Why Amplitude Acquired Kraftful: CEO Yana Welinder on Unlocking Product Insights with AI

Hello and welcome to the latest edition of the digital CXO Leadership Insight series. I'm your host, Mike b. Today we're with Yano.

Well, Linda, who is CEO for Kraft Ful, and they were just acquired by a company called Amplitude. And we're talking about the two of them together or analyzing data. And on one side.

And then the other side of it is, um, Janna's company is collecting the data and getting the feedback in the first place and understanding what people are trying to, uh, say. It's not just a matter of data, it's also understanding their intent. Yana, welcome to the show.

Thanks so much for having me, Mike. Really excited to be here and, and proud of that. Yeah, I'm not sure most people know who both companies are, so walk us through what's the motivation for this merger and acquisition and how it all came together?

Yeah, absolutely. Um, so for on, on our end, obviously, we, uh, on, on the craft full side, we help over 60,000 product team listen to users at scale by collecting user feedback from all these different sources of support tickets, call transcripts, survey data, app reviews, and make sense of it all. Um, amplitude is this incredible, um, broader platform that really helps teams build better products by providing product analytics, but also things like, um, session replace surveys, um, AB testing, really a whole, um, tool set of different, uh, ways to help product teams be, be more, uh, successful building products.

And so the missing piece to, to the puzzle on their end was really to be able to understand what users were saying and find that needle in the haystack, and that's what Raffel does. Um, so we ended up chatting with a team. We talked about a few different ways to partner together, but ultimately as, as the, uh, as the Amplitude CPO told me was we should just be one team.

And so, and so we started talk, talking about, uh, joining forces and becoming one team and, and, uh, and it's been an incredible conversation so far and ultimately we ended up joining them. What is challenging about understanding what end users are trying to say? You would think that they just fill out a form and everybody could figure it out, but I'm sure there are nuances in there.

So what exactly are we surfacing? Yeah, so the big challenge is really the volume of feedback that, that every company gets. And my, my background is I've, I've led product at various companies, everything from being PM number two at a, at a flash screen unicorn to leading product team with, uh, where, you know, our products were used by millions of users.

And the challenge is really that you're getting a ton of support tickets every day. Your, you, your sales team has a bunch of calls with, with customers. Um, if you're a mobile app or if you're some sort of like online product, you also have online reviews or mobile reviews and you have survey data.

And so what what ends up happening is that a lot of this feedback either gets ignored or you have teams spending lots and lot, you know, hundreds of hours trying to analyze all this data and make sense of it all and try to incorporate it into the product team's development process. So what we ended up doing is really saving teams hundreds of hours, um, and instantly telling you, here's your list of feature requests that are coming from all of these different sources. Here's what, here's the feedback you got today, and here's the actionable list of things that people are asking for.

And then you click through and see what exactly people said and and, and turn that into, um, tickets that your engineers can start working on immediately. But it's, it's sort of like making sense of it all. We are of course living in the age of AI and we're all obsessed with data, um, as the two companies come together or there are ways to think about applying AI to the analysis that will help people maybe surface some more interesting insights.

Absolutely. Well, FU is an AI native platform, and I should say a large language model, native platform. So we have our proprietary LLM process that we apply to all this data to, to turn it into actionable product insights.

So that's kind of our better butter is to really use AI in, in all of product development. Um, and Amplitude on their end have already started the journey of, uh, really building out different AI capabilities in, in different, uh, products. And so now we're really bringing all of that together, um, and making sure that we're building the platform in a way that's thoughtfully applying AI instead of just bolting it all on onto an existing product, which I think a lot of teams do.

Mm-hmm. So what is the ultimate impact on product managers and the people who are trying to decide what the next product to build is? Um, will they take all this data and make better decisions?

'cause a, a lot of times I feel like they're always wrestling between, um, instinct and data. That's right. Yeah.

So absolutely. I think, I think the key is really understanding everything that users are doing and saying, and that's, that's what the joint amplitude craftable experience is gonna provide. And then it is up to product teams, um, to then decide what to do with that.

Um, and a lot at that point, a lot of that is product intuition and being able to make thoughtful decisions. But if you don't have the underlying data to even make those decisions, then you are acting on instinct. And that instinct is often, uh, misguided, right?

It's not, it's not based on anything that users are, are actually interested in. Um, so we are really making sure that you have the relevant data points to then use your intuition in, in the, in, in the best possible way. And can I experiment more readily because maybe I can segment different, uh, types of people that might be using my product for different use cases, and as such, I might discover entirely new use cases for something Absolutely right.

You can, you can, you can segment better, you can survey users better, you can in moments understand if something is appealing to your users, um, and sort of, uh, act much faster on, on different hypotheses that you are also forming based off of all of this data that you're getting. So there's so many different ways in which you can, you can faster iterate in the process. Are the teams themselves gonna become more integrated?

And I ask this question because a lot of times when I see folks who are doing data analytics, they're often their own little group somewhere and they don't always interact so well with the folks in manufacturing and sales, et cetera. So are we gonna get, uh, better able to kind of take all this data and surface something that feels like more actionable intelligence across those teams? Absolutely.

I think, you know, one, one thing that we've thought a lot about is bringing in insights from, um, data sources that product teams didn't have to or haven't traditionally taken into account. And by doing that, by listening to customers in, from sales calls and getting insights from those conversations and getting insights from conversations from, uh, support ticket channels and all of these different channels that product teams traditionally haven't necessarily listened to, uh, product teams are building empathy not only with the end users in those specific situations, but also with their colleagues that are having to respond to, to those requests or having to pitch the product in a sales meeting, right? So I think that does make the whole team a little bit more interconnected, um, in ways that, um, that hasn't been possible before just because the volumes have been so large that no one's really tried to take that on.

So who in these organizations kinda wakes up one morning and has the aha moment that says we need to rethink analytics and, and drives this kinda investment? Yeah, so it's primarily gonna be product managers, um, in, in, um, in folks within product teams, product and engineering team. So it goes a little bit broader than that, but I would say that the kind of the primary persona is usually the CPO, uh, and, um, and the product team.

And that's true, that's true for Amplitude and that's also been true for crackles. We had that joint persona, uh, that we've been serving, but with different solutions. And so now we're gonna be serving, serving them jointly in a, in a much more effective way.

Mm-hmm. In the age of ai, how smart will smart get? And today, I feel like a lot of the issues with the analytics is I need to know what question to ask in the first place, but I wonder as I go along here with things like AI agents might not the AI agent know what question to ask before I even think about it.

Yeah, so, you know, one big piece of this is that you shouldn't have to think about what question to ask. Your user should be telling you what's, uh, what they care about. And they already, the good news is that there already are, right?

Um, and so this, what this does is just make sure that you can effectively listen to those, uh, to all of the feedback you're getting from users. Um, now we, uh, at Raffl actually have developed a survey method that takes your past feedback into account and builds follow-up questions for users. Uh, we can do that either, uh, before you set up a survey or proactively in real time while a user is providing feedback.

So it takes the user's prior response and formulates the follow-up question, um, that's gonna help you dig deeper from, from a product insights perspective. So definitely lots of opportunities to, um, to leverage AI to make sure you, you get to ask the right questions, but it's also just a matter of listening to everything users are already proactively sharing because there's a lot, there's a lot of insight there. And is all this happening at a higher level of scale?

Because in my mind, at least historically, product managers would, you know, they'd get a panel together and they'd stick 10 people in a room and hope for some sort of insight to come out of that. But whether that was applicable to the entire customer base was anybody's guess. So are we just gonna get smarter about all this?

Absolutely. This, this is a way to make sure we can listen to users at scale. And when I say scale, to give you an example, we had a customer at Craft Full that surveyed users and they got, uh, they got 16,000.

They were able to survey 16,000 users in, uh, in, in just a few days. Um, and, and incorporate all of the, those insights instantly because they instantly got a list of feature requests and complaints and all the different, uh, topics that users were saying. So that volume is something that hasn't been possible before.

As you say, you sort of, you had to listen to a small group of people that wasn't necessarily representative of the user base and had gave you sort of a skewed view on what, uh, users needed from a product. So what do you see people doing today that just makes you shake your head a little bit and go, folks, we need to just be a tad smarter than that. Um, you know, know, I think a lot of the, I I still do see folks do that, you know, focus groups and, um, kind of small scale feedback where you, you could gather just so much more data.

Um, and then another thing that I see that, that, that personally irks me is when folks listen to users have a way to collect feedback from users and then dismiss it as sort of like, well, this isn't helpful. This is just a bunch of people telling me that they need a faster horse. I need to think bigger.

You know, um, and, and, and as they're thinking bigger, they're not thinking about that car that they could be building for the users that are asking for a faster horse, they're just thinking about something completely different, right? Um, that, that isn't at all helpful to what folks are asking for. So I think that, uh, the, the kind of the, the arrogance in, in, in product teams is the thing that sometimes, uh, it really gets on my nerves or I'm sort of like, well, you have lots of data.

Why don't you kind of creatively think about what is, what, what is it that, what's the thing that would really help these people be successful? Help your users and customers be successful based on what they've told you. Yeah.

Hey folks, we're all trying to make better data-driven decisions, but um, frankly, you just need to be able to sort through that data at scale. And well, AI will hopefully sort all that out for us a little bit, but um, maybe we'll just make happier, better customers. Who knows.

Hey, Yana, thanks for being on the show. Thanks so much for having me. Really enjoyed it.

All right, and thank you all for watching the latest episode of the digital CXO Leadership Inside series. You can find this videos and others on our website. We invite you to check them all out.

Until then, we'll see you next time.