to get a month free on the…”
Users care about broken workflows, not minor UI styling inconsistencies
“The user does not most of the time really care, to be honest. Like we've put it in front of so many people and they're like, oh yeah, I didn't really notice that. But they do notice when they can't figure out, when they got frustrated.”
Repeated specific questions unlock deeper usability test feedback
“If you just ask, like spend a few minutes on this website, people are gonna be like, okay, it looks good. You know, Keep asking in different ways, and then you'll get people to kind of open up and say more and give like really interesting feedback that way.”