Jason Harris, Mekanism: Agency economics after acquisition and generative AI in creative advertising

Welcome to another Founder Wisdom podcast. We have Jason Aris with us today. Jason is a co-founder and co at Mechanism with a K. He's a podcast host uh all about soul and science. He's a best-selling author uh co of the year by the drum uh campaign US uh 40 over 40 in NYC. I see a Hampton member as well. Hampton's a pretty nice uh mastermind group for high level CEOs.

So today we're going to discuss about um his background and everything he's achieved. So Jason, welcome to the pod. Uh tell me a bit more about yourself and what you're up to. Sure. So uh yeah, I've been in the you know pretty much the only business I've ever done is um running creative advertising agencies. I've been doing that for uh 20some years.

So I started a few I started a branded content production company was the first company I started uh did that for a couple years and then I I'm still running Mechanism which I uh started in 2006 we're a full-ervice agency got about seven offices um you know hundreds of employees I sold it uh to a private equity group in 22 and uh I'm still still on on

on the uh uh earn out train and working at the business and growing it. Very cool. That's that's that's what I've I've you know I'm an entrepreneur by heart but I'm really been focused on sort of one type of business which is in the uh marketing advertising space right and since you sold out what changed uh in your life like financially I'd guess you're doing better but uh sometimes when I see founders being acquired they get sort of bored working for the business they that they've just been acquired into.

So how's it been for you? So I I wouldn't say I'm bored. Um most of the of the team that kind of built the business with me uh stayed on. So we're still here and we're doing a lot in terms of you know we expanded into some other markets. So that's interesting. I'd say the biggest change though is uh it took me about a year to deal with having a boss and that is a big change for a founder or an entrepreneur uh to have to actually report to someone or you know maybe if you're running a big company you're already reporting to a board uh but I had a double whammy of reporting to a boss and having

to prepare you know a couple times a year to report to a board and so that um that first year was rough for me. Um, it was rough to uh, you know, learn how to think about other people when making decisions and to think about the everything is about the financial impact. Um, so I I had to really change my the way of working and that took that took some time.

Right. And in creative agencies, what has changed since uh J Chat GPT4 dropped in uh in 22? I mean, what hasn't changed? Um you know, it's it's all the AI tools have really uprooted the the business and I think um costs have gone down. So you know fees have gone down because whether you're you know using these tools to a large extent or a small extent client the expectation and the perception from clients is that you've you're doing more with less people.

So your cost should be less. So the you know FTE model uh should should kind of change and reflect these tools that are now at your disposal. So that's one big thing that's changed. I think it's also impacting um because of Gen AI it's really impacting production budgets. So what comes on the big or small screen uh that's expected to be you know much much less expensive to make as well because can't you just enter a prompt and and make it the way I want it and have it come out and look amazing.

Uh so I think those you know doing more with less fees have gone down and it's really impacting on the production side. I'd say on the on the optimistic side of it there's you're you know you're skilling up workers to perform better and to have better creative to get to ideas quicker. um to have more uh variety and more uh options when you're when you're picking out creative.

The ability also to make uh unlimited content or tweaks or changes uh at scale quickly uh is is a good thing for businesses and for marketers, but it has um radically radically changed our our business. Yeah. Do you got like all these big clients like Pelaton, NBA, I've I've read Rosetta Stone. Like I need to know your secret on how you got these clients because generally I'm pretty bad at reaching enterprises.

I do GTM as a service with top leads and think you need to be so patient with these um larger business. So longer sales cycle uh they take forever, lots of bureaucracy. What's been your secret to uh closing all these uh flagship brands? Well, we um you know, typically in in my business, there's it's you know, you get an inbound inquiry and then it's a competitive um you know, pitch process where you might go through a month or two of pitching your idea of where you would take that brand to the client and then the client either picks you or doesn't pick you.

So, that's another um in the advertising marketing business space. That's the it's it's been like this forever, but it's the only business I know of where you give everything away for free in hopes that you win. You know, like you may have a one in 20 or 20% one in five, 20% chance that you land the client once you've gotten down to the finals and then you're giving everything away and they can decide if they like it or not.

and then you get you know the wind falls millions or zero. So it's uh that part is is is tough. Um but to answer your question the way we really built the business is you know when we started we would do we would proactively pitch ideas. We'd call clients and say, "Hey, we have a big idea for you." And you would do that to 20 clients and you'd get two responses and then you'd go pitch them an idea and see if you could land them on your roster.

Over time, you're known for your work and you know, after doing it for a long time, your name is you have a a name recognition in the market and then it's word of mouth. it's consultants reaching out to you to pitch the business and and so that takes a lot of you know that takes a decade plus to build up your reputation. So, you have enterprise clients coming to you, but before that, you've got to go out to them and, you know, have some type of value exchange where you're you're going to help their business for a low investment on their part so that you can get their big client on your roster, big brand on your roster.

That's how we build a business, you know. Yeah. Straight up. And I think that's how it is when you run a creative agency of of your scale. And yeah, most uh new agencies, new agency owners, they're quite naive, right? They just think that they'll launch an agency and one or two years uh after they'll be like super big. But yeah, these logos take time to acquire.

Once you acquire them, these attract more logos, right? And like you said, there's a one in 20 probability. Uh then there's like time and patience in that equation. So nice that uh you could handle you can hang out for so long right and eventually be being acquired. Do you think that um most agencies are acquirable uh nowadays like with AI? What is there to acquire like employees, SOPs, clients?

Do do you think agencies are good vehicle to to be acquired or is that changing with AIS? Well, I think the more valuable um media companies are very valuable now. So, in the advertising space, media um martekch companies are are valuable. Creative agencies are still valuable because they're the tip of the spear. So, they're that's still the way in.

That's the door into the client to sell other services. So, it's still very valuable. um creative agencies if you have a brand and you have a roster of clients but it's less valuable than it was before the advent of you know before the AI revolution it was more valuable then will it still be important yes creative agencies aren't going anywhere um but their um value might be but perceived as less and they have to take on more clients to get the

same revenue um but they're So, you know, we're like cockroaches. We're not going anywhere. There you go. I think like AI agents are going to be very valuable, right? Like because the AI agents are not going anywhere. They're actually increasing in their uh productivity and capacity and output basically. So, I think the agency of the future would probably have a bunch of AI agents and as an investor I'd be looking like, okay, what model is this?

What are they doing? And that that would be making like an agency very predictable almost transforming it into a a SAS right at some point. So like one CEO and multiple uh AI agents and then that I think that vehicle would be very viable as an investor. What do you think? I think so. I still think you need I I don't think it's as you can't boil it down to just that the agents.

You still need the editorial filter of experience, creative minds to um determine what the brand's going to stand for over time. That's still valuable. I think the you know there's there's definitely going to be a transformation between how many levels you need though uh for sure and I I see that coming but um you still need the you you know you still need craft and taste and I think that's something that you can't really replace with agents right uh could you also define like craft and taste and why these are

human super skills why they're not replaceable or what they not be replaceable and how do you nurture uh these skills? How do you scale them? Yeah, that's a good good question. Why they're not replaceable? I think I think craft and taste are um you know based on experience and what's worked over time and not everything can be uh you know not everything can be uh predictive in what's going to happen to the outcome.

So, I think that that experience and that knowhow is still going to matter, but it's going to be partnered with the AI agents. It's going to be putting those that human experience together with, you know, how you how do you um automate something. Um, so that's definitely going to be the future for sure. But that's what I I think craft and taste is still even when whenever you're doing anything on your own with with Gen AI, you still have to make choices.

you know, you still have to talk about, I want it to look like this director or this film or this actor or whatever it might be. You're still making taste and and choices and decisions even if you're using prompts. And if we would have to define craft and taste, for example, is it product market fit? Is it like what's in my brain like what has accumulated in the the past couple of years and also my prediction of where the market is going?

Is is that what it is? Uh yeah, I'd say that's that's part of it. And I think I think it's about um it's about design. It's about quality. It's about is this this might the the engine might spit out this as the answer, but does it really is it differentiating enough? Is this just a good idea that might it might have a market fit and it might be right for the brand, but is someone else already doing it?

Is it differentiated? So I think those all those factors go into when I talk talk about craft and taste. I think it's you know quality, it's choices, it's differentiation and that comes from experience. Um so that's where we sit right now. Will will that change as AI gets smarter and smarter? You know, perhaps right. And yeah, can't AI learn that and make even better prediction of uh of what is trendy and what is not trendy right now?

Like I I think it get pretty good at product market fit and what would be hip or not. It could have uh I think it would have to consult human brains, but I do think it can get pretty good at that. It's yeah that I don't think that's like a particularly human skill but I do see some skills for example uh like empathy right like that uh that seems uh pretty uh good that as a human being like just to understand EQ and and all of these things as a human I do think that this this is uh scalable I do think like trust as well as a

skill right is very human I think orchestrating these agents is also very good to your point That would be taste but also managing these systems and becoming technical. Do you think these are super skills of the future as well? Yeah, definitely. I and I think skill like you mentioned is the skill is the ability to do something what the work has to do which machines will be able to have the skill.

But then I think taste is knowing if it's good or not you know recognizing is it good? Does it make me feel something? Is it right for the brand? So, I think yes, I think skill is rapidly becoming um uh a level playing field with man and machine, but then taste is that that X factor, I think. Right. I don't know. What do you think? Yeah, I I think well brand is interesting, right?

Like brand is very abstract still. I don't think it's like a a defined science. I I very much think that yeah it I've heard it various times on this podcast like yeah like a top skill of taste and I would sort of relate that to intuition as as humans being as human beings I think our super skill is intuition which is computing data in or brain at a very low energy cost and incredibly rapidly and I think that is not going

to be replaceable anytime soon it's sort of like quantum computing what we're doing and I don't think we understand it well also from a neuroscience perspective what's happening what is intuition right uh then we call it like another synonym is gut feeling right and we have brain cells in our guts as well right so it's really weird like how this whole system works and I think we

can scale it you know like for example I'm a biohacker and I I do think that yeah I can increase these skills um by controlling some variables in that equation. I also think this is going to go up, right? We we're often talking about AI revolution, but we're we're not often talking about like the the human intelligence revolution, right? Which is like biohacking, Brian Johnson, right?

And all that. So I do think um this is going to be interesting and yeah possibly even neuro tech in the future like how to increase the output of our brains, how to sort of merge it with um AI and increase the output of both. I do think that uh that can compete with um AI and AGI itself. Wow, that's uh yeah, that's insane. How much Let me ask you because you do you use a lot of automated uh processes in your workflow.

Um has it how much how much would you say you use automation versus h human uh capability? Yeah. or uh AIS I would guess like automation VS AI because automation is just like sending some predefined variables through AI. Yeah. Yeah. And nowadays um agentic itself in my opinion it's automation but with AI so I'm doing a lot of agentic stuff follow-ups maybe I'm doing a bit too much you know for this podcast but like I just want to avoid the the no shows which were a problem like I probably had like 15% no shows and nowadays like 2%.

Oh that's great. Yeah I'm annoying people but uh I'm getting them on the pod and then once they speak to me they're not annoyed anymore. they they know that it was worth uh the follow-up. So, I mean, 90% of what I do is automated. I've always been obsessed with that. I used to have a an agency with more than a thousand employees a couple of years ago.

I let them all go and I replaced them with AIS. That was uh three to four years ago and it was the best move I ever made. Same company and you got rid of all the employees. Yeah. Yeah. top leads which is uh my GTM company and yeah I have an AI copywriter I have an AI researcher I have uh infrastructure that self adapts you know like so that my emails they actually reach uh the the inbox so yeah and back then it was very much painful right like uh the copywriters performance was unstable sometimes they would do well sometimes not their uh list pulling

capability was also terrible they because like targeting people especially in B2B is is pretty complex right there's so many roles there's so many nuances to these roles and different situation and AI is very very good at that so yeah the next steps I guess for this podcast would be to have an AI Charles to do the interview with like your AI Jason uh and that's coming in

like how do you think how do you think that podcast would would go the same thing would less human, but it would be very good because we could crank up the intensity or the like contrarianism and it would make for a very good conversation, right? And I think with us humans, it's always hard to start a podcast, right? Like and and break the ice. There's always like a slope we need to go through.

And I haven't found a solution for for that slope. Have you found one and for your podcast to make guests feel at ease and to quickly get in the the good juice? No, I think you have to just go through that awkward like first few minutes to get into it. Uh and build a rapport, but there's no way there's no way to not go through that that part. Like you have to do that.

Like there's no magic formula, you know? Yeah. And that's what being human's about, right? Like we're we're walking trust systems like we cannot get deep into people if if you're rational, right? Like uh if you're not naive, right? It it takes some time to just evaluate someone and adjust your own systems. Like humans are complex as well, right? Like it's very and we're always adapting, right?

Like we're adaptation creatures. We we are always sensing the person and adapting our style, how we speak, how we think when we're we're talking to someone else. We're very like social creatures, right? So, I don't think there's any magic solutions to these warm-ups. I mean, I've tried to do it, right? You've seen some of my automation. I sent you some of my favorite movies, some of my favorite music, but even that, you know, I'm like, "Yeah, how how the do I get over that slump?" Wow.

Well, that's amazing that you made that transformation. I uh that's not something I think about in my business. I think about um I certainly think about adding and and how much AI can do and how to upskill employees with it. But um you know, you made you made a huge jump and it seems like it's work. Obviously, you're behind it because it's working. Yeah.

Yeah. Wealthwise, it's uh it's pretty incredible because as you know like agencies uh yeah there's so much cost in employees that you pay like 80% of your cost most of the time and yeah there's like less churn because there's less errors. Uh there's also like more energy on my end because as a CEO I would spend probably 50% of my time on uh HR related stuff either recruiting new people or managing people right so yeah it it made a gigantic impact you know and about isolation in your work what about your day-to-day isolation yeah very good question the mental impact on you um it was nice back then

But I think it uh it had even a more negative impact on me. Why? Because I I would I would have to manage all my employees problems and emotions. And I used to hire a lot of Gen Z's, right? So I would always be the one like trying to fix their problem like Hero Charles, you know? And I think that sort of word on me. Um they were also like pretty stressed, right?

And they were putting these stressors on me. Um and yeah, that was less nice. I have this podcast. I'm talking to a lot of cos and founders. So, that's good for the a mental health part, but it's not like connecting physically with people. So, I'm still working on that aspect of my life. So, yeah, I have uh friends here that comes to this house. We go to the restaurant, have my wife, have five dogs.

The dogs are like a love machines, right? So, right. Love. I love that. Yeah. They they Yeah, they're they're my world basically. So I go with uh them like out every day and then I have my wife as well and my family. So I think that's conversation our our AI bots wouldn't talk about. I would I wouldn't know about your dogs. Yeah. Like I I think the AI know that about me.

I'd also use AI as a psychotherapy which is pretty good, right? like every time I have a a micro problem in my day, if like I don't know, I have an angry client or a email, right? I'm going to pass it by Chad GPT, my psycho therapy GPT and it's sort of gonna ease the micro pain, right? And I also AB tested that like with the physical psychotherapist and uh I think it's doing way better.

Why? Because it's always with me and it knows everything about me, you know. Amazing. Yeah. A lot of people uh are freaked out about that and that's really contrarian like Zuckerberg went on a podcast not so long ago and um he talked about like AI companions and all of that. I I wouldn't take it that far yet, but I think people that are usually against these ideas, they just haven't tested them, right?

Like they first I don't know, they're like too privacy oriented people, which is in my opinion. Think you should share like most of uh what you have if you're like in a free country. Um but yeah, once you you AB test it and see the solution, then you can compare it to the old solution and clearly see that it's better. That's incredible. I'm I want to try that.

Yeah. No, it's I'm gonna fire my therapist and try that. But I'll tell you like if you're using chat GPT for example, their voice API is is not like the the chat GPT text responses you would get. The voice API is like a simplified version of it. Or maybe I prompted it that way at least. But on my side, I don't think I prompted it that way. It's just like very short responses.

So, I wouldn't try to call um one. I would just try to chat with it. The chat one like gives out a long longer answer that's more prompted with you. And their voice is like is very short and it's it's pretty dumb actually. It's so I don't recommend that one. All right. Well, that's cool. I'll try that out. So, we should do uh we should do a followup and see how much of my business has become uh AI focused in six months.

Yeah. Yeah, we should do that. Maybe I should uh should be on your pod as well. We should do a followup at in six months or something like that. I'll get you on there. That'd be awesome. Where can people find out more about you, Jason? Uh you can just um you know uh I'm on uh my ma most people listen to the pod on YouTube. So YouTube soul and science with Jason Harris.

You can find me there. I'm Jason Harris on most of the socials. And uh yeah that's that's how you can find out about me. And then my my company is mechanism with a K like you said.com. It's always