Transcription
We need to talk about what it takes to be an actual A+ AI consultant because I see a lot of people who are either considering that as a career path or actively pursuing it who aren't able to deliver the kind of extraordinary value that would let them retain clients over the long term. In this video, I'm going to talk about what I see as positive behaviors that actually promote long-term value for clients when you're consulting on AI and also the opposite things that I see people actually practicing a lot that aren't supportive of the value they're going to need to sustain to have a long-term consulting practice.
But first, let's find out why everyone's running to the field. It's exploding. Revenues are projected to reach $630 billion in AI consulting alone by 2028. I don't know, that sounds like madeup money. It's a lot of money. Whether it's 500 billion or a trillion, who knows? It's a big pile of money. So, people are running at it. Over 50% of large enterprises are already using AI consulting services. And that doesn't even count the small-medium businesses all over this country in the United States and also the rest of the world who are looking for support on AI, let alone individuals, sole proprietorships, indie hackers, etc. The vast majority of consulting firms, and I'm not even including AI in that. I'm not saying you'd like segment it down to just AI consulting firms. Consulting firms overall are reporting a bump in demand for AI expertise.
Which gets us to our first point, which is that winners keep winning. In other words, the first rule of being an AI consultant is to be conveniently enough a consultant about something else with an existing book of business. Why? It goes back to one of the foundational laws in the AI era. Distribution is king. If you already have distribution relationships with other clients about other work, right? Maybe you're a marketing consulting agency. I don't know. Maybe you actually consult with folks on their development team or on team formation or on process improvement. There's a million in one things that consultants already do. And the convenient thing about a new general purpose technology is that AI can help with all of it, right? Like yay AI.
And so what I find in practice, and this gets at our first pitfall, is that a lot of these consultancies that already exist tend to just AI wash what they're doing. They just paint the big sign, make it AI now, right? And then let's put some AI into our proposals, then we'll be fine, right? But the people delivering the service don't actually have A+ knowledge of AI. And so when they go and they talk to the client and the client just asks a perfectly reasonable question like when and where should I use RAG? I hear a lot about RAG. You can just sort of see the business style consultant kind of blanch visibly and they're like, I don't know, probably should talk to a technical person about that. That's not a super technical question, guys. As a business consultant, you should have some kind of high-level opinion on when RAG should be appropriate or not because you should understand what RAG is for. And that's just one example. I'm not picking on RAG. It's just an actual example I've seen. So, it was top of mind.
In other words, when you are thinking about consultancies, one of the things that is a drawback is to use your existing distribution and AI wash. And yet, the strength is having that distribution and being in the consulting business. And so this should encourage you if you are not yet in a position where you have an existing book of business but you aspire to be a consultant or you perhaps wish to join a consultancy as an AI expert. They need people like you and if you want to start your business it's never been a better time but you have to start it on the basis of stronger knowledge.
And this brings me to the second sort of bigger point. The first point was talking about the state of the market, what existing consultancies are doing now. I want to talk about this idea that we might start from scratch and what does that take and how do you start from scratch or even how do existing consultancies retool? What are the key skills and key proof points they need to deploy to have a truly A+ relationship with their clients around AI?
The first thing is to get specific. So much of AI washing, you can tell it's AI washing because it's vague in general. It's this sort of buzzword-filled consultant language that makes sense maybe in the boardroom, but once you try and drill down to what it means for an actual individual contributor, you couldn't say what it meant. And so if you want to talk about the agentic mesh, right? What does that mean? No one really knows what the agentic mesh means. It's just a term that consultants made up. There are other terms that consultants tend to throw around like AI-powered that also don't mean a whole lot. And so I like to suggest if you are trying to succeed long-term as a consultant, not just sell the book of business, right? Like if you just try and sell to the board, you can use terms like that. Most of the boards will buy it. But if you want to actually deliver the service, you have to get to the point where you can plausibly talk about what you will build in a way that is attractive and clear and specific.
And when people are choosing between different proposals, you know, often times the assumption is, well, they choose on price and so we need to be competitive on price and we'll be fine. AI actually opens up more price headroom. You need to compete on intelligence delivery in a way that is plausible. And you need to show you can do it by showing a specific domain expertise that goes beyond AI. And I find that this is where a lot of consultants, whether they're established or whether they're new, go astray. They tend to say, "We're going to zero in on AI. All we do is AI." But AI is it's hard to put your fingers on. I've talked about that already in this video. It's a general purpose technology. You can do it with development. You can do it with process change. You can do it with a dozen other things in the consulting world, marketing, etc. You need to pick a particular domain that you can own. You need to pick a particular domain where if you have an opinion on how AI agents should be deployed in that domain, you should be believed. You have authority there. You have a decade of experience. That is part of why when I speak about product and AI, I tend to have very strong opinions because I did product management for a long, long time. When I speak about founding an AI, I tend to have opinions because I have founded it before, etc.
I think the thing that I want to call out is don't be afraid of that background in yourself. You are not so much pivoting into AI as you are letting AI grow like a strangler fig tree, which grows all around the existing tree until eventually it transforms it or kills it. Let's just be honest. But you're growing a new AI product surface, a new AI service offering, a new AI way of thinking around what you have already done. And that current experience is critical. A strangler fig tree cannot grow without a tree to grow around. Similarly, you can't really sell your AI expertise without something that is a core domain expertise besides AI. The best consultant engagements that I have seen come from that core of expertise and then there's a deeply thought-out AI enabling layer that helps that particular domain expertise go farther. And I'm not talking go a little bit farther. I'm talking significantly farther.
This is part of why the software IDE market has exploded. IDE, as hard as it is to believe, development environments as a SaaS business were not an attractive business at all for a long, long, long time until Cursor came along. Cursor absolutely transformed them because it put AI at the heart of a previously boring business, but it owned the previously boring business. It is a development environment at its core. I'm not here to make a video about Cursor's product strategy, but I think that that picture in your head is useful as you think about the kind of deep engagement and the deep mesh that you need to have, to borrow a term, between the domain knowledge that you have and the expertise on AI that you need to deploy. You cannot be a general AI consultant. You need to be a specific one. Even if you have multiple expertise levels, multiple domains you can play in, it is stronger to have three or four specific offerings around AI that you're good at than to just make it one general offering. It also gives you more pricing power and more opportunity to earn the right to win because you are going to be able to say, "You want to talk about SEO in the age of AI? I have a whole thing for that. I know SEO, I've known SEO for a decade. These are the 15 principles I'm laying out in the new world. This is how we're going to do SEO in the age of AI. This is how we're going to protect your search traffic. This is how we're going to get you sort of situated with large language models," etc. It's the specificity that helps you win.
Another piece I want to call out. We've talked about distribution, we've talked about AI washing, we've talked about the importance of specificity, which I think is like if I had to pick one thing, I would call out you got to do that. I want to call out also it is really important in AI to be a part of the value chain. You need to understand where you can make friends and influence people inside the consulting world because people who consult tend to only have a piece of the puzzle if they're doing specifics. And so they need friends and partners. You need to be able to say, "My buddy James is going to come in and help you with this other piece. I'm maybe not the best in the world at that, but James is incredible and James can help you. We can come in on this engagement together." Have a lot of James's. Have a lot of people you can say, "My buddy is really good at this." And you know how you get that? This is yet another point in the theme of what makes a good consultant. You put your work out there. You have got to be able to talk very specifically about the kind of work you do. And yes, I know you're under NDA sometimes. I get it. You can still, even if you can't do a specific case study, talk very specifically about the kind of work you do in a way that lets other people remember you. You can be their James. You can be the person they tap on the shoulder for. I guess we're following this example, the SEO example. "Hey, I'm not an SEO guy, but my buddy here is an SEO guy. Bring him in, right?" You want to be that person, too. That is part of how you develop your pipeline.
And the reason why that's especially important in the age of AI is that AI is that general purpose technology that means that organizations need AI everywhere they look. And they may want to write multiple engagements to help them get there. I've seen over and over again that an organization will pick a vertical and they'll get into the vertical and they'll say, "Okay, we got to expand from here," and they'll want to write a wider engagement. And sometimes they want to write with the consultancy they know, distribution, but they write outside the expertise of that consultancy. And so you need a friend at that point. And you get friends by winning at developing and distributing your content so people know you and can kind of map you. It's not just a content game. It's also a social media game. It's also a game of referring people where you meet people on engagements and then you keep up with them. It's all of those old tricks. It's just especially important now.
And I think the content piece is especially important to signal credibility because I regret to tell you, as I'm sure you're aware, whenever there's a gigantic zero to $630 billion number, as we're having with AI consulting, it gets big. It gets big fast. It draws fraudsters. It draws hucksters. It draws people who sell snake oil in the AI sense. And so people, even people who are trying to hire consultants, have their hackles up, let alone other consultants who are trying to size up and say, "Is this person someone I actually want to deliver with?" So the more you can be clear and specific in the way you talk about yourself, the less likely you are to be mistaken for a fraud.
And people sometimes fight me on this if they're thinking about going into consulting because they say, "I want to be everything for everybody. I want to be general enough. I don't want to close the door on potential pipeline." I get it. I get that you want the business off the ground, but if you're too general with the content you put out there and the messaging you put out there, people can't tell the difference between you and 600,000 other AI consultants on the web. You just look like one of the crowd. Even if you have genuine expertise, maybe your expertise in AI is AI for fitness centers. I don't know. Maybe it's AI for clinics. Maybe it's AI for grocery stores. Whatever it is, own the domain expertise that you're talking about. Lean into that area of focus.
And what's interesting is that boldness enables you eventually to claim adjacent verticals. And so if you lean into your area of focus, let's say that you're super focused and you're really, really good, we'll just pursue the SEO piece. You're really good at SEO now. Like, what happens to Google right now in the age of AI? What happens to your placements inside GPT? I've actually made videos about this. Maybe this is why this is top of mind. I also used to work in marketing at Gray Gray Hairs and Sat Mini Shares. So if you were that person, you can then extend to paid marketing eventually and say, "AI for paid marketing." You can extend to "AI for social marketing." If you stack up enough of those chips, you're eventually going to become an AI-powered consultancy for marketing that is deeply plausible. But it tends to come more plausibly by stacking those individual verticals as opposed to just coming over the top. And existing consultants that have distribution relationships framed around marketing are going to try and AI wash their way over the top and almost always what I see when I look under the covers of those decks is they don't know what they're talking about. Maybe in one or two sub-verticals they have one person who's good, but like for the most part, they do not know what they are talking about. And that costs all of us trust. It costs trust in AI as a whole. There's a reason that AI is something that people are skeptical about. I do not think it's hard to tell wise on polls. Polls are notoriously unresponsive to rationale. Like, you can't. You can speculate about why people answer something, but you don't know. I am speculating. It is not just a distrust in AI as a new technology we don't understand. It is also a distrust, especially if you look at business polling and sort of business surveys. It is a distrust in AI because consultants have burned bridges with bad implementations. I know consultants who make their entire living cleaning up after other consultants' messes. There's going to be a lot more of that. I don't want there to be, but it burns the bridges.
And so one of the negative consequences of AI washing, of what we talked about at the top of this video, is that if you do that, you don't only poison enterprise value for your own business. You poison value for the entire ecosystem because you teach people that AI is for frauds. AI is for hypocrites. AI is for people who will tell you they'll do the AI thing and then you look under the covers and it's not really it's not really AI, right? Like this reminds me of Amazon when Amazon would boldly claim they used AI and special fancy scales and cameras at their Just Walk Out grocery stores. Actually used those grocery stores back in the day and it turned out they never got it to work. It was never really AI. It was always a bunch of people sitting in India looking at videos. And that happens a lot. And I think people are sensitive to that in AI and I think there should be. And it's up to us if we are working with clients to have a different perspective and show that you can be authentic and be yourself.
The last thing that I want to call out is templatization. And this is something that, you know, we have gone after for a long time. If you've ever done consulting, people talk about templatizing your services all the time. The difference with AI is that AI enables you to do much more interesting things with templates. Prompts can be living templates. Templates can be extended through prompting at the end of the engagement surface with the client. Templates themselves become more customized and distinct service offerings because AI makes it so easy to extend tokens to the value chain that the customer is bringing. This is reflective of the fact that software itself is changing. Software can be more customized now because it is so cheap to write code. Similarly, consulting can be more focused on individual customers than was practical before and still utilize the power of templates because it's cheaper to customize those templates.
Now, that doesn't mean that you should forego templates. It doesn't mean you should give up on templates. Templates are still really important because they enable you to say distinctly what you intend to serve and how you intend to provide value in a way that is packaged up that a customer can understand. Customers are already struggling with AI. They are wrestling with AI. They need AI to be clearer. Packaging is part of how you do that. And the way you power packaging is through templates. But it's not just your grandfather's templates from the old consultancy days in the '70s. These are templates that you can extend with AI. You can do that through prompting. You can do that by working to customize a particular offering after you listen to the customer and pull the transcript. There's a dozen ways to do it, but the point is the template still matters and customizing matters more than ever.
So wrapping all of this up, if you were to look at the picture of an AI A+ powered consultant, they're specific. They know their domain. They're not afraid to use it and deeply wrap it into AI. Number two, they have distribution. They have figured out how to build relationships in the book of business. And this sounds self-fulfilling, but I don't think it is. I think that distribution is one of those things that consultants are able to achieve when they are able to demonstrate disproportionate value even to one customer to start and they generate word of mouth and referral. So distribution does matter as a marker of authenticity and a marker of good work, especially if you can maintain and grow that distribution in the age of AI. They are clearly differentiated, clearly differentiated from other work. And having your own domain helps with that, but so does explaining how you engage differently. Otherwise, you're just going to wash into the background of the hundreds of thousands of other AI consultants. They define the part of the value chain that they go after and they have partners they can bring in for other parts of the value chain whom they trust, whom they have working relationships with. They put their stuff out there, which is part of how they build that relationship and you can see it specific and actionable without breaking NDA. They templatize. They templatize. They templatize so that when you want to talk about what they offer, they have packages and templates that help you easily understand where the edges of their engagements are and where they're not.
So, there you go. Those are my tips for being a strong consultant. I'm sure there are others. If you have examples of positive behavioral traits that make great AI consultants, I'd love to hear them. Let me know. Throw them in the comments. It's really critical to get this right. I think that the thing that I want to come back to is that if we get AI consultants wrong, we end up risking poisoning the AI ecosystem and nobody wants that. So take the time, bring your real A-game, dive in on expertise, and have.