Learning Library

← Back to Library

Sustainable Success in AI Consulting

Key Points

  • The AI consulting market is exploding, with revenues projected to hit $630 billion by 2028 and over half of large enterprises already seeking AI services, attracting many newcomers to the field.
  • Long‑term success as an AI consultant hinges on leveraging an existing consulting practice and client base—“winners keep winning”—because distribution and established relationships are the primary drivers of sustained business.
  • A common pitfall is the “AI‑wash” approach, where firms rebrand their existing services with AI buzzwords without genuinely mastering the technology, leading to superficial proposals and weak execution.
  • Lack of real AI expertise quickly betrays consultants, as even basic technical questions (e.g., when and how to use Retrieval‑Augmented Generation) expose their knowledge gaps and erode client trust.

Full Transcript

# Sustainable Success in AI Consulting **Source:** [https://www.youtube.com/watch?v=9-xvYoIMHcQ](https://www.youtube.com/watch?v=9-xvYoIMHcQ) **Duration:** 00:18:58 ## Summary - The AI consulting market is exploding, with revenues projected to hit $630 billion by 2028 and over half of large enterprises already seeking AI services, attracting many newcomers to the field. - Long‑term success as an AI consultant hinges on leveraging an existing consulting practice and client base—“winners keep winning”—because distribution and established relationships are the primary drivers of sustained business. - A common pitfall is the “AI‑wash” approach, where firms rebrand their existing services with AI buzzwords without genuinely mastering the technology, leading to superficial proposals and weak execution. - Lack of real AI expertise quickly betrays consultants, as even basic technical questions (e.g., when and how to use Retrieval‑Augmented Generation) expose their knowledge gaps and erode client trust. ## Sections - [00:00:00](https://www.youtube.com/watch?v=9-xvYoIMHcQ&t=0s) **Untitled Section** - - [00:04:09](https://www.youtube.com/watch?v=9-xvYoIMHcQ&t=249s) **Beyond AI Washing: Concrete Consulting** - The speaker criticizes vague AI buzzwords in consulting, urging firms to replace “AI‑powered” jargon with clear, domain‑specific solutions and intelligent delivery rather than competing solely on price. - [00:07:37](https://www.youtube.com/watch?v=9-xvYoIMHcQ&t=457s) **Specificity Over Generality in AI Consulting** - The speaker explains that embedding AI into a specialized domain—exemplified by Cursor’s development‑environment focus—yields deeper engagement, pricing power, and market advantage, urging consultants to offer a few narrowly defined AI services rather than a broad, generic offering. - [00:11:04](https://www.youtube.com/watch?v=9-xvYoIMHcQ&t=664s) **Niche Positioning to Avoid AI Frauds** - The speaker stresses that clear, specific content and a focused niche are essential for AI consultants to establish credibility and differentiate themselves from the flood of fraudsters and generic competitors. - [00:14:55](https://www.youtube.com/watch?v=9-xvYoIMHcQ&t=895s) **AI Enables Living Templates** - The speaker critiques overstated AI hype, stresses authenticity, and explains how AI transforms consulting templates into dynamic, customizable service offerings. - [00:18:34](https://www.youtube.com/watch?v=9-xvYoIMHcQ&t=1114s) **Seeking Traits of Ideal AI Consultants** - The speaker urges listeners to share positive behavioral traits of effective AI consultants, emphasizing the critical need to get these traits right to prevent damaging the AI ecosystem. ## Full Transcript
0:00We need to talk about what it takes to 0:02be an actual A+ AI consultant because I 0:05see a lot of people who are either 0:08considering that as a career path or 0:09actively pursuing it who aren't able to 0:12deliver the kind of extraordinary value 0:14that would let them retain clients over 0:16the long term. In this video, I'm going 0:19to talk about what I see as positive 0:22behaviors that actually promote 0:23long-term value for clients when you're 0:25consulting on AI and also the opposite 0:28things that I see people actually 0:29practicing a lot that aren't supportive 0:31of the value they're going to need to 0:33sustain to have a long-term consulting 0:36practice. But first, let's find out why 0:39everyone's running to the field. It's 0:40exploding. Revenues are projected to 0:42reach $630 billion in AI consulting 0:47alone by 2028. I don't know, that sounds 0:49like madeup money. It's a lot of money. 0:51Whether it's 500 billion or a trillion, 0:53who knows? It's a big pile of money. So, 0:55people are running at it. Over 50% of 0:57large enterprises are already using AI 1:00consulting services. And that doesn't 1:02even count the smallmedium businesses 1:04all over this country in the United 1:06States and also the rest of the world 1:08who are looking for support on AI, let 1:11alone individuals, sole proprietorships, 1:13indie hackers, etc. The vast majority of 1:16consulting firms, and I'm not even 1:18including AI in that. I'm not saying 1:21you'd like segment it down to just AI 1:22consulting firms. Consulting firms 1:24overall are reporting a bump in demand 1:26for AI expertise. Which gets us to our 1:30first point, which is that winners keep 1:34winning. In other words, the first rule 1:36of being an AI consultant is to be 1:38conveniently enough a consultant about 1:40something else with an existing book of 1:42business. Why? It goes back to one of 1:44the foundational laws in the AI era. 1:46Distribution is king. If you already 1:49have dist distribution relationships 1:52with other clients about other work, 1:55right? Maybe you're a marketing 1:56consulting agency. I don't know. Maybe 1:58you actually consult with folks on their 1:59development team or on team formation or 2:02on process improvement. There's a 2:04million in one things that consultants 2:06already do. And the convenient thing 2:07about a new general purpose technology 2:10is that AI can help with all of it, 2:12right? Like yay AI. And so what I find 2:15in practice, and this gets at our first 2:16pitfall, is that a lot of these 2:18consultancies that already exist tend to 2:21just AI wash what they're doing. They 2:23just paint the big sign, make it AI now, 2:25right? And then let's put some AI into 2:27our proposals, then we'll be fine, 2:29right? But the people delivering the 2:31service don't actually have A+ knowledge 2:34of AI. And so when they go and they talk 2:38to the client and the client just asks a 2:41perfectly reasonable question like when 2:43and where should I use RG? I hear a lot 2:44about RG. You can just sort of see the 2:46business style consultant kind of blanch 2:49visibly and they're like, I don't know, 2:50probably should talk to a technical 2:52person about that. That's not a super 2:54technical question, guys. As a business 2:56consultant, you should have some kind of 2:58highle opinion on when rag should be 3:01appropriate or not because you should 3:02understand what rag is for. And that's 3:04just one example. I'm not picking on 3:05rag. It's just an actual example I've 3:07seen. So, it was top of mind. In other 3:09words, when you are thinking about 3:11consultancies, one of the things that is 3:13a a a drawback is to use your existing 3:16distribution and AI wash. And yet, the 3:18strength is having that distribution and 3:20being in the consulting business. And so 3:22this should encourage you if you are not 3:24yet in a position where you have an 3:26existing book of business but you aspire 3:28to be a consultant or you perhaps wish 3:30to join a consultancy as an AI expert 3:33they need people like you and if you 3:34want to start your business it's never 3:36been a better time but you have to start 3:38it on the basis of stronger knowledge 3:41and this brings me to the second sort of 3:42bigger point the first point was talking 3:44about the state of the market what 3:45existing consultancies are doing now I 3:47want to talk about this idea that we 3:49might start from scratch and what does 3:51that take and how do you start from 3:52scratch or even how do existing 3:54consultancies retool? What are the key 3:56skills and uh key proof points they need 3:59to deploy to have a truly A+ 4:02relationship with their clients around 4:04AI? The first thing is to get specific. 4:06So much of AI washing, you can tell it's 4:09AI washing because it's vague in 4:10general. It's this sort of 4:12buzzwordfilled 4:13uh consultant language that makes sense 4:16maybe in the boardroom, but once you try 4:17and drill down to what it means for an 4:19actual individual contributor, you 4:21couldn't say what it meant. And so if 4:23you want to talk about the agentic mesh, 4:26right, what does that mean? No one 4:27really knows what the agentic mesh 4:29means. It's just a term that consultants 4:30made up. There are other terms that 4:32consultants tend to throw around like AI 4:34powered that also don't mean a whole 4:36lot. And so I like to suggest if you are 4:38trying to succeed long-term as a 4:42consultant, not just sell the book of 4:44business, right? Like you're if you just 4:45try and sell to the board, you can use 4:46terms like that. Most of the boards will 4:48buy it. But if you want to actually 4:49deliver the service, you have to get to 4:52the point where you can plausibly talk 4:56about what you will build in a way that 4:58is attractive and clear and specific. 5:00And when people are choosing between 5:02different proposals, you know, often 5:04times the assumption is, well, they 5:05choose on price and so we need to be 5:06competitive on price and we'll be fine. 5:09AI actually opens up more price 5:10headroom. You need to compete on 5:12intelligence delivery in a way that is 5:14plausible. And you need to show you can 5:16do it by showing a specific domain 5:19expertise that goes beyond AI. And I 5:22find that this is where a lot of 5:24consultants, whether they're established 5:26or whether they're new, go astray. They 5:28tend to say, "We're going to zero in on 5:30AI. All we do is AI." But AI is it's 5:32hard to put your fingers on. I've talked 5:34about that already in this video. It's a 5:36general purpose technology. You can do 5:37it with development. You can do it with 5:39process change. You can do it with a 5:40dozen other things in the consulting 5:42world, marketing, etc. You need to pick 5:45a particular domain that you can own. 5:49You need to pick a particular domain 5:51where if you have an opinion on how AI 5:53agents should be deployed in that 5:54domain, you you should be believed. You 5:57have authority there. You have a decade 5:59of experience. That is part of why when 6:01I speak about product and AI, I tend to 6:05have very strong opinions because I did 6:07product management for a long long time. 6:09When I speak about founding an AI, I 6:11tend to have opinions because I have 6:13found it before etc. I think the thing 6:15that I want to call out is don't be 6:17afraid of that background in yourself. 6:19You are not so much pivoting into AI as 6:22you are letting AI grow like like a 6:25strangler fig tree which grows all 6:27around the existing tree until 6:28eventually it like transforms it and or 6:31kills it. Let's just be honest. Uh but 6:34but you're growing a new AI product 6:37surface, a new AI service offering, a 6:39new AI way of thinking around what you 6:43have already done. And that current 6:45experience is critical. A strangler fig 6:47tree cannot grow without a tree to grow 6:51around. Similarly, you can't really sell 6:54your AI expertise without something that 6:56is a core domain expertise besides AI. 7:00The best consultant engagements that I 7:03have seen come from that core of 7:05expertise and then there's a deeply 7:08thought out AI enabling layer that helps 7:12that particular domain expertise go 7:14farther. And I'm not talking go a little 7:16bit farther. I'm talking significantly 7:18farther. This is part of why the 7:21software IDE market has exploded. IDE, 7:25as hard as it is to believe, development 7:28environments as a SAS business were not 7:30an attractive business at all for a long 7:32long long time until cursor came along. 7:35Cursor absolutely transformed them 7:37because it put AI at the heart of a 7:41previously boring business, but it owned 7:43the previously boring business. It is a 7:45development environment at core. I'm not 7:47here to make a video about cursor's 7:49product strategy, but I think that that 7:52picture in your head is useful as you 7:54think about the kind of deep engagement 7:57and the deep mesh that you need to have 7:59to borrow a term between the domain 8:02knowledge that you have and the 8:04expertise on AI that you need to deploy. 8:06You cannot be a general AI consultant. 8:10You need to be a specific one. Even if 8:12you have multiple expertise levels, 8:14multiple domains you can play in, it is 8:17stronger to have three or four specific 8:20offerings around AI that you're good at 8:23than to just make it one general 8:24offering. It also gives you more pricing 8:26power and more opportunity to earn the 8:29right to win because you are going to be 8:30able to say, you want to talk about SEO 8:33in the age of AI, I have a whole thing 8:34for that. I know SEO, I've known SEO for 8:36a decade. These are the 15 principles 8:39I'm laying out in the new world. this is 8:40how we're going to do SEO in the age of 8:42AI. This is how we're going to protect 8:44your search traffic. This is how we're 8:45going to get you sort of situated with 8:47large language models, etc. It's the 8:49specificity that helps you win. Another 8:51piece I want to call out, we've talked 8:52about distribution, we've talked about 8:54AI washing, we've talked about the 8:56importance of specificity, which I think 8:57is like if I had to pick one thing, I 8:59would call out you got to do that. I 9:01want to call out also it is really 9:04important in AI to be a part of the 9:06value chain. You need to understand 9:08where you can make friends and influence 9:10people inside the consulting world 9:13because people who consult tend to only 9:17have a piece of the puzzle if they're 9:18doing specifics. And so they need 9:20friends and partners. You need to be 9:22able to say, "My buddy James is going to 9:24come in and help you with this other 9:26piece. I'm maybe not the best in the 9:28world at that, but James is incredible 9:29and James can help you. We can come in 9:31on this engagement together." Have a lot 9:33of James's. have a lot of people you can 9:36say my buddy is really good at this and 9:38you know how you get that this is this 9:40is yet another point in in the theme of 9:42what makes a good consultant you put 9:44your work out there you have got to be 9:46able to talk very specifically about the 9:49kind of work you do and yes I know 9:51you're under NDA sometimes I get it you 9:53can still even if you can't do a 9:55specific case study talk very 9:58specifically about the kind of work you 10:00do in a way that lets other people 10:03remember you you can be their James. You 10:06can be the person they tap on the 10:07shoulder for. I guess we're following 10:09this example, the SEO example. Hey, I'm 10:11not an SEO guy, but my my buddy here is 10:13an SEO guy. Bring him in, right? You 10:15want to be that person, too. That is 10:17part of how you develop your pipeline. 10:19And the reason why that's especially 10:21important in the age of AI is that AI is 10:24that general purpose technology that 10:26means that organizations need AI 10:28everywhere they look. and they may want 10:31to write multiple engagements to help 10:33them get there. I've seen over and over 10:34again that an organization will pick a 10:36vertical and they'll get into the 10:38vertical and they'll say okay we got to 10:39expand from here and they'll want to 10:41write write a wider engagement and 10:43sometimes they want to write with the 10:45consultancy they know distribution but 10:47they write outside the expertise of that 10:49consultancy and so you need a friend at 10:51that point and you get friends by 10:54winning at developing and distributing 10:56your content so people know you and kind 10:58of can map you. It's not just a content 11:00game. It's also a social media game. 11:02It's also a game of like referring 11:04people where you meet people on 11:06engagements and then you keep up with 11:07them. It's all of those old tricks. It's 11:09just especially important now. And I 11:11think the content piece is especially 11:13important to signal credibility because 11:16I regret to tell you, as I'm sure you're 11:18aware, when there whenever there's a 11:20gigantic 0 to630 billion number, as 11:23we're having with AI consulting, it gets 11:25big. It gets big fast. It draws 11:27fraudsters. It draws huers. It draws 11:29people who sell snake oil in the AI 11:31sense. And so people, even people who 11:33are trying to hire consultants have 11:35their hackles up, let alone other 11:36consultants who are trying to size up 11:37and say, "Is this person someone I 11:39actually want to deliver with?" So the 11:40more you can be clear and specific in 11:43the way you talk about yourself, the 11:45less likely you are to be mistaken for a 11:47fraud. And people sometimes fight me on 11:49this if they're thinking about going 11:51into consulting because they say, "I 11:52want to be everything for everybody. I 11:53want to be general enough. I don't want 11:54to close the door on potential 11:56pipeline." I get it. I get that you want 11:58the business off the ground, but if 12:00you're too general with the content you 12:02put out there and the messaging you put 12:04out there, people can't tell the 12:06difference between you and 600,000 other 12:10AI consultants on the web, you just look 12:12like one of the crowd. Even if you have 12:14genuine expertise, maybe your expertise 12:15in AI is AI for fitness centers. I don't 12:18know. Maybe it's AI for clinics. Uh 12:20maybe it's AI for grocery stores. 12:22Whatever it is, own the domain expertise 12:24that you're talking about. lean into 12:26that area of focus. And what's 12:28interesting is that boldness enables you 12:32eventually to claim adjacent verticals. 12:34And so if you lean into your area of 12:36focus, let's say that you're super 12:38focused and you're really really good, 12:39we'll just pursue the SEO piece. You're 12:41really good at SEO now, like what 12:43happens to Google right now in the age 12:44of AI, what happens to your placements 12:46inside GPT. I've actually made videos 12:48about this. Maybe this is why this is 12:49top of mind. I also used to work in 12:51marketing at gray gray hairs and sat 12:53mini shares. So if you were that person, 12:56you can then extend to paid marketing 12:59eventually and say AI for paid 13:01marketing. You can extend AI for social 13:03marketing. If you stack up enough of 13:05those chips, you're eventually going to 13:06become an AI powered consultancy for 13:09marketing that is deeply plausible. But 13:11it tends to come more plausibly by 13:14stacking those individual verticals as 13:16opposed to just coming over the top. and 13:18existing consultants that have 13:20distribution relationships framed around 13:22marketing are going to try and AI wash 13:24their way over the top and almost always 13:26what I see when I look under the covers 13:28of those decks is they don't know what 13:31they're talking about maybe in one or 13:33two subverticals they have one person 13:35who's good but like for the most part 13:37they do not know what they are talking 13:38about and that costs all of us trust it 13:42costs trust in AI as a whole there's a 13:44reason that AI is something that people 13:46are skeptical about I do not think it's 13:49hard to tell wise on polls. Polls are 13:51notoriously unresponsive to rationale 13:53like you can't you you can speculate 13:55about why people answer something but 13:56you don't know. I am speculating. It is 13:59not just a distrust in AI as a new 14:01technology we don't understand. It is 14:03also a distrust especially if you look 14:06at business polling and sort of business 14:08surveys. It is a distrust in AI because 14:12consultants have burned bridges with bad 14:14implementations. I know consultants who 14:17make their entire living cleaning up 14:20after other consultants messes. There's 14:22going to be a lot more of that. I don't 14:24want there to be, but it burns the 14:26bridges. And so one of the negative 14:29consequences of AI washing of what we 14:31talked about at the top of this video is 14:33that if you do that, you don't only 14:35poison enterprise value for your own 14:37business. You poison value for the 14:39entire ecosystem because you teach 14:41people that AI is for frauds. AI is for 14:45hypocrites. AI is for people who will 14:48tell you they'll do the AI thing and 14:49then you look under the covers and it's 14:51not really it's not really AI, right? 14:53Like this reminds me of Amazon when 14:55Amazon would boldly claim they used AI 14:57and special fancy scales and cameras at 15:00their just walk out grocery stores 15:02actually used those grocery stores back 15:04in the day and it turned out they never 15:06got it to work. It was never really AI. 15:09It was always a bunch of people sitting 15:11in India looking at videos and that 15:13happens a lot and I think people are 15:15sensitive to that in AI and I think 15:17there should be and it's up to us if we 15:20are working with clients to have a 15:22different perspective and show that you 15:24can be authentic and be yourself. The 15:25last thing that I want to call out is 15:27templatization and this is something 15:29that you know we have gone after for a 15:31long time. If if you've ever done 15:32consulting people talk about 15:33templatizing your services all the time. 15:35The difference with AI is that AI 15:37enables you to do much much more 15:39interesting things with templates. 15:41Prompts can be living templates. 15:43Templates can be extended through 15:45prompting at the end of the end of the 15:47engagement surface with the client. 15:49Templates themselves become more 15:51customized and distinct service 15:52offerings because AI makes it so easy to 15:55extend tokens to the value chain that 15:58the customer is bringing. This is 16:00reflective of the fact that software 16:01itself is changing. software can be more 16:03customized now because it is so cheap to 16:05write code. Similarly, consulting can be 16:08more focused on individual customers 16:10than was practical before and still 16:12utilize the power of templates because 16:14it's cheaper to customize those 16:16templates. Now, that doesn't mean that 16:19you should forego templates. It doesn't 16:21mean you should give up on templates. 16:23Templates are still really important 16:24because they enable you to say 16:26distinctly what you intend to serve and 16:28how you intend to provide value in a way 16:31that is packaged up that a customer can 16:33understand. Customers are already 16:35struggling with AI. They are wrestling 16:38with AI. They need AI to be clearer. 16:42Packaging is part of how you do that. 16:44And the way you power packaging is 16:45through templates. But it's not just 16:48your your grandfather's templates from 16:50the old consultancy days in the 70s. 16:52These are templates that you can extend 16:54with AI. You can do that through 16:56prompting. You can do that by working to 16:59customize a particular offering after 17:01you listen to the customer and pull the 17:02transcript. There's a dozen ways to do 17:04it, but the point is the template still 17:06matters and customizing matters more 17:08than ever. So wrapping all of this up, 17:10if you were to look at the picture of an 17:12AI A+ powered consultant, they're 17:16specific. They know their domain. 17:17They're not afraid to use it and deeply 17:19wrap it into AI. Number two, they have 17:22distribution. They have figured out how 17:24to build relationships in the book of 17:25business. And this sounds 17:26self-fulfilling, but I don't think it 17:29is. I think that distribution is one of 17:31those things that consultants are able 17:33to achieve when they are able to 17:35demonstrate disproportionate value even 17:37to one customer to start and they 17:39generate word of mouth and referral. So 17:41distribution does matter as a marker of 17:43authenticity and a marker of good work 17:46especially if you can maintain and grow 17:47that distribution in the age of AI. they 17:49are clearly differentiated clearly 17:51differentiated from other work and 17:52having your own domain helps with that 17:54but so does explaining how you engage 17:56differently otherwise you're just going 17:58to wash into the background of the 18:00hundreds of thousands of other AI 18:02consultants they define the part of the 18:04value chain that they go after and they 18:06have partners they can bring in for 18:07other parts of the value chain whom they 18:09trust whom they have working 18:11relationships with they put their stuff 18:12out there which is part of how they 18:14build that relationship and you can see 18:15it specific and actionable without 18:17breaking NDA They templatize. They 18:20templatize. They templatize so that when 18:22you want to talk about what they offer, 18:24they have packages and templates that 18:25help you easily understand where the 18:27edges of their engagements are and where 18:29they're not. So, there you go. Those are 18:31my tips for being a strong consultant. 18:32I'm sure there are others. If you have 18:34examples of positive behavioral traits 18:37that make great AI consultants, I'd love 18:39to hear them. Let me know. Throw them in 18:41the comments. It's really critical to 18:43get this right. I think that the thing 18:44that I want to come back to is that if 18:46we get AI consultants wrong, we end up 18:48risking poisoning the AI ecosystem and 18:50nobody wants that. So take the time, 18:53bring your real agame, dive in on 18:55expertise, and have