AI Pretenders, platforms and scaling extraordinary talent: Phil Fersht with Malcom Frank

Hey, great to see you, man. How are things?

I know. I'm glad we don't have to sign checks anymore because I'd always get the date wrong. So, at least there's that benefit.

Sure. Gosh, I do a lot of things and I've been at this game for 35 years in a lot of different companies, but Phil, if I was to distill the whole thing, I don't know many things in this world, but I know how to grow tech services companies really fast. So, I've been fortunate to work with a lot of market leaders, but in each case, we would go into markets that were growing 4%, 5% and we were able to grow those companies 20, 30, 80%. So that's what I know well, just how to get into these firms, get service market fit, and help them grow really, really quickly.

Prior to that, I was one of the founders at this firm, Cambridge Technology Partners. So that was a darling of the '90s. Went through hyper-growth. We started on MIT's campus and that was at the client-server wave and grew it from pure startup to 67,000 people and that was actually the best IPO in the '90s not just of our industry of any industry. That firm was important for lots of reasons but one is it was the first tech services model like that that went public. And I could remember speaking to bankers and going on road show and all the rest and people would laugh at our faces. You say, 'Hold on, you know, the assets walk out the door every night and your visibility is only 180 days and why should you float this thing again?' And so I think about that a lot when we're all looking at what's happening with an EPAM or an Infosys or a Wipro and some of the market leaders of today.

I'm very, very bullish on the services space in the medium and long term. But there's a logic to what Wall Street is going through that they recognize the current model is broken. And so when you look at these valuations that have crashed, it's just very, very simple. As AI washes over this market and washes over the business model that currently exists in IT services, they don't trust the 5-year cash flows. It's that simple. If there's something that could be delivered for 10 units today and it's going to be delivered for three units three years from now, what is the associated value with that? Is there going to be enough price elasticity that when that unit cost collapses that the market opportunity is going to commensurately grow? So the street can't figure that out yet. This is why I'm so excited about this market transition because this is where fortunes are made. This is where new leaders emerge. But that new model in the services space has yet to break out. These are exciting times. But to answer your question, the street is looking at these very large pyramid-based organizations and are saying, I don't know what the future's going to hold, but I don't think that model's going to win.

That's exactly right. That's exactly right. And so, you know, you go with first principles thinking on the pyramid business model. You know, getting a pyramid of humans, you know, assigning a rate card to each of them. It's going to be this much time so forth and so on. That pyramid is very, very quickly turning into an obelisk number one. But then when you look across each service line and across the entire SDLC, it is getting eaten by AI. If there's anything I've learned, Phil, that one of the hardest things in the world to change, aside from yourself, is a publicly traded tech services company. They're extraordinarily hard to change. And so who is going to be able to navigate these seas and go through that fundamental change where at a unit level how you deliver? What is the split between the human and the bot? What are the types of people? How do you build the structure? How do you price that? These are a lot of questions that nobody has properly answered at scale. I think there's a few green shoots, but that is the $64,000 question in the market. I think that's why HFS is so vitally important during these times, but nobody has quite figured it out yet.

Well, that's right. You said a couple of really important things there. The first is what do they say about pioneers? They're the ones with arrows in their backs. So, the R&D model instead of it being internal, I really advise folks during these times. It's the old rip off and then deploy. You should be a student of the industry watching every firm, watching every move, every new capability and then say which ones are working and how do I become a very fast follower. And these markets evolve. It's interesting in a way they evolve slower than you think. They're sort of slow, slow and then all of a sudden. And we've seen this before in technology markets when you think of the internet browser from Netscape through AOL through IE and then it was Chrome. You know, that was a multi-year journey about half a decade to get through that. The same as with search. I can remember when Google came out and I was like, oh hell, the last thing we need is another search engine. You know, there's AltaVista and Excite and Yahoo and the rest. I didn't understand Boolean at the time. But the point is that that was a late entrant if you will. The same is true. You think of all of the mobile devices that you had from I started as a Psion guy. So you went from Psion to Palm Pilot to Blackberry and of course iPhone. So these things take more time than you think, but you have to lay the foundations today so that you can really scale. So that's thought number one. Thought number two, I was thinking over the holidays, what did 2025 really represent and how are we going to view that in our space? And Phil, I thought of it as it's sort of the parade of pretenders. That's what we had. And Carlotta Perez, I don't know if you follow her work, but she's this Venezuelan economist. That sounds like an oxymoron, but she now lives in the UK, but absolutely brilliant. And she's spent her entire career on mapping these large-scale economic transitions that were created by a new machine. And she said, 'The first few years, you get all of these speculators come in, all of the pretenders, all the charlatans.' And you know who they are. Think in your personal life of latest thing guy. You know, when we had the credit crisis, all of a sudden this person was an expert on CDOs and the mortgage market and then COVID comes along and they know mRNA and intubation and then, you know, Ukraine breaks out and suddenly they understand geopolitics and they're correcting you on how to pronounce the capital of Ukraine. So that same person has become, you know, self-appointed expert on RAG models and the rest. It creates a lot of confusion, but at the same time, I think that's a positive indicator of just how large this is. So, if you're running one of these firms, how do you make sense of this very confusing environment? And I think you're right. It's this sense and respond. Sense and respond, recognize what's working, pull it in quickly, and then start to scale against that.

If you keep that aperture that you've seen just in the last six months, Anthropic and Google have really started to break out in the enterprise because they understand the needs of those buyers and that there's a heavier lift and there's considerations that they all have to deal with that consumer markets just obviously don't have to. And I often think of one of my greatest mentors, it was Bruce Rogow. So, you know, Bruce was, you know, he built the magic quadrant. So, if people hate that thing, then you can blame Bruce. But he was at Gartner for ages and then he became basically in his later years a CIO concierge and one of the smartest people I know in the industry. But Bruce would always talk about this, Phil, that it doesn't matter what gets created on the supply side. What matters is how that can be ingested on the demand side and how you can work with a large organization that has their infrastructure, their systems, their data, their process, their people, their culture, their security needs and how do you navigate through all of that to make it easy for them to consume what you offer. And so when you look at firms that win in B2B, often it's not quote unquote the best technology, but it's those organizations that can work with large enterprises at their pace and deal with all of those issues. And so I think that you're starting to see both Anthropic, which is interesting as a young firm, and Google have started to do that very, very well from a model basis within the enterprise.