Typeform's CMO on Why AI Still Needs Human Judgment

Malinda Sandman 0:00
You're getting one piece of feedback, right? You're not getting feedback at scale. And so the squeaky wheel bias, the challenge is you may be let listening to the loudest voice or like customer reviews, right? You're listening typically customers write a review because they either absolutely love something or absolutely hate something, right? And you're some, you know, not necessarily getting what does your market at scale, what does your customer at scale really think about that. Um, and that does where research flow is really solving that for marketers is being able to give this context at scale. So you get the, you still get the quantitative and your ability to be really confident in the um, I think the scale and where this direct, where the feedback is going, but then you get all that rich context to say, okay, I see what's happening in the data, but I don't understand why.

Kevin Kerner 0:45
Hey everyone, this is Tech Marketing Rewired, and I'm your host, Kevin Kerner. My guest today is Malinda Sandman, VP and head of global marketing at Typeform. Typeform has been one of the most recognizable names in form for a decade, trusted by over 150,000 customers and 95% of the Fortune 500. Malinda's leading the push into AI now with two new products out in the last two months and getting the market to see Typeform as more than the form builder that they already know. This one's part of the Pass the Pilot series. Um, it's a series where we ask senior marketers what they're doing now that AI experiments are over, and Malinda answered that quite well. So this is a fun one. Let's get into it. This is Tech Marketing Rewired. Malinda, great to have you here. Thanks for joining.

Malinda Sandman 1:38
Great to be here. Thanks for having me, Kevin.

Kevin Kerner 1:41
Yeah, I was very excited to talk to you, as you know, from the couple conversations that we've had. I am a big typeform fan. I know you guys have been around forever. And you know, it's just really interesting nowadays, like like voice and data is context. And it's it's the more um information I can get, even transcript data or forms and those type of things, it just becomes so useful. I don't know in your daily workflow, but in mine, I'm always like hitting the download MD file button on everything these days.

Malinda Sandman 2:13
AI has has completely changed the game. That that's um on the ability, I think, to be able to get that that rich context on everything. So same. It it's the how do I get as much information as possible and then figure out how to get the those quick insights.

Kevin Kerner 2:29
Yeah, and making it easy. So you guys have done some of that. So I've followed you guys since you launched probably in the tens of 2012, I think or so. 150,000 customers, maybe more at this point, 95% of the Fortune 500. It's such a cool heritage brand that was began as a Google form, better Google Forms, and now it's it's sort of building into new things. I'd love your take on the market environment you're playing in, how things have changed um in some of the new stuff that you're you guys are announcing just now.

Malinda Sandman 3:01
Yeah, absolutely. Um, I mean, the market has changed so dramatically, I think even in the past, um, even in the past month, right? It it changed.

Kevin Kerner 3:10
Yeah, right. Yeah.

Malinda Sandman 3:12
So uh but I think I think that our vision, um, you know, and I think our heritage uh is still so relevant to to marketers and to businesses, you know, ultimately it's around um the heart of how do you get closer to your customers? Um, you know, so one of our beliefs is when asking feels human, people share better answers. And I think kind of that belief and that philosophy is so um so much as the heart of this kind of human first design, the conversational forms that typeform innovated. Um, but it goes back to the end of the day, the job to be done and the use case and how we can help marketers, businesses, right? A whole host of researchers, product managers, right? A whole host of different um roles in companies be able to get closer to their customers. Like that ethos is still there. And now, with what has happened with what AI can do, um, it just allows us to now grow in our um applicability and use cases to these customers. Um and in kind of in two ways in particular, like so um I've been at Typeform since August. So I'm almost at a year, and it's been great to kind of really dig in there and see, you know, what are people really using Typeform for? And kind of the two things that keep coming up. Um, customer feedback is one of our biggest use cases. Um and so that's been kind of as we've been evolving, is really thinking around now what's available, because obviously in 2012 um and even a few years ago, right, that this wasn't available. What can we now do to be able to supercharge customer feedback for our customers? Um, the other one would be kind of lead gen, right? And how do you kind of grow your customer base? Um, and that's also been a big application to really just be more to our customers.

Kevin Kerner 4:55
Um on the research flow side, um, from a product perspective, I mean, that market seems to be a something that's growing. And there are quite a few AI moderated sort of qualitative research companies that are out. Um so now you have, you know, you're in this research flow uh space, qualitative research space, and you're competing with these AI native companies that are coming out, even though you have the you have the install base, which is great. I'm really curious, like as a CMO, how you're thinking about the levers that you pull to I guess compete with that market dynamic, especially with all these AI native companies coming into this stuff.

Malinda Sandman 5:37
So I think that is a competitive advantage and how we think about kind of how do we position ourselves against these AI native companies. But yeah, it is it is definitely a it is a great opportunity and a great challenge to think about how do we approach this in a different way, right? We're not we're um founded in Barcelona. We have a globally distributed workforce. Um so we're not, you know, a Silicon Valley in native tech company. But I think that gives us um a real advantage to come one, to come to the market. And like you said, we have this incredible customer base. Um, we've been around for a while, we have that brand recognition. So it's that blessing and that curse to be able to now evolve and change perception. Um, but still really granted in what are people using typeform for? Right. And so it's that it's that next step of these incredible forums and surveys, right, and quizzes that get you access to this data. But now we can take it a step further and be able to really democratize access to that context.

Kevin Kerner 6:32
It's such a missed thing. So many people miss on the product side and the product marketing side, the research angle. Like they go right into the product. It's like, oh, we we have a great idea. So many people don't go like the research angle. It seems like you guys would, because you your uh you know, forms generally are used for research. And so it kind of leads you in that direction. But what's it are there any interesting things that you found in the research that drove you in a direction from a marketing perspective, either messaging perspective or channel perspective, um, ICP, like anything that you found that was surprising or maybe even solidified the path you were going down?

Malinda Sandman 7:09
Honestly, we are in this iteration phase right now. So uh we launched it in beta in May as part of our spring release. Um, and so right now we so we have our hypotheses. Um and you know, that's actually the some of the fun part is like we are using the product on our own. We were actually, I I just met with my team and we were talking around creating a messaging house for Research Flow. And I asked, can we, I'm assuming we can use Research Flow to be able to go out and see right what's really reconating. Um and that really, and and so we we kind of laughed. We're like, absolutely. Um, and um I I think that's where, you know, when you talk about kind of the unlock for marketers, um, I've always been a lot of my background is more on the the web and more on kind of the product conversion funnel. Um so I've I've sat really close with a lot of product teams and a lot of uh UX designers and researchers, and this ability now to be able to unlock this um combination of qualitative and quantitative. So the work that I've done in my past is how do you use kind of the qualitative? How do you get really close to what your customers, you you've got maybe five different hypotheses, narrow it down, right? And then you can get into A-B experimentation. Um, that's where I I think that kind of thought process and that methodology is what I'm kind of bringing to the marketing team and thinking around, all right, how do we now, we can even use our own product now to be able to get more of that contextual insights and then be able to go and test it from a quantitative perspective as well. Um, but yeah, we're in terms of kind of how we're thinking of it, um, you know, it feels very meta in that we're using our own product to be able to how we want to go to market. Um, but that's really where we're. We are we have a hypothesis of um kind of really two core ICPs. Um, and now what we're trying to do is trying to just get as much kind of signal and confirmation of, you know, does one side is one ICP more valuable than the other? Where is this market in general? And then also kind of looking competitively at where the other kind of AMO research tools are going and and who else is also in our base um that's you know historically using type form.

Kevin Kerner 9:15
Yeah, it's really uh I think we on the pre-call, I was talking about how we use voice data quite a bit now, transcript data. Of course, we're using we're using maybe a Google Doc that has questions in it, but we would rather get the voice data than the typed data. And I know you guys have video ask and you have all these other ways you can the AI can ask you the next question. But when you're in code, it's better to download that very vast transcript and MD file. Like if we put our transcript here into something, it's gonna be better than me sending you a form that has you type out. And I just cure I'm curious whether that I wonder how many people are know that yet. Do they really like our market do marketers know that? Is it better to get boy, you know, the rich transcript? Coders know it, you know, developers know it for sure.

Malinda Sandman 10:03
Great question and call out. I that is um what we are absolutely seeing in some of kind of our early data and what I think we know of, like you said, with with video ask and with um in our growth flow product having the ability to do the the um the video assisted forums, we see that in the effectiveness for our customers. So some of you know the early data that we're seeing is by having the video aspect of research flow, right? And having this AI moderator ask the questions in um in in voice and video versus in text. We're seeing about four and a half times more feedback per question in in video versus text.

Kevin Kerner 10:40
Yeah.

Malinda Sandman 10:40
Also that kind of nuance of then what do you get from kind of the emotional insight and how do you get more of that nuanced context? Um, so we're seeing about three times the amount of insight there. And then probably more than anything else, the engagement is twice as long as a service. Um, so right, we have the ability to kind of pull someone in, and I think that's that is the the form could get you there, but then this next level with having the AI um moderator is the AI moderator can change, right? And can the next nuanced question as the interview keeps going to kind of keep that engagement out and really dig to the surface or dig deep into what truly is going on here to get, you know, the market or the product manager, the research, right, the right amount of context that they need.

Kevin Kerner 11:23
Yeah, love it. Yeah, it's like rambling. I think I mentioned on the pre-call too the idea of rambling. Like rambling is a thing that we're told not to do, and I do it a lot. But when you're good, just when we tell our clients when they're when they're um doing uh transcript for us, we're not gonna watch the video. We're just gonna listen to the words. So so be as ramble as much as you want, as awkward as you want, because you'll this video will never get seen, and it's really good context for us. So rich. It's really rich. Um, I wonder, um, so you needed to message these new products. They're just coming out. So I did I I guess I had forgotten that um research flow is just in um May. That's uh how did you just let's let me take me through the thought process of how you message them? How did because there's so much AI messaging out there? How did you how did you come up with messaging for the site? You knew how many site are you added to the site, uh Webflow site. I'd just be curious about your process to get messaging, at least the first level of messaging aligned.

Malinda Sandman 12:24
Yeah, yeah, the first level, um, so that that definitely was going out and trying to test with some audiences. Um, but really I think that kind of builds on our heritage, right? And our ability to, we know what the most valuable use cases were for typeform using our um heritage products. So if you're using form builders, how are our existing customers using form builders surveys, quizzes, right, to be able to get that data. So I think we were able to see um, you know, through different case studies that we had, through different customer um testimonials, how were they using typeform and what were they really trying to accomplish? And then basically building on top of that, right? What was um and we we had a fair amount of messaging that we could leverage from um we launched growth flow in a couple different iterations over the last couple of years. Um, and we also had launched Typeform AI, um, which kind of is that underpinning across everything. Um and so we launched that kind of in earnest um in GA last fall. So that was like that was our first big product release um since joining form. And so I think as we were getting um kind of feedback and seeing what was resonating, seeing what was resonating in some of our paid media, seeing what was resonating across the website and seeing kind of where we were getting whether it was interaction or maybe we were getting better signups and conversion of how we message type form AI, we almost like fused those together in a lot of ways, thinking of that. Um, but I agree with you where we've debated this a lot is how do you um how do you make sure that you don't go too AI heavy? And I think it's been really important for the entire marketing creative team to really continue to sit with our brand differentiation and and what um what customers right and the market kind of expects from typeform is that is that human-first interaction. Um we actually even we had released um a study um about how marketers feel about AI um earlier in the year. And I think even taking some of that feedback, right, and continuing to have that context when we think about reaching that marketer ICP of how they're feeling about AI, then really pressure testing our messaging and positioning against the emotional uh, you know, perception and how AI is helping them in their um in their job daily, but then what are they also afraid of? And then thinking around how to kind of bridge that connection with research flow. Cause ultimately is about, well, you know, how do you give marketers this context? Um, you know, one thing um that we talked a little bit is like this the squeaky wheel bias that I think all marketers can can fall into. Um, you know, I think we're we're kind of trained to be to be as close to our customer as possible, right? And so you want to be able to talk to customers. Um, I know it's always been a big thing that I've I've seen in in multiple roles that I've been in, you know, the the leadership is like I sit down with a customer weekly and I talk to the customer. You're getting one piece of feedback, right? You're not getting feedback at scale. And so the squeaky wheel advice, the challenge is you may be listening to the loudest voice or like customer reviews, right? You're listening typically customers write a review because they either absolutely love something or absolutely hate something, right? And you're some, you know, not necessarily getting what does your market at scale, what does your customer at scale really think about that. Um, and that does where research flow is really solving that for marketers is being able to give this context at scale. So you get the, you still get the quantitative and your ability to be really confident in the um, I think the scale and where this direct, where the feedback is going. But then you get all that rich context to say, okay, I see what's happening in the data, but I don't understand why. And I think that's the the game changer with marketers being able to use Research Flow is to really understand the why and to make better decisions there. Um, so we're we're also trying to think about that, right? As we're kind of looking at all of the data we have across, whether it's you know, releasing typeform AI and growth flow, how do we think about how you know how our customers are using typeform and kind of I mean we can use AI right to and research flow to kind of mass all the other and sit there and say, how do we actually now go to market um in a way that's gonna resonate?

Kevin Kerner 16:23
Yeah, it's really yeah, it's really important. And you're right, it's a very human, you have a very human brand from the beginning, you know, one question at a time and how the the the the it's not about just AIing everything, it's that enablement piece. Um it's it's interesting the surface where you can engage with the product. I don't know. I didn't look deep enough into it, but is there an MCP or any access that I can get to typeform data through Claude or any other LLM?

Malinda Sandman 16:49
Yes. So we are actively working on that. Um that's a big strategy for us. So stay tuned.

Kevin Kerner 16:54
Yeah.

Malinda Sandman 16:55
Um you know, in general, that's obviously a big strategy for us for discoverability, right? And and you know, customer awareness and acquisition. Um, so yes, that's that's something where um I think we have uh our product team is very um attuned to that. And so we've been working closely between product and marketing to say, okay, what is the best use cases and how do we bring this to market as quickly as possible?

Kevin Kerner 17:16
Yeah, yeah, I think that's that's common. I mean, there's so many brands that I learn from through in a coding session, you know, actually in it, I'm asking, look, I need an email platform, or I need a database platform, or I need a form generator or whatever, or I need a research assistant. And it will, there's some brands like resend is a great example for email. They've figured out a way, it's super base for database. They've figured out a way to get the coding agent to mention you. And I think there's something really powerful about the back-end tables you have of survey data and research data and being able to query that in a surface that's not the application itself, but in Claude or something like that. It seems like where the future's going.

Malinda Sandman 17:57
Absolutely.

Kevin Kerner 17:58
Um I was gonna ask about the site. The site's great, it's beautiful. Um, the new products on there. You had you do have a um little chat box. I think it's TY is called, ask TY. Um and I was wondering how that's working for you. What's what's fueling it? Is that tied into sales? Does it work to have one of those on the site, especially as early as you are on some of these products?

Malinda Sandman 18:21
Um, that is also a relatively recent development. Um and I would say I I have history in this, and I I think that's a good uh tie into thinking around also the um what have been kind of all the different kind of tools available to marketers and and how I think my um AI and kind of research flow kind of changes that. Um so yes, so very early days. So that it's really thinking around um how is the customer behavior in terms of shopping for. So I think we talked about MCPs, right? Is gonna, I think, really disrupt um how um our customers are using the product. I think that's also what how our customers are discovering, researching, finding out, right? Education, educating around products, um, definitely seeing that the chatbots help to kind of uh tighten the funnel a lot. And so that actually is for both um self-serve reasons. So, you know, instead of browsing potentially maybe a couple pages, it's a way to really, you know, aggregate. So it's to have your own kind of AI agent right on the site to say, okay, you have this question. Let me aggregate this information for you and um share it up to you. Or yeah, um, being able to connect you with sales. Um and I feel like that's it's a pretty common and you know paradigm now across the the different um the the SaaS websites. But um yeah, I think that's that actually was kind of an early application, I think, of of how you know I I had been using AI in in previous um roles is we had all this rich data from the chatbots.

Kevin Kerner 19:46
Yeah.

Malinda Sandman 19:47
So, you know, it's a combination of um, you know, you could you could, as a marketer, when you're trying to get closer to your customer, you can sit on a lot of calls, you know, with the sales team, right? Talking to customers. You know, that's changed a lot of times in terms of all the tools to be able to aggregate it. Um, you can do the same thing in terms of how you're seeing on the self-serve funnel, you know, what's happening on the chatbot and being able to get you more web data, or maybe there's a web intercept, you can consolidate it. But I think then research flow basically puts it all in kind of one tool. And then you can action off of it too. Um, so it's just kind of incredible how much, how much richer, right, all of this context is. Um, now our challenge is gonna be, you know, how do we then how do we now act on it and how do we kind of consolidate? And you want to you want to talk a little bit about kind of like workflows and things like that as well. Um yeah, it's it's incredible how much it's changed only in a couple of years.

Kevin Kerner 20:36
Yeah, it's really amazing. And there's so much context because you do have now you have your chat bot, you got the research data that you have, you have your gong data or however else you capture the sales data. There's just so much in it now uh that sort of gives a lot more context. And then pulling that all together and making something, you know, useful out of it is really you know, that's the challenge, you know, and doing it quick and making the right decisions as a human. Um I'm wondering where oh, go ahead. Go ahead.

Malinda Sandman 21:04
Oh, I was gonna say, I think the, you know, the other thing that um when you say kind of we have so much data, right, is it's also then, you know, how do you lean on the right or how do you have the right type of training to be able to know what to do with that data. Um I think, you know, when I was talking around kind of our our our research team and kind of our philosophy at typeform too, um, I think the the difference between having kind of tool versus thinking, all right, I'm gonna I'm gonna have kind of different data and different context in different places, and then how do I consolidate and how do I um how do I listen to it? Kind of back to that that squeaky wheel biases, even if you're consolidating through, you know, through AI using kind of all of those different um pieces of information, where where do I actually see um the spikes, right? And where do I think about kind of the qualitative and quantitative? So I think having that kind of expert research guidance built in is gonna be that next level for marketers. Who are not researchers, right? To be to figure out where is the context that I do want to listen to versus where is maybe something that's not statistically significant.

Kevin Kerner 22:09
Yeah, that's real. That's super interesting. We work in the supply chain space a bunch and they have control tower-esque language when they talk about the supply chain. Something goes wrong here and it spikes and then this thing happens here. That same type of need, really, probably even more so is needed in research because we're not professional researchers. We don't know the I don't know what statistically is valid. So if you if there's some way to enable that using the AI, which I'm sure there is, especially nowadays, um that's really useful. I actually hadn't thought of it. I was I think people are just so used to just chatting with the thing um and getting the answer. Uh, but it's almost like a skill-based, you know, tell tell me when something spikes and something's useful for me. That's a really that's a really good use case. I hadn't thought of. What's the um what's the what's next on the agenda for you as a CMO as you get these product launched? Like as you look down the next next six months as you're getting the things launched, what's on your agenda? This be helpful for people to hear that might be launching products that are in market now. You now what?

Malinda Sandman 23:14
I I mean, Kevin, the work to be done now is to really show the market that Typeform is not the typeform that they've expected, right? The typeform that they know. Um, you know, and that and it's really kind of getting there and telling our story. So, you know, I think I think tactically it's around we still want to make sure, right, that we're serving this um, you know, incredible customer base with our best in class conversational, like human, human first forms. But now it's the you can do more, right? And so we want to make sure that um, you know, we get across that message of the form is still so important to running your business. And you know, what that means is like, right, the data intake is is key, but now you can you can act on all that data within type form in a variety of different ways, whether it's I'm gonna grow my business and I'm gonna do lead gen and enrichment um and be able to do automations through GoFlow, or then you know, with what you talked about with research flow, I can take all of that data and now make decisions and I can even act on some of those decisions within typeform. Um, you know, the research flow can create um a highlight reel for you to share with like leadership. Like as a marketer if you're in a bigger company, that's cool. I've done this and now I can go share this highlight reel. Or I can get a smart we're working on a deck. So, you know, in terms of kind of like thinking about the roadmap and where this goes, it just continues to be more and more um all in one. You know, uh, one of the things we've talked about is like an insights deck, right? And so once how does research flow will just create this here. You go, ready to go, insights. Um, and so the time, both like you know, feeling like the marketer has now you really feel confident in your decision, then you also have freed up time not to go work on other strategies, right? And work on something else um is incredible.

Kevin Kerner 24:56
Such a great highlights real is such a great idea. Man, that's cool. Because I I uh I'm outputting a lot of stuff. I had to train a junior dev this morning on a thing that was going on that we were doing, and I I it was very complicated. So I just did an HTML view of it and made it J junior dev uh readable and he we reviewed it this morning. It was better than anything I could have ever done. So it's like finding new formats to get things synthesized in the context that whatever the person can understand senior executive, junior dev, junior marketer. I got I gotta look at one of those highlight reels, I gotta see what that thing looks like. This sounds really cool. I'm imagining like on all trails when you're hiking, there's this little you can view the like mountain you're going up and where it's gonna be as little roadblocks along the way, something like that. It's a great idea.

Malinda Sandman 25:46
That's a great visual. Yeah.

Kevin Kerner 25:47
Yeah. Once you um when you think about budget allocation and where you can put like channel uh to to launch this as a small rel, you know, not a very small business, but you still have to be efficient. Are there any things that you rule out? Are there, I don't know, are there any any channels or tactics that you'd be like, well, this is where we're gonna lean in?

Malinda Sandman 26:09
This is also the the other part of the the fun of um research flow. I think in the the challenge uh for for me and my team is you know, this is um this is a product that we think has enormous relevance up market. So it's kind of twofold here with research flow is we think this really unlocks and democratizes research for SMBs, uh, you know, and smaller businesses that have never had access to this type of research, um, which is gonna should be a a a complete game changer. And then as you think about kind of mid-market and into enterprise, this is a tool that could maybe aid a in-house research team's roadmap. There's there's never, right? Like roadmaps are always completely full if you have a research, you know, in-house research team. There's always more questions than you can answer.

Kevin Kerner 26:56
Especially now. Yeah, especially now. Yeah.

Malinda Sandman 26:58
So it can be a supplement or maybe can help um from a budget perspective. It's so much faster, so much cheaper than a traditional research study. Um, so I think it can unlock and kind of supplement, right? Research that's already being done, or actually create this entirely new category that doesn't really kind of exist um in an SMB's um, you know, budget or time. So we're thinking about it in two different ways. Um, and I think with type form, you know, historically we've been um majority self-serve, and we've we've got, you know, definitely reach right into um a lot of the enterprise and the Fortune 500. Um, but this is kind of about building two different motions now. And so as a marketer, it's really thinking around, you know, we have um really great tried and true, and we've got an you know an amazing kind of PLG engine. Now we've got to think about how do we go and reach kind of from a um a sales motion, these bigger companies. So yeah, we're definitely thinking around a lot of things to invest and expand into. And so yes, that from a budget perspective, it's kind of all around that mix. Um, but I think it just goes back to continue to iterate. So we've got a lot of tests coming up in the back half to try to figure out um where can we find um where can we find the these um these customers because some of them may not be on typeform, and you know, how do we sell into the researcher, how do we sell into the marketer, how do we sell into the product manager?

Kevin Kerner 28:14
Yeah. And it's in and in that category in the enterprise, research is important, but it's probably more like a feature than a than a large platform. They wouldn't think of it in the way that they're thinking of that if they should be thinking about typeform. It's not a hot, it's not like a CRM platform. It's not that level of it's more of a feature that you use to to to improve one part of your organization. So it's a really interesting challenge at the enterprise level because they might even be thinking about, yeah.

Malinda Sandman 28:45
Yeah, and we have a we have um we were saying we have an incredible research team um with PhDs right on this team that um that are what we're calling kind of in our ICP, like our expert researcher. Like this is the team. And you know, they've talked about the fact that this is this just becomes a tool that they can use too. So it I think it definitely has um, you know, extendability to be an enterprise platform that a research team can go in and say, all right, I have 50% more requests from my stakeholders and I can actually manage this quarter. Research flow now help me actually be able to fulfill all of the requests, right? Maybe and then some because of how fast I can work now, um, how much cheaper it is than even going out to any of these market research firms. Um brainer. Yeah, no brainer.

Kevin Kerner 29:33
No brainer in terms of the cost, the cost angle. That's great. Well, a lot of a lot of things to discuss. I could ask more questions. I want to get through three. We ask uh every guest on the podcast three questions to create somewhat of a benchmark. You're the second uh person that I've asked these questions to. So as we get more data, we're gonna put more data out on these. So let me ask these questions one at a time and and I'll get your thoughts here. So question number one is what's one AI workflow you put into production this year that moved a real number and one that you killed when you said didn't this is not gonna work? AI workflow.

Malinda Sandman 30:06
So for I don't know if we consider this a workflow, but I think what we're really seeing that's moving the needle is um the speeding up the time to market to do experimentation. And so when we think about kind of prototyping, so it's kind of twofolded. There's the okay, how do we kind of speed to market from a prototyping perspective to just get more creative and to market to be able to get more signal um and and then get and get clarity. Um, so definitely on the A-B experimentation side, um, you know, like using Claude to be able to create prototypes of like the homepage you see right now, right on typeform.com. Um, that this was one of the um kind of experiments and workflows that we did. And so now we're able to kind of start to roll this out kind of across the board and and seeing. And really what I think it is, it's also by prototyping, it's helping us kind of create that shared language between strategy and execution. So we can have a lot of the marketers be able to really quickly say, okay, I'm I'm here are my requirements, here's kind of what I'm thinking, here's the messaging, here's what I'm trying to get across. And so the creatives have a real-time example of what that looks like. And so I think it's it's cut out a lot of the okay, I'm saying I'm I'm here and marketers are here, maybe creatives are here, and we we're we're we're going through several iterations to get to the on the same page. Um that's one that we've seen, yeah, through A-B experimentation that's definitely moving the needle.

Kevin Kerner 31:26
Love it. Well, is there anything you're killed? You're like, ah, this isn't gonna work.

Malinda Sandman 31:30
Um, you know, on the the killed side, um, I think where we're going is more on the how to create more AI workflows. So I think we're, you know, as I'm sure kind of everyone is now, I think everyone's pretty proficient. Um, me in the tech marketing space of like, I'm using AI for a lot of one-off use cases. And so I think what it's less about like what we're not using, but more I think where we want to go is the how do we now like templatize and get everyone on the same page and say, okay, I created this skill, right? Or I created this gem, I'm gonna share it with everyone, and this is now how we do X or Y. And we're all gonna do it in the same way, and then we actually all have the same information feeding into it as well. Um, so I think that's really more where it's going. And we're so we're trying to figure out how to evolve the one-off, which has definitely sped things up into a much more structured, uh, I think like templatized processes for for a variety of use cases.

Kevin Kerner 32:23
Yeah, that's good. That's great. So you haven't killed the one-off, but the one-off needs to like move over to be scalable, like get to the one-off fast so you can scale it into something. I love that. Um, second question is what's still stuck in pilot? Something you expected to be running by now that isn't?

Malinda Sandman 32:39
This one is is very similar to I think the my last example, and maybe my my um specific example here is when you think about um, you know, kind of everything we've talked about today of like there's so much in any company, right? There's so much context information, um, ability for whatever AI tool you're using to be able to summarize all together and give you the output. I am surprised that we haven't um we're working on this, uh, that we haven't evolved our marketing briefing as much as we have. And so, and use that. So, one of the things we're working on is we're using, we're working on a um a briefing agent. But when you think about we know what's really hard for marketers is how do you take all this information? Some of it's in people's heads, but a lot of it is out there right over across a variety of different decks from your messaging and positioning deck to maybe some information of what happened on like in a campaign or what happened on this website. You put it all together and say, all right, here's my strategy, right? Here are here are my audiences. Um that that's one of our hypotheses of like this has to be something that we can templatize and probably make um much richer, right? To be able to move faster in creating um amazing creative.

Kevin Kerner 33:50
Yeah, I love that idea because we're we're we're doing the same thing ourselves. We go with a tool called Flow, and we're finding that when we do client strategy, like we have one client who has 26 different brands, and we're having we had to build website strategy and actually build all the sites across 26 sites. And so we created these skills that like gave us this knowledge set around these individual brands, and we used them for the sites. We're looking at it going, oh my gosh, this is like you could ask this brain any question about any brand or a combination of brands. And so we put that in a in our tool called, which is also called Flow, which is uh which is interesting. It's an end it's kind of an agency uh uh tool that we created ourselves, and um we call it brand intelligence, so it allows people to not only uh talk to that brain, but also it um documents their current brand, you know, colors and types and you know, pictures and words and that type of stuff. And then you can chat with that brain. Yeah, uh, so you're on to something there. It's really we have the same vision is that you shouldn't have to do you should tell the brief that you're you want to you need an outcome, but the brief should have enough context to where you don't have to do all the things anymore. Yeah, it just does a lot of that for you. So really, really cool one. Um and then the last question is what's one thing your team won't hand won't hand to AI and why?

Malinda Sandman 35:13
I'm speaking for this as a people manager. The one thing I won't hand to AI is managing my team, right? Is build you know, is building the relationships. Um, you know, and then coaching and developing marketers. So I think there's, you know, there's a ton that we can hand over from an execution perspective. But when it comes to the strategy and even thinking, I think around how we're even using right the the AN automation, I think there's so much from, I mean, it's like our conversation today, right? Like learning, sharing, like swapping ideas, talking about how we're approaching things. Um, and I think that, you know, it it sounds kind of like maybe a little cheesy, but I think that's very much um everyone that I work with at typeform, right? When we when we think we put our money where our mouth is in terms of like what we're developing for our customers and how we want our customers to go out and and work with their customers, I think it's the same thing, right? I think we have that kind of philosophy of we need to be human first. Um, by being human first, I think we'll have such an advantage um in kind of working and growing and um being able to launch these amazing products. But yeah, I think it's that I see as something that will not be handed to to AI automation. I think so critical for that human connection.

Kevin Kerner 36:19
Meet your new manager. It's the it's this AI agent that I that I've created for you. Yeah.

Malinda Sandman 36:23
Because you can make it. You can say I'm gonna create an AI agent. Here's all the here's all the to-do list. Just tell me, tell me.

Kevin Kerner 36:29
You know someone's doing that. You know that's being done somewhere. I don't want to know where, but it's not that good. Yeah, that's a really that's really good. Okay, so I have one more ask of you. I do an AI roulette question where I'm going to ask my AI robot here to ask you a question. I fed in the brief that we talked that we're talking through here, and then um your LinkedIn profile. So I don't know what it's gonna come up with. So let me just hit send here. This is uh real quick. And this is from the AI, so it's then the AI's voice here. Okay. My file says web flow publishing went from three days to three hours. Automation did that. Question, what happened to the other two days in 21 hours?

Malinda Sandman 37:11
Is this making me a mad question?

Kevin Kerner 37:14
Yeah, it's basically like, hey, I think you must have I saw somewhere, or maybe it was in our previous conversation, about your web flow process was a lot faster now. It ought you automated a lot of it. And you talked about like prototyping. So if you say if if you went from three hours, three days to three hours, what are you doing with the rest of the time?

Malinda Sandman 37:34
Overhauling the rest of the website.

Kevin Kerner 37:36
That's right.

Malinda Sandman 37:37
We have a lot of things.

Kevin Kerner 37:38
There's always work. There's always work to do.

Malinda Sandman 37:40
There's always work. And you know, even when we get to the point where is we're working on this right now, is like continuing to make sure that all of our own assets, right, reflect our new product lines, right, and our new product vision, then our company vision. Um, but then you just you keep experimenting, right? I mean, yeah, what's happening in the market, right? The even the composition of traffic's changing so fast that, you know, what may be working in one month is not working in another month. So yeah, you're constantly iterating.

Kevin Kerner 38:05
Yeah, I don't know how you feel about this, but I don't see work slowing down. Like there's the AI doomers who say that this is gonna take all our jobs. And I just think there's gonna be a bigger backlog on everything.

unknown 38:16
I agree.

Kevin Kerner 38:16
Just more work.

Malinda Sandman 38:17
I agree. I mean, if you think about where even we are um with the ability to do remote work, right? All of the tools that were available at our fingertips before, you know, um uh ChatGPT rolled out, right? Like that feels like so much more work than people were doing 10, 20 years ago.

Kevin Kerner 38:33
Yeah.

Malinda Sandman 38:34
So I agree.

Kevin Kerner 38:35
Yeah, 100%. Well, Malinda, this has been great. Uh, I'm really pulling for you guys. I love the product. I'm definitely gonna go check it out right after this. Um, people want to follow you or get a hold of someone about the product, where do they go?

Malinda Sandman 38:47
That's a great question. Um, typeform.com. Um, we're on all all the socials, and yeah, you can feel free to also reach out to me on LinkedIn.

Kevin Kerner 38:55
Sounds great. Well, thank you so much for your time. This has been awesome. I'm sure a lot of people will love it. And uh have a great rest of your day.

Malinda Sandman 39:02
Thank you so much, Kevin. You too. Really enjoyed our chat.

unknown 39:04
Okay.

Kevin Kerner 39:04
See ya.