183 - Verbal identity as the antidote to generic AI content with Ken Marshall

Welcome to Product Led Growth Leaders. We've got a great guest for the show today, Ken Magma Marshall. He has a background in SEO, has started a bunch of companies. And today he is the co-founder of Meet Sona, which helps you create content for LinkedIn posts based on voice interviews.

Ken, welcome to the show. Thanks for having me, Thomas. You know, great set of hair, and I'm ready to rock. So let's do it.

For the listeners, we we've we've got a very similar hairstyle. I think this is cooler though, because you've got the shave side going on. So, so okay. So let's let's get into your career.

I mentioned there that you have a background in SEO, but you did a lot there. And I think you used it to segue into the stuff that you're working on today. So how'd that come together? Yeah, I mean, you know, without taking up too much time, I started off in sales and business.

I actually like won a sales competition in college. I went to, you know, go accept my prize and realized that I was going to be a rich, bored, like depressed person someday. So I got into like coding, web development, and realized, oh, I can create something online, this sort of new wave that people notice in a second. Um, and that evolved into yeah, SEO, content strategy, and marketing.

And eventually I just decided that my boss didn't know more than me. So at age 23, I just started a company on Thursday and then on Monday I was cooking. So that led me down, yeah, seven years of just mostly for um SaaS companies, like early stage venture-backed or uh private equity-backed SaaS companies. Sold those two agencies two years ago.

And I'm like, you know what? We help founders and consultants and CEOs do all day long, grow with thought leadership, unique stories, and their ideas that only they can tell. And so I built Meet Sona for myself, started offering it to people, and seven months later, here we are. Excellent.

I love that story. I mean, we've all got to take the things that we learn at earlier parts in our career and just as times change, technology changes, you find new opportunities and use the skills that you have. So let's steep ourselves in the problem that you guys are solving. So obviously, you know, I uh it's heated up with the uh LinkedIn posts and everything like that.

But let's let's take some time to appreciate the problem that people face. Yeah, I if I may break it down to the three like constituent components that I hear most often. The first is a capital and a time constraint that every busy executive has. If you have employees, if you ever run a company, you know you can't spend four hours developing your taste and writing from scratch every day.

Anyone who says you can't is lying and has not been an executive, I promise you. Um, and so that's the first major constraint. The second, and the one that nobody believes until I open up, you know, a recording and say, speak to me about what you do, is self-esteem. A lot of leaders with actually a lot of executive presence lack the confidence to speak about what they only they know in the way that they feel comfortable with.

They sprinkle thought leadership tokens on it, they change their style and the templates to fit what's viral. And so that gets me to the third thing. In a world where Opus 4.7 can replicate anybody's style with absolute precision, what is the thing that's your differentiator in the market, Thomas?

It's unique specialized knowledge, experience, and your own taste and judgment. And that's my thesis. Okay, so so this some things to unpack there. So the the aspects of the problem, the capital constraints, yeah, because and that's changing.

That calculus is changing a lot, right? So if I wanted to start some kind of um video campaign eight years ago, um that would take a different amount of money necessarily than um then, but the landscape has changed. There's also a lot more content. Um, the time to do it, and that's one of those tricky ones because I think we haven't gotten into how you deliver the solution, but I can make some guesses as to how that you are creating content from small sets of kind of meetings and turning it into a lot more than it would have been or could have been.

And then um I like the self-esteem point that you're making because there's a lot of folks who are pretty confident in being able to lead and um and to be in charge of things. But when you put them in front of a microphone, they might not really feel that they're the they're the person who can kind of flow with it and talk about it. And um, when you're creating content, that's something that's important. So these are problems that you noticed and you thought that you, hey, you know, let me go ahead and try to tackle these.

Yeah, the and the reason I start with those sort of fundamental issues is because it maps completely onto my methodology and the how we design this the system within the product. So the from those statements of if you if you believe that those things that I just said are true, then the issue becomes how can we leverage this, you know, first and any generation technology that's before us, with the fact that we don't want to just put out, I think they call it slop nowadays, like nonsense, basically something that doesn't add um any intellectual value or entertainment value in a way that is repeatable and in my opinion, enjoyable, which is my third pillar.

I want people to have fun while they're doing it. And so the way that that's done is you you nailed it. Um, I read a Harvard Business Review study about people that write things down, talk about things, and then engage in a conversation. And I think they essentially rated that their ideas were more uh distinct, so like the uniqueness factor, and that they liked them more out of being drawn out from a conversation.

So, you know, what we essentially are doing is saying, let's develop this verbal identity like algorithm, right? So verbal identity is your unique personality, what you always say, never say. So some tools will say, Hey, Thomas, you know, if you go on Claude, do you want to sound modern? Do you want to sound funny today?

And we're just like, no, no, no, no, no, no. All you have to do is show up and talk, and we'll um extract and then isolate, you know, each of those components based on a matrix, you know, blah, blah, blah, boring stuff that we've developed. And then we'll ask you a bunch of questions back and forth in a conversational interview. Take the transcripts and use the transcripts and your verbal identity as the foundation for the content.

And then whether you're on LinkedIn or X or Medium, we have different platforms. We just provide the template and a little bit of polish on to you, that you are at the center of everything. And that's what makes the system faster, cheaper, but then with your own taste and judgment and stories at the heart of it, which is what performs well on platforms like LinkedIn or in Google Search now with their uh trust embedded in the algorithm. So yeah, and I think that's part of the key, right, Ken, is getting it to where it's easy to produce the content.

Because if you, you know, you try to say, like, okay, well, we're gonna create, you know, a hundred videos and and you overthink it, and you you get a big team and cameras and everything like that, they're never gonna get it. Yeah, and and and you've probably had this mind. I think a lot of listeners probably have had this moment where you might be just having a conversation with a friend, like over the phone or over coffee, and you both might be saying funny things or interesting things.

And have you ever stopped to think like, what if there was an audio this was being recorded? I think that you actually could have like other people would want to hear this. Like, you know, it's a shame it's not being recorded, right? But that is, I'm not even kidding.

That's what the first thing you hear when you get on board. I guess I'm giving away this the surprise, but is Sona's like, don't overthink it like you're talking to a friend over coffee. Because when I do the so my methodology is based on my like hundreds of interviews, you know, with executives and founders. And at first, they're like so tight and nervous, and I'm just like, You ever heard of free association, like how they teach you in psychotherapy?

We're gonna do that. And so you're you nailed it. You nailed it. Right.

Another, another kind of commonplace example is have you ever been in maybe a social setting and you're around friends, and then you see maybe two of your friends talking, or you see a certain moment, you grab your camera and you kind of want to capture it, and then they notice there's a picture being taken, and then they straighten up and then they start doing a fake smile. No, do what you were doing before. I want to capture the live moment. Yeah, and and and some people think that's um um like amorphous and squishy and not tangible enough.

And and my point is not we're we're we're not replacing the yearly report that Gartner puts out. I'm not claiming that there should be one type of content that has your stories and ideas that's emotionally resonant at the heart of it, but everybody seems to want to do the 20% serious thought leadership stuff, and they forget that somebody's having a bad day somewhere that needs to just hear your unique perspective on a take to let the information sink in and integrate it into their lives.

And people completely disregard those other four categories. They just want to do the yeah, information. But no, no, literally, it's it's it's more even than that. It's it's that if you studies have shown that if you have uh stock photos of people on uh where it's like the two business people doing the perfect handshake and they're smiling as they're doing it, right?

And when you have stuff like that, people literally, if you look at their eye movements when they're scrolling a page, they like pre-see that it's fake and their eyes don't even really jump to it and and look at it. They just kind of scroll past it. It's similar to this other phenomenon. Like ad blindness, but yeah, yeah, yeah.

Banner blindness is is um, and they're all forms that you from the my design psychology side, they're all forms of inattentional blindness where our brain learns that's not relevant. I don't need, I'm not gonna get because your brain is trying to survive and create a model of the world. So your brain learns that's not gonna give me anything useful. That's a fake thing that's just there to look nice.

So we intuitively think, like, oh, this is the picture we should put on the page because they look perfect and they're smiling. What you do want is something a little bit messier because it's believable and it looks like something that we should be looking at and getting information from. Yeah, I mean, well said. I didn't know any of the um specifics about that.

But I it's like I call it uh AI blindness. We have a one of my blog posts talks about like why we I we start with your stories and your lived experiences at the heart of then, even if we're gonna dive into thorny topics in the interviews, it has to be with your use uh lived experience and stories because um I just I'll be reading, you know, in my feed or I'll be reading a blog post. And not that I think actually humans are about 50-50 on being able to tell, interestingly, of what's AI or not.

When I I think the test had some flaws. But anyway, I feel like it is, and so I check out, and I'm sure you do too. And so yeah, I think I think what people are maybe we're early, but I think in five to 10 years, that proliferation of AI blindness, because most people, you know, in our industry, we're in a bubble within a bubble in tech on LinkedIn of use, nobody knows what cursor or lovable or opus 4.7 is when I go out to dinner with my homies.

Literally not a single person has used pawn. But in 10 years, when I think that proliferation of this kind of blindness that you eloquently described, I want to be there saying, you can give good information, but if you lean into your unique verbal identity, your stories, the dope life that you've lived, you'll capture people's attention. And from that attention, they'll be able to take what you said that helps them, but couple you with it. Whereas if you just give them the information, you're no better than the robot.

So do both. Have them like and respect you while giving them good information. Now, let me give you an example of a service that I used to pay for um a few years back with largely positive results, but you know, a little bit expensive. And I did it for a period of time and stopped.

So um, what I liked about this service, and and like you two opine on this and compare it to what the solution that you're offering, um you I would meet with the team, and they would record me for you know an hour, hour and a half, answering questions, kind of talking to them. And, you know, and if I said something that they thought was good, then I'd repeat it or whatever and try to say it. Um, and after this, they would go through a post-production process of going and finding the gold within it and editing it, and giving me video clips, and then we edit certain things about it, the animations around it, and and so on.

And we would end up with a library, and then we would kind of like have this periodic meeting of I would kind of meet and and do all of that. Um, and that's a framework that I found pretty good, better than like me on my own, saying, like, okay, now I will create a video, and then I turn my camera on. You're never gonna do it. Yeah, exactly.

So, so so what are your thoughts on that? And how, and extending from that, how does your solution kind of come together for the person who's doing it? Yeah, uh two, two things on that. So, one, um, we did that at Revenues End, the agency, right?

It's just we had a team and it costs between seven to twelve thousand dollars a month. And so it's not that that isn't valuable, it's just picture yourself as you know, you have three employees um, or you're an early stage company, maybe you didn't raise, like you're just bootstrapping now. Will that $80,000 potentially get, you know, get you your return on investment? Perhaps.

But the CAC payback period is just unacceptable for a lot of early stage companies. Or, and that's just software. If we go into consultants or coaches who are about 70% of our uh customer base and user base, there's a 0% chance most of them are gonna make that back in a reasonable period to pay their bills and for their kids. And so everything that I'm trying to um build into the product, and I still do interviews, actually.

We my co-founder was like, How are you gonna sustain this? But I still do workshops with people myself to like teach them how to do an interview so that they can use the product better. But it's the fact that it's at a cost where it's in the middle of the concentric circles. I mean, I think the pro plan is like 59 bucks a month.

So if you're ready for a ghostwriting agency or a premium content marketing agency to interview you, I've got some homies who will just like blow your mind. I mean, they're incredible, but you have to shell out the 80K for the year. You're gonna get excellent quality, but you may not make your money back in an acceptable payback period. So, does that kind of make sense?

We're going for the middle of both of those from my experience. Yeah, absolutely. And and one of the things I've been kind of told to think about from my various, you know, folks who help me produce content, you know, I do the podcasts and I do stuff online. There's um, I forget the term, but there's one type of metric that's sort of like, hey, I'm winning on social media and look at my statistics.

And people, this many people liked the comment and this many people viewed, but it doesn't map on the revenue. And then there's the more kind of pragmatic thing of like, are we getting an actual return on investment for what we're doing? So one might make you kind of feel good and like you've made it, and you, you know, you can show people like, hey, we're we're growing, but it's like you really the need for the real stuff really is going to kick in at one point in time or another.

Um, so when folks use your service, let's let's pivot to talking about um, we start using um uh Meetsona, and we I I I fire it up and I'm I step in front of it and I'm about to use it. What is that experience like? Yeah, so first you're gonna people hate this, but it's because of the second thing I mentioned, the self-esteem and confidence aspect. But we ask you three questions to onboard you.

We're in in three minutes, we're simultaneously like we have a LinkedIn sign-in. So we're looking you up for context, which means you don't have to upload a bunch of stuff or do a bunch of prompts because I have many Claude projects and skills. I have Claude Cowork, I have Claude Code. I love building context files, my man.

I love connecting my own digital brain. I call it Kenba. However, the average American who is not doing what we're doing will never do that. They will never achieve that level of um proficiency.

And so while you're talking for three minutes with three simple questions, we are simultaneously using the context of information we find out about you. If you sign in with LinkedIn, we are starting to build the beginnings of your verbal identity. Remember that if you ask somebody from a blank page, and I've done this many times, to define the unique aspects and characteristics of how they speak, their style, and how they want to present to the world, most cannot answer those questions.

They can tell you their ICP and their business model. They can tell you about their competitors, but when you ask about them, they struggle. So we're building that profile. And then all the other, you know, sort of boring onboarding things like what you do, who you help, through three very simple and kind of deceptively um specific questions.

From there, you know, you go straight to the interview itself, right? And so by that time, we've already calculated enough to have a very relevant first interview. You answer five to six questions. Sometimes there's follow-ups if needed, sometimes there's not.

Some people just like to kind of burn through in about five to six minutes. But in six to ten minutes, you have about 1500 words on average. Um, if we round down. And then from there, you click a single button.

You decide if you want four or five LinkedIn posts, a blog post, two newsletters, and an ex post. You click go, it create it takes about two minutes. It creates all of those content pieces simultaneously with the best practices built in per platform one at a time. Um, just parallel, but it builds, you know, them at one at a time.

Um, and then once those are done, you click schedule and publish to LinkedIn, it picks the best day and time for you and sends it. And then you go do the actual work you're supposed to be doing. The point is not to make the platform sticky for hours, it's 15 minutes maybe a week for three to four sessions per month. You know what I mean?

And then get out of here, go sell. So yeah. Okay, you know, you're making me curious about this, bro, because I I might actually literally uh check this out. Um for yeah, and I know you're not trying to sell, but uh you're you're doing a pretty good job regardless.

So um, so let's talk about the interaction. So I I love that that you you uh you make it easy to set up. The typical person is not gonna know how to generate MD files and upload them into Claude and right and and do that efficient setup with the projects where it knows all this stuff. They're just regular folks trying to run their business.

And then I like the output part. Do you somehow manage to capture how do you capture the, and maybe if this is your secret sauce, you don't have to get too deep into it, into the weeds, but capturing that naturalness of the conversation. So just like the uh podcast producer who says, uh, hey, uh, you're great, but we're gonna get you a co-host so that you get you can have like the back and forth banter and the co-fo-host is gonna have like a comedic background or something like that to contrast you and you'll it'll great create good content, uh content easily.

How do you capture that? Is that um is that the AI uh chatbot doing it? Or um, you know, how how does that come together? Yeah, so we're actually using uh this this is not sauce, but it is nerdy technical stuff.

So people who don't care, uh tune out. But but um we were experimenting with um 11 Labs real time and then open AI's real time API for a truly natural, like like Sona would interrupt you. And and it's it's fascinating because this is why, this is why listening to your users, watching what they do and interviewing them is so critical. I'm like, we gotta go more natural, we gotta interrupt them more, we gotta move on naturally.

And everyone's like, whoa, whoa, whoa, whoa, whoa. I didn't have time to think. Give me a button or something to like pause or have Sona wait. So we actually recently, I'm not even, I know this might not be the coolest thing to say, but we took a step back.

It's not actually as um informal as you going back and forth with a friend, interruptions. We've we've um kind of pared it down to Sona will ask you the questions, you have the ability to control which and when follow-ups are given. And then either auto proceeding like you would in a natural conversation, Sona will listen for pauses, or if you're the kind of person who takes a lot longer to think, you can toggle that auto proceed off. So we went uh beta was like really pared down, no natural language at all.

Then it was like, let's go full real-time conversational. And now we're kind of just retaking the steps based on our customers' feedback. So yeah. Okay, so this is really, really sounding cool.

What um what does the price points look like for this something like this? Yeah, there's a seven-day free trial of the creator plan that's 24 bucks a month, and then the next tier is 59 bucks a month. And the difference is two things. One, you can't really grow only doing 20 pieces of content across your whore whore portfolio.

So, like get your feet wet, get your 20 pieces on the the other plan. But then the most important thing on the um uh the $59 plan, other than just the the rate uh limit in uh increases, is that we just got approved for LinkedIn Analytics. They are trying hard, they are coming after companies who use their API incorrectly. So it took just under, I think, three months to get approved for analytics.

So that's being offered now on the the $59 tier. And all I'll say is that before I had a thesis that this kind of content, I mean, through my own posting, through my business partner, that's where we generate a lot of business. Almost all of our customers and a good chunk of our users have come from my profile. But now the analytics are going to start to say, not just what the thing I didn't like about other tools is they said, here's your impression count.

This post did this well, you know, you got this many shares. That's table stakes. The layer underneath that that I'm really excited about is hey Thomas, when you talk about um presentations that you've given, it performs 19% better than baseline. When you use a contrarian hook on your ex post versus you start with a story, it performs 100% better in terms of engagement underneath in your feed.

Those level insights about you are what's going to be at the heart of our analytics product, not the table stakes, comments, and impressions, because I believe that your own personal development is completely consistent with the growth of your business. And I'll argue that point with anybody. So we want to help people change and get better as their strategy changes and get better. And that's just my heart, you know?

Amazing stuff. No, what does the compete uh competitive landscape look like in this type of tooling and this kind of software? And where do you guys see yourselves in either that ecosystem or um high on that list, low on the list, underdogs, uh early uh first movers? Yeah, the category is early.

I mean, I think there's legitimately like 10 companies that use actual um, you know, uh conversational AI voice interviews precisely because it's it's uh it's difficult. So when you're doing 50 concurrent interviews with other users, right, and uh one of the web sockets closes because uh open AI or Claude uh goes down, tell me what to do about that, Thomas, right? How do you ensure redundancy for your WebSocket architecture? Or you know what I mean?

There's all of these considerations from a technical perspective that just um putting a wrapper around Chat GPT to get some written content gives you. And so the category looks like this there's a ton of chat products. I'm not a chat product, I don't want to be a chat product. I hate the little boxes.

So we're not even gonna compare to those because the the fun the foundational models, they own that category. Then there's writing products. Think of like writer or Jasper or Grammarly. These are writing tools for people that want to write.

And writing is excellent. I write every day, I have a journaling practice, I wrote a book, I love it. It's fun. But you cannot get an expert who's not used to writing to sit down on a blank page and crank out five pieces per week.

If you can't, I'll give you a thousand dollars. But like I haven't seen it. So most people struggle with writing from a blank page. So that's the other part of the category.

There's voice note tools where you just talk into it and it gives you a piece of content. It's good and it's a step in the right direction, but it's incomplete. Again, I trust that Harvard Business Review study from my own anecdotal evidence that when you ask people questions in a specific way to extract their uniqueness, you get certain parts of their brain firing in ways that I think they wouldn't have connected the dots prior. So we're going fully to the interviews.

And then, like I said, there's probably 10 companies and two have my respect who have been who've raised capital and have been around for many more years. We're seven months old. So I think we're up and coming. But I think the fact that I've done this at an agency level for so long, and my goal is to help people self-actualize, not just create content.

We're gonna carve a nice little niche that I don't even know if is a good idea or not, but I'm doing it. So do they work with you on setup or is is that a fully automated onboarding? Fully, uh, you know, we're at the PLG podcast, so we've done so many rounds of onboarding interviews, but yeah. Nice fully um, but I they I can do a um a workshop style onboarding for an extra price as well.

Ladies and gentlemen, we've been speaking with Ken Marshall of Meet Sona. Check it out at Meetsona, S-O-N-A, Meetsona.ai, and uh see where you can take it from there. Where else do you want people to go, Ken, to uh keep keep up with you?

Yeah, uh Ken Magma Marshall, type it into Google or Safari. You'll I'll pop up the SEO guy, uh, our was. And so have a chat with me on LinkedIn. I I share exactly what I'm working on in the same way that you guys should use the product.

Awesome stuff. Uh look forward to see where this all goes. Ken, thank you for joining us today. Yeah, cheers.