In this episode of The Digital Edge, host Mark Reed-Edwards sits down with Phil Fersht, founder and CEO of HFS Research, to dissect the critical operational and financial realities of the enterprise AI boom. Phil pulls no punches as he warns CMOs about the skyrocketing economic constraints of automation, highlighting how unmanaged token costs can explode into millions of dollars a month.
The conversation explores how marketing leaders must navigate these massive financial shifts by moving beyond premium platforms toward more cost-effective model alternatives. Rather than letting cookie-cutter, passionless machine output drain corporate budgets and sanitize brand relevance, Phil outlines why the future belongs to clever, adaptable “chameleon” marketers who master cost-governance alongside creative taste. Learn how top enterprises can reject expensive, commoditized workflows, implement strict AI management, and successfully balance raw efficiency with authentic human judgment.
Read the Full Transcript Here
Mark Reed-Edwards: This is the Digital Edge from Incubeta. I’m Mark Reed-Edwards. This podcast is about how you can balance technology and humanity, how as AI eats the world, you can integrate efficiency with empathy. We’ll talk with leaders from Incubeta and across the industry as we traverse the digital edge into tomorrow’s world.
This week, I’m talking with HFS Research founder and CEO Phil Fersht. In our discussion, Phil urges CMOs to get trained on AI and avoid commoditized cookie-cutter output.
Let’s get to the discussion. Phil, welcome.
Phil Fersht: Hey, Mark, how are you?
Mark Reed-Edwards: I’m doing really well. It’s great to have you here. We’ve known each other, I was thinking this morning, we’ve known each other for more than two decades, so I know you pretty well. But can you tell the audience about your career and what HFS does?
Phil Fersht: Oh, that’s a big question. My career is about three decades old now. I started out as an analyst with IDC in the ’90s, in the UK. Got a move to the US during the dot-com boom, then moved to Singapore with IDC and Australia with them and ended up returning to the US. That’s where I met you with Yankee Group, one of the great old analyst companies, and then went into consulting for a while, went to Deloitte and went back to research, worked for a company called Everest Group, helped start up their outsourcing research and then eventually found my way back to research with a company that got acquired by Gartner, AMR, and then founded HFS in 2010 because I felt like at that point I had a specific idea on what a research firm should be like, how it should be run, and, the culture and that sort of thing.
And then I’ve never looked back. Honestly, it’s just grown and grown since then and today we’re about 70 staff. We punch way above our weight. We have quite sizable revenues and a very large analyst and advisor team spread across India, US, UK, Canada, Dubai.
It’s been a fairly exciting career and now I’m getting to the point of figuring out, I want to write a book and do some other things maybe at some point. But yeah, it’s been a good ride. Met a lot of great people along the way, and that’s really been the secret to my success: surrounding myself with talent and good people and learning from them and having them contribute and having them be part of this.
And a lot of my clients today are my former employees and friends and stuff like that. So it’s been a really good ride.
Mark Reed-Edwards: It’s all about the network, isn’t it?
Phil Fersht: A lot of it’s about the network. A lot of it’s about the people and your reputation and these types of things.
Mark Reed-Edwards: One of the reasons I asked you on here, other than just to get a chance to talk with you, is to talk about AI. And you’ve done a lot of work around AI, going back before ChatGPT hit the scene. So now AI can generate competent content, code, and analysis at scale, but there’s a lot of slop in there, isn’t there?
And in the HFS services-as-software model, where does human judgment actually sit? And is it becoming more valuable or is it just more invisible? What do you think?
Phil Fersht: I think we’re all just figuring out how to use these tools more effectively. I’m starting to notice a lot of people who weren’t particularly strong writers, content people, possibly relying on it more than anyone because suddenly they have powers they didn’t have before.
I’m literally seeing people who I’ve never seen write anything, send me like twelve-page white papers, and I’m like, “Where did this come from?” And then you read it and think, “Oh God, this just sounds like ChatGPT,” a bunch of blah. Nice fancy title, and the rest is just lots of manufactured artificial stuff.
But on the plus side, if you’re good at what you do these can be very good tools to help you structure things better. I like to write something first now myself, and then I might go to Claude and get it to help me structure what I’ve written better or maybe give me some other ideas or actually go in to say, “Hey, can you look at the market and see if there’s any other data pools that I could have missed?”
That sort of thing. So I think as a kind of research assistant, it’s a fantastic piece of technology that’s only getting better. But I just think it depends who you are, how you use this. And I think there’s also a tremendous amount of fatigue. People read things, and they feel they’re just reading artificial stuff.
It’s just such a turnoff. I don’t want to read it. I have people commenting on LinkedIn posts I’ve put out, which is clearly written by their LLM, and I even called someone on it yesterday and the person denied it. Just said, “Oh, it’s just that’s how I write,” after saying, “Great framing and that lands well,” and blah.
Come on, man! And I actually think certain people are now putting everything through their LLM, everything. Every email, every response they write, every proposal to the point of everything is just a load of em dashes and canned, passionless verbiage. It’s, I think it’s a problem.
I think people are going to suffer from some type of cognitive decline because they’re not thinking for themselves anymore. They’re just literally getting a machine to do their work for them. And the problem is, if that’s what you’re doing, you’re going to become pretty irrelevant yourself pretty quickly.
Mark Reed-Edwards: Yeah, what about human judgment? There are lots of people who use AI every day, but they don’t use it to replace themselves. They use it as an augmentation. So where does human judgment sit in that?
Phil Fersht: Human judgment, I think, becomes very important because you’re now having to trust the technology you’re using and the information you’re getting to be accurate, to be reliable, to make decisions on. And you need to be accountable. So you’re relying on these tools. Ultimately, it’s the human at the helm who’s going to be fired if you get it wrong.
So I do think human judgment, if anything, is getting more important because a lot of the grunt work’s being done for us now, and now we need to be better at judging and deliberating and thinking and being smart. And then the other thing we need to get better at is communicating and relationship building.
One thing I would say is people are out at conferences more than ever. People want to interact more than ever. People are feeling, I think, a little anxious as well as excited. And it is driving, in some respects, a lot of good behavior as well as, I think, some rather concerning behavior.
Mark Reed-Edwards: So you’ve talked a lot over the years about autonomous AI-driven enterprises, but if, you know, every marketer uses similar models, where does the real differentiation come from? When you look at the companies that you cover, the companies that you talk to, the CEOs that you talk to, or the CMOs that you talk to, you know, where are they gaining differentiation?
Phil Fersht: The more they commoditize, the harder it is to find differentiation, and then you end up with maybe a smaller number of models. But I think what’s going to happen next is companies right now are paying a premium for their Anthropic, their OpenAI, these tools. What’s going to happen next is we’re gonna see this flood of Chinese tools like DeepSeek and some others really impact the markets because CFOs don’t want to spend $10, $15, $20 million a year on tokens.
And the cost of this is starting to explode. And I think that’s going to have an impact on how we engage with this. And because I think already the ROI is becoming quite clear for certain businesses, certain people, these types of things. And now it’s going to shift to how well you manage this, how well you cost it up as well.
Because ultimately, I think I heard one company racked up five hundred million dollars of token costs one month because they weren’t managing this properly. I do think that we’ll continue to see better advancement with these tools for a couple more years in particular, because, look at the sophistication today even compared to a year ago.
You can literally run entire PowerPoint decks on Claude now, which I’ve been waiting 30 years for this. I remember Windows 95 and I’d never seen any advancement in PowerPoint until recently, till Anthropic’s come in and developed a nice plugin, and you can actually go in and create beautiful slides for yourself.
And we’re seeing it in Word, Excel, it’s fantastic, so it’s really changing how we work, how we operate, and it’s going to get better before it maybe hits a plateau. And that’s when I think it’s going to get down to: How can we do this a little faster, little less costly, burn a bit less energy?
Mark Reed-Edwards: Yeah. And a lot of AI strategy is framed around productivity. So thinking of a CMO, how should they look at the tension between producing work, the productivity side of it, but ensuring that work is meaningful and actually matters?
Phil Fersht: I think for a CMO, this impacts them more than most professionals in my opinion, because suddenly a lot of what you did can be done using these tools: campaign runs, email automations, managing social media, these types of things. A lot of this can get better automated. But I also see a lot more small to medium businesses springing up that only need one, two, or three people in marketing because you can do so much without these armies of staff.
I remember the days where you’d have somebody just managing your Twitter and somebody just fielding press calls. Suddenly now I think there’s less tolerance for just throwing more labor at issues and thinking a bit more about what are we trying to achieve here.
So I work for a company that competes with billion-dollar enterprises and I’m not a billion-dollar enterprise, but clients no longer care that much that I don’t have a thousand salespeople scattered around the world and things like that because we can compete in terms of data, insight, impact, and market focus.
But we were doing that before AI got as sophisticated as it is. All that we have now is it’s just the technology’s much quicker. I think there’s less tolerance for people who don’t understand it. And that’s the other thing is you need to be really good at using these tools, and I think you will get found out very quickly if you really aren’t using them.
And I would urge mid, late career CMOs to get themselves trained up, really understand what these tools can do. Because I’ve seen some people have got really good at this stuff and others who I think are still living in a world of denial, and they want to go hide away in a sort of corporate environment and pretend that what’s happening isn’t happening.
Mark Reed-Edwards: Are you seeing more high-value work coming out of people because of AI, more creativity? Or is it just more output, like you referenced, you know, the person you see posting a lot on LinkedIn who can’t really write?
Phil Fersht: It’s a really good question. I think that if you asked me this question about six months, nine months ago, I would say absolutely seeing really high quality information out there, and I think a good analyst would become a great analyst because they’re really, “Oh my God, I can now get what I want really quickly.”
What I think you need to do now is you want to read something that’s authentic. You want to hear insight that’s authentic. You want to read from a human. And I read far too much stuff from people who I know and respect, who I can tell they’re putting stuff through their tools, and it doesn’t sound like them anymore.
And that’s what worries me. Maybe it is them, but it doesn’t sound like them. And I think what AI has done is it’s sanitized insight and sanitized product to the point of what’s truly great versus what is just standard fare. So it’s probably raised the plateau in terms of content, but it’s destroyed the plateau in terms of personality and authenticity.
And I really want to see that start to come back into daily work life. Right now things are getting a little synthetic, a little artificial. Everything has to be automated. And you really want to get back to that world of more human interaction, more collaboration, more culture.
Even when I look at my analyst team, what keeps me up at night is worrying that they’re spending more time collaborating with Claude than each other, not a good thing, in my opinion. The real magic happens when you put humans together, and you really start to think about where the world is going versus we’re all just pumping out very productive content and delivery, and maybe losing that sort of authentic touch.
Mark Reed-Edwards: And a lot of that AI work is, you know, it’s solid, but it’s indistinct. So if you were talking to a CMO right now what would you tell them about, thinking about taste and creative judgment? Can that be taught or systematized or is it inherently a human thing?
Phil Fersht: I’m gonna say the problem with marketing in general is it has become a bit too standardized and cookie-cutter in general. And so I see a lot of marketing people loving these tools because they just want to put out standard fare. To them, as long as the words make sense and it’s well constructed and put together and looks professional, that’s what they’re thinking versus that human touch.
And maybe, you’re talking to somebody who built an analyst firm which had a personality and, I got vilified for it a lot when I was younger because I would dare to call people out and things like that. But I do think a lot of marketing in general has always been a little bit cardboard, especially in corporate situations where you just want to tick the box and make sure the brochure looks good. So this has actually made it even more kind of artificial versus marketing that is real personality-driven. Like when I look at successful companies today, you’ll often have a leader who has a voice in the industry and some good people around that leader who also have personalities and people want to engage with.
So when I think of, I’ll pick on a company, Accenture, right? So 15 years ago, I would probably be able to name you 50, 75 people I knew were engaged within Accenture. And IBM as well, they’re similar companies. They had great people they would send down in front of you who would educate you and get you excited about the value of dealing with that company.
Today, I would probably say I can name maybe 20 people or 15 people in each of those companies because everything’s got a bit more sanitized and a bit more mechanized, right? A bit more corporate. And I think what has to happen next is companies who want to be more successful need to get away from the corporate facade and need to think, “How do we get more personality back into what we do? What values do we hold as an organization? Are we investing in a junior talent which has particular cultural values that we want to promote?”
Mark Reed-Edwards: So big question to end. You know, If you look three to five years out, what do you think the real role of the human marketer in an AI-first enterprise will be? And what do the best people that do that, what kind of people are they?
Phil Fersht: I don’t think they have to be different people than the ones who are around today. I just think that clever marketing people are the ones who are good chameleons. They can adapt well to delighting customers, creating personalities for their businesses, creating personas around the leaders within the company, the ethics within the company.
And I just think something’s gone a bit lost, and I think AI has really helped exacerbate that in the last couple of years. So I think successful marketing people need to be creating more of that corporate personality, working with their people to help create the culture. So I think marketing needs to be internal as well as external within a company, like creating that culture, creating that raison d’être for a business than just that glitzy external view from outside.
So we need to get away from commoditized content and value, everybody sounding like everybody. I look at the tech firms and the service providers, and it’s almost insipid how they’re all sounding like each other. It’s let’s put out a puffy press release, a meaningless toothless partnership, and all say the same. That needs to change. And I think marketing is going to be more valued than ever as a differentiator in the next two to three years because without it, I think companies are going to get lost in this sea of bland, and they need to find a way to rise above that, be more authentic and, find a way for people to want to engage with them again.
Mark Reed-Edwards: Well, that’s a great way to end, Phil. I really appreciate you joining me on “The Digital Edge.”
Phil Fersht: Thank you, Mark. It’s great to catch up.
Mark Reed-Edwards: That was a great discussion with Phil. My takeaway: differentiation will come from authenticity, creative taste, and rebuilding corporate personality and culture. Let’s take that to heart. Thanks for being here today. I’m Mark Reed-Edwards. Join me on The Digital Edge next time.
Speakers: Host: Mark Reed-Edwards; Guest: Phil Fersht
Take A Deeper Dive Into The World Of The Marketer
We are entering an era where the rules of marketing are being quietly rewritten – not by tools, but by how decisions are made. In Incubeta’s latest research report “The Marketer’s Confidence Paradox” – conducted among marketing leaders in Retail and eCommerce – it’s revealed that while 70.4% of marketing leaders are confident their budgets are deployed effectively, 41.6% acknowledge a significant portion of that investment fails to deliver its full value. Incubeta’s research explores the gap between perceived performance and real growth, providing a roadmap to bridge the divide.
