Family Feud · 3:40 PM · Wednesday, May 13, 2026
All right. So this is the family feud. We have the buyers and the providers sitting in the opposite direction. But this is the team buyers. I'll get you to introduce very quickly. And this is the team providers. Everybody knows me and Phil. So maybe if we can start with the provider team, if you guys can quickly introduce yourselves and then we'll get going.
All right. Hey, everybody. Atul Vashistha, I'm the founder and CEO of Aokah, a platform to manage the entire lifecycle of a global capability, explore, build, and optimize.
Hi, I'm Mike Hobday, senior advisor at HFS, formerly CEO of AntWorks and a long-time IBMer.
I'm Amresh, I manage the digital transformation at Citi. My role is to build trust and then increase the sales.
That's an easy role.
Yeah. What's easier? Good afternoon, everyone. Joe Montrosse, senior advisor with HFS research, and I've spent a lot of time advising organizations on how to, well, make these deals happen.
Yeah. All right. Fantastic. So let's get going. So this is how it will work. We have two teams. We have team buyers. We have team providers. There will be three rounds of voting in this. There will be a blame game, who's blaming each other? There will be a relationship counselor game of counseling each other and then who's winning? You know where's the power going? For each question you'll get 20 seconds or 30 seconds to discuss what's the best option. You'll get the same time and the audience will poll. The poll results will not be shown to either of you. They'll be hidden from you. And depending on how your answers match the audience reactions, you'll get each other a point. Audience role is very, very important. At the end, whoever has more points gets to win and sponsors all the drinks outside. After that. And by the way, it's not a democracy. Phil and I are really bad at math, so we don't keep a good score, and we can choose whoever we want to declare the winner. All right? All clear? Everybody's got their apps, you know? Phil, you want to start?
Sure. So we'll start with the blame game. So pilots never grow up into production, so who kills them? Is it buyers, they scope them too narrowly, never give them a chance? Providers, they over-promise in the sales cycle and under-deliver in execution? Is it both of them? They shake hands on a vision and then fight about the SOW. Is it procurement? They turn a transformation partner into a line item vendor? Or is it leadership? They want pioneer outcomes on exploratory budgets. Honestly, the steering committee did it.
You've got 20 seconds. Don't show the results of the poll yet. Otherwise, we'll cheat in the game. First one is always tricky. Let me know if you're ready. Why am I on? Oh, here's.
We just found out.
What do you think is the answer?
What do you think?
I tried to make it really complicated with giving options that you know you would have to discuss.
I think it's a combination of C and a little bit of D. Combination of C and D.
I think it's D. D? All right. All right. Time's up. We have a buzzer also. Wow. All right. Providers? What's your answer?
Well, I'll blame the buyers because they don't scope projects properly. They don't give them time enough to succeed, and they don't understand that they need to train AI models. Why did I even ask?
They had to blame the buyers. Well, we were going to go with the lawyers, looking at you, Akiba. But, no, what do you think?
No, we'll also go with D, is the procurement. They turned a transformation partner into a line item vendor.
So we had a chief procurement officer after the last panel he left. Good he's not here. So you have A, yours is D. Let's see what the audience pulled. Let's see if we can make it work. Unanimous. There it is. Oh, my God. That's a nice idea. I think it's leadership. It's leadership. It's leadership. All of you were wrong. Yeah. Even Phil was wrong.
Oh, no. I think both.
At least we didn't blame. Oh, thank you very much. Sorry?
At least we didn't blame the sellers.
So nobody gets a point on this one. Yeah. It's basically. That's fine. Nobody was good enough. All right. No matter. Who is more guilty of AI theater? Is it buyers? We green light pilot after pilot to show board level progress. Is it the providers? They rebrand existing services as agentic. You know, every provider is agentic first today. Both. We co-produce the tech and never read the footnotes. D is a very bad choice, but... The analyst... Maybe Phil takes the hit on that one.
And the reason was...
Let's see. I think the analysts are very, very right. I think D is definitely not the answer. Do you think the analyst community is guilty of AI theater?
I don't think they're as... Oh, that's a wrong buzzer. I don't see enough from the analyst community on it, to be honest with you. I see more noise from the tech. Already.
But we don't need to show the answers. All right.
Can't read them anyway. I'll jump in on this. So look, I think the answer in this case is C. The most common that I see today is buyers eager to experiment with AI because somebody is wanting them to show progress. Literally, you're often being measured by what you're doing rather than what you're getting to accomplish. Suppliers agree to the deal very easily in many cases without fully understanding what it takes to actually get to the outcome. So in one case, outcome's not clear. In other case, you'll make a commitment without understanding the outcome.
And no wonder things don't work out. So, Atul, your answer is perfectly right, but you don't know how to game the system. The answer was right there in front of you.
Can I give an alternative answer? Yes. AI theater, Anthropic. Fair marketing. Every time I go on X or threads or Instagram, it's like, Claude is about to take over something else. And everyone just has this feeling of dread. And then they make up their like, messengers from God. So come on. And then they want to IPO.
What was your answer? We clearly went C because it obviously includes A and B. Yeah. That's basically all of the above.
Again, nobody gets the answer. The answer is B. The audience answer is B. Yeah. We are at zero, zero. I think we were.
We're keeping it like that. Solid.
Yeah, okay.
Phil, you want to do this? So who owns the debt problem? We know what the debts are. Process data, talent, and tech. So buyers, it's our mess. We knew it going in. Buyers, we kept asking for cost cuts and got exactly what we paid for. Providers, they've been managing our operations for years, never flagged it. Just carried on selling, doing the same old, same old. Providers, this is D, they built dependencies instead of capabilities and they called it a partnership. Or both, they optimized for contract renewal. We blocked modernization. What do you think?
I know the answer to this one.
We're going to make Saurabh buzz for this one. Yeah. I think they're both true.
All right. You can raise your hand early as well. Yeah.
Ready? Yeah. We'll go with A, the buyers. We're the guilty party here. I think that means you got it right. Yeah. Because we do believe that they know exactly what the processes are. They know the internal data. They know how siloed it is. And they already know what they are in. So they kind of use, they have the complete knowledge. And then going into this debates, they're trying to see which of these they can lower and make it more manageable. So one point for the humility.
We're in, I think it's B, which is kind of a bias. They're looking for cost cuts. And they're just driving solutions, knowing that the vendors have got to kind of manage it all so that as long as they get their cost cuts.
They're getting exactly what they ask for. That's what they incentivize and that's what they're getting.
So basically it's the buyers. It's the question of what's the nuance on the buyers. So let's see what the audience said. See if we can get it right.
I was going to say that. I think I've seen so many contracts where both the buyer and the provider, teams, middle managers, completely comfortable with the status quo, just don't want to change it, renew it, renew it, renew it, and it's only when the pressure comes from above. Come on, where can we squeeze some more cost out of this?
So, Phil, what are we doing? We are at 0-0. After the end of this round, do you want to give one grace mark to any of the teams? No. We've already won this point. Penalty shootout. All right, we are at 0-0, but maybe we should give a point to the buyer team because they were humble enough to blame themselves.
Yeah, it is A actually. If you see the votes.
Oh, is it? It changed. Somebody voted. Oh, that's not fair.
You didn't see him on his phone? We did earn a point.
You did earn a point. Can you put your phone away?
Yeah.
All right. So we have buyer team one, provider team zero. We're like being on our —
Yeah. Yeah.
Relationship counseling. The trust is gone. Who broke it first? The buyers, we keep moving our goal posts and we call it agile. Buyers again, we treat providers like vendors and then wonder why they act like vendors. Providers, you guys put the A team in the sales pitch and an average team to actual deliver. Providers.
Again, all the 30 years.
The staff are most strategic work with the least experienced people, both it's a transactional relationship or let's just blame somebody else, legal and procurement.
Yeah, that's good, yeah, I'll get that.
So who broke the relationship?
We're ready to go. Yeah, we are. So look, I think, you know, when you look at deals where things are not working on long term and you're not driving the outcomes, it's very clear they run it, they, sorry, they contract it as a transactional relationship. They run it like a transactional relationship, and whenever they're in trouble, they treat it like a transactional relationship. Suppliers, even though they know the problem, are not being proactive about it, so we say E. Yeah, yeah.
E, both. Okay. Yeah, we have the same opinion as well. Both, we built a transactional relationship. We said it first. We said it first. Though he would like to blame the legal and the procurement. Yeah, who wouldn't? Who wouldn't want to blame others?
So, but we'll go with E. So, both are E. Let's see what the actual answer is. Yeah, it is. It is, so both teams get a point here.
Oh, thank goodness. Half a point extra for going first. Yeah, I think that's true. And that's gonna be unsustainable in an AI world where you're gonna share in data, so I think we should definitely get two points.
All right, Phil, your chance. Single biggest thing killing the partnership right now. So commercial models still priced for labor. Buyers measuring providers on cost while demanding transformation. C, neither side willing to share risk or real reward. D, the RFP song and dance or dog and pony. Six months, 400 questions, and we picked the prettiest deck. E, governance structures designed to catch failure, not enable spend. F, providers managing to the contract instead of managing to the outcome. Yeah. I don't think anyone ever pays attention.
Anybody ready? We're going with A because it is quite evident in the market that contracts are still based on headcount and per capita. And that's not, although there is a demand for outcome-based contracts, procurement are getting in the way.
It's what's driving the incentive. Yeah. So commercial models?
Yep. We'll go with F as well. You're going with F? We're going with F. Yeah, we're focused on being right, not with being popular. So why F? Joe? I mean, I think when I think of managing to the contract, I think of figuring out every way you can get a change order to get it to procurement, to get it through a system, when instead you should be trying to make sure stuff works.
You've been worn out by change orders. You have no idea. All right.
Don't reveal. Is it C? It looks like it's C. It looks like it's C, right? Either side willing. C? They're hedging every part.
I think the buyer is getting big too. Yeah. It's C I think.
Yeah, it's C as part of that.
Yeah. A and B are pretty close. Maybe we should give half a point.
It's getting desperate, isn't it?
F versus. Pretty point.
You know, they said commercial model is pretty strong.
It is. Well, that was the incentive point, right? Yeah.
So I think one and a half and two. All right. All right.
All right, we're closing the gap.
Who needs to change more to make the relationship work in the AI world? Buyers, stop treating AI transformation like a traditional outsourcing deal. You know, it's important. Buyers, give providers access to be innovative, not just, you know, you want them to be innovative but you don't give them access to your systems or anything. Providers, if your AI pitch still leads with cost take out, what's the point? Providers, can you actually retire the pyramid and go to the triangle that Phil was talking about? Or both? Really, it's the contract which is the problem.
We can go. Go for it. Yeah, we'll go with E, both. Which one? Contract is to blame. Yeah. And, yeah, contract is essentially to blame because, like you said, it is also about innovation. From a buyer perspective, they want to see the results as fast as possible and tangible. From a provider perspective, they want to retire the pyramid. And then so we would, I think in this case, both are to blame. One is overpromising, and the other one is expecting immediate results. That's why we feel both.
With nobody looking forward to signing up for the incentive structure that would at least help get you on track.
B? Yes, well, we were B with a shade of A. Yes. Right?
B was green and A was amber.
Yeah, so let me maybe talk about B and then you jump in on A. So look, I'm thinking about not just what you already have as relationships, but the fact that you're looking at new. And when you think about that, if you're not gonna integrate your provider or give them view into what you're doing, what your intent is, give them access to your systems, have them understand your business, how are they gonna be innovative? That's right. Right? Yeah, read A for me again, 'cause I can't. Stop treating me, oh sorry, stop treating AI transformation like a traditional outsourcing.
Yeah, so the issue here is, in these new relationships, the world we're moving into, there needs to be a much closer ecosystem. You can't continue with the vendor enterprise relationships we've had in the past. It's got to be much more closer ecosystem where the divide between the two parties becomes much more blurred, in my view.
But I would like to also reiterate that it's already working. The relationship is working in AI worlds. You already have all those use cases where companies have successfully deployed AI solutions. So I think it's a myth that doesn't work, but we've seen it work as well.
You're an exception to the norm.
Yeah, you can read about it. Let's reveal.
I would say B, but only if the vendor actually has something innovative to offer.
So the audience answer is E. The contract is the problem and nobody wants to be the one to rewrite it. So I think you guys. Right and popular. Right and popular. Do you know what the score is, Phil? I think they just took the lead. Yeah. Three to one and a half. Thank you. Trust the professor. Trust the professor. We have somebody else.
They're buying the drinks. I was at like 50.
All right. Round three.
Phil. Power shift. Who holds the leverage in this AI first world? Large providers. The world's flattening mid-tier providers who squeezed vendor consolidation. C, buyers. We never built internal AI capability. Now we're more dependent than ever. D, buyers. We outsource so much institutional knowledge we can't even supervise the agents. Or anyone still billing by the hour in this room?
I think this should have been a three-way family feud with the technology provider family as well. Or maybe...
It's being hard to represent. That's what I wanted to do. Sorry. So I think... Perfect. So now I get full time. So we think B because large providers at the scale to both invest but also those deep relationships. So it's harder to kind of remove them but they have the ability to invest. Mid-tier providers I think are the most challenged because they're not just being challenged by the large providers. They're being challenged by in-house capability and innovative providers like us.
It's the talent gap, isn't it? Because the mid-tier will find it very difficult to recruit. I think so.
But, Mike, all the mid-tiers are growing the fastest right now. You know, take Coforge and Persistent and, you know, everybody else is struggling with 1 to 5%. They're growing at, like, double digits.
This reminds me of when I did my first company in 2000 where Accenture and IBM were refusing to build offshore capabilities. It took them literally until mid 2005, 2006. It gave the ability of Cognizant and Infosys all this ability to rise. I feel that's what's happening to the mid-market right now, but I think they're going to be disrupted very heavily, but it might take another three, four years, three to five years.
I'll go with A, and the reason is right now, the infrastructure and the companies that provide the infrastructure, let's say the Geminis and the Googles of the world, are taking their advantage, but they have that leverage for the next probably two or three years, right? Once that kind of settles in, then they're the ones who are going to lose most of the leverage because everything by that time would have got kind of commoditized. And at the same time, you would have already had the data, the friction points, the efficiency, all that would have already shifted in favor of the buyers, right? So then the next three to seven years is probably going to be more of a buyer's play. So the biggest leverage that these infrastructure companies have now will be the ones who lose all the leverage. So we have A, large providers, mid-tier.
Phil, who do you think is going to lose? Is it large, mid-tiers, buyers, large? Okay. I think everybody loses HFS wins. What is the answer? A, large providers. This room thinks the large providers are going to lose. Sorry, guys.
We will find out in a couple of years. Then we'll come back and reset.
Maybe, four and one and a half?
Yes.
Okay. Who holds the leverage in renegotiating contracts? Okay. Buyers. Because the new economics of AI make every existing contract, you know, you've got a lot of cost saving. Tech providers, because they control the models, so they control the roadmaps. The service providers, because they're embedded in the operations, so they have the leverage. Buyers, again, because switching is much easier now. With AI first contracts, the exit costs are low. Tech providers, we used to sell to the IT. Now we sell directly to the CEO and skip everybody else. Or service providers because you guys don't have the talent. You don't have the talent. Or buyers don't have the talent. So who holds the leverage? Is it buyers, tech providers, or service providers?
Well, we were going for, we haven't done very well up to now. We decided to try and second guess what they're all thinking. Because we think that, or I think, I think we think that the buyers will think they've got the power. But we think the tech providers have got the platforms and the models. So you will be A? We're going to go with A because we think they'll go for A.
All right. Yeah, we're going to go with D. The more that clients are holding on to the licenses, data, some of the development around everything and everything in the cloud. It's just getting easier and easier.
Would you switch windows?
Yeah. Why not?
You already have the return on investment. You already have everything quantified. And the agentic AI is probably going to give them the answer. What is overpriced? What is underpriced? We negotiate at that time.
Yeah. And companies love to mix it up.
Phil, do you think buyers have all the negotiating power?
Absolutely. And it's going to get even more towards them. We've got a chance.
All right. Let's see. Let's see. So maybe they'll be over. Oh, my God. Yeah. You guys get a point?
Yeah. You see, I've been told you we've been overthinking it.
So it's two and a half and four.
We've only just woken up.
Yeah. Do you mind if I ask Phil a question? Sorry if you. So, if you think about the tech providers, it's kind of like the old, the spades and kind of the platform. Why do you think they will not have the power in the future? You know, because you build on their models, right? It's not that easy to switch models. Or do you think in the future it will be easy to switch models?
I think the technology is gonna get increasingly and to the point where you'll be struggling to differentiate between a Gemini or an Anthropic or an OpenAI, and you'll almost get blended models and things, and it's going to be more down to how's that control layer operating and who operates the control layer. Do you do it yourself? Do you use partners?
Can I have a go? I think the answer will lie in the knowledge graph because I believe in two years' time, we'll be saying that the knowledge, the learning in an organization sits in its knowledge graphs attached to its agents and not in the language model. And the answer to competitive advantage and your moat is that you train your knowledge graph and not your large language model. And then you can swap in and out a large language model anytime you want.
Yeah.
I don't know. Is that on the question? Do I get extra points for that?
No. Well, that's just extra time. All right, let's keep going. Who will own the AI first operating model of the future? This one is important. It's buyers because it's your data. So if whoever owns the data owns the game. It's tech providers because it's their platform. Whoever owns the platform owns the game. It's service providers because the service providers will own the workflows, you know, and whoever owns the business outcomes sort of owns the game. It could be buyers again because, you know, they can build all this internally and they don't need you guys anymore because the agentic is so easy. It's tech providers because of what Anthropic just announced. They'll build their own consulting companies. Or it's service providers because you sit between the tech and the business and that seat only gets more value. You're the one who will bring IT and business together.
What do you think? I have to go with what we think. I think it's right. Okay.
I think this reinforces what Phil just said, which is A, look, whoever owns the data owns the game. Right? Because if you can, you own that intelligence layer and you can switch LLMs, and we've seen that. I mean, like... Whenever you speak, there's a buzzer. Yeah. I should wait till it goes to 20 and then speak. Yeah. But it's one of the reasons where, like, look, I'll talk quickly about the company that I just founded a year ago. It has massive amounts of data, right? And we leverage that for exploration. So when you're thinking about a GCC, instead of going to an LLM, imagine if it's data that's constructed for that, and you leverage it to make decisions on location, talent pools, risk, infrastructure, and others. What we decided is starting the 1st of July, we're releasing it to HFS customers at no cost. It's going to be free. Is that like a gimmick or is that... No, this is true. This is true. I didn't know that. No, this is true. Again, not for service providers. Sorry. For the buy side. Right? And the reason we think is that we think the bigger problem with AI is actually the execution layer. Which is, well, you know, do you have the intelligence at the right moment to actually make those decisions? But do you actually execute well? And so in this whole aspect of this answer, when we think about data is, look, eventually LLM's program management, like what we're trying to do, enable you to leverage that decision, but the secret sauce is having the ability to execute it. So when drift occurs, you know it's happening ahead of time. So you can come back to what your original intent was, what your original outcome was. So you know during the journey whether you're getting there, not at the end of the journey.
And I'd add to that, if the buyer doesn't own the AI first agenda, what relevance do they have?
If nothing, I'll give you an extra point just because you're sharing your data with us.
Two points. It's very valuable data.
Yeah, there is very valuable data. We'll also go with A, buyer, whoever owns the data owns the game simply because of the fact.
You can't do that, this is gaming it.
They're not gaming it. No, only simply because they understand the business as well as whatever, let's say, even if it's a geopolitical situation that is changing because they own the data, the difference between any historical, you've seen in the recent, the thing that whatever historical trends were there, no matter what model you use, you could not predict what was going on, right? And that's only the companies who own the data kind of knew what the factors were and where they were fluctuating, and where they could basically mine the data, and when it made no sense to mine the data. So that's the reason why we would go with that. I'll also note that we're already ahead, so if we just pick whatever they say.
Yeah, exactly.
We did spot that. That should be a negative point for that.
You can give them that extra point if you want for the HFS. We have given them an extra point. They're still losing, but let's see. Let's see who wants the answer. So I think this one, at least we have a little bit of a consensus. I think for all the other eight or nine questions, we could have gone either way. I don't think, but this one, I think it's whoever owns the data owns the game. And so it will be a buyer's market, it seems, both in the game and in the real world.
Indeed.
Indeed.
So I think the service providers are misunderstood. And what I mean by that is, I don't think service providers have always done a good job of helping the client understand how to be a good client. So like what does it take for the client to get the outcomes they want and what conditions should exist for the suppliers to do their best work? And I think that's, so when you have that environment, you'll never move to a AI first contract, as an example, or an engagement. Because you're always trying to figure out what's my incentive and what incentive is being taken away from me. And that continues to be the problem.
Yeah.
Well, to that point, I mean, organizations are trying to see. I mean, they used to be thinking about mobile first as an organization and all that. They are shifting towards AI first as an organization. Not just from a productivity perspective, but what the amount of data that is there at the fingertips to take the right decisions. It's probably headed that way.
Yeah, so look, I agree with that intent.
This is a real family feud.
Yeah, yeah, yeah. I agree with that intent, but the environment is not being established to make that intent happen. And so if your incentives are not aligned, the service provider is never able to do its best work. And the buy side is never going to realize the results if they don't create the environment for the supplier to be successful.
All right. Look, I think the results of the game are congratulations. Drinks are sponsored by Citi. Thank you, everyone, for being a good sport.
If you don't have an account, Register here |
With the exception of our Horizons reports, most of our research is available for free on our website. Sign up for a free account and start realizing the power of insights now.
Our premium subscription gives enterprise clients access to our complete library of proprietary research, direct access to our industry analysts, and other benefits.
Contact us at [email protected] for more information on premium access.
If you are looking for help getting in touch with someone from HFS, please click the chat button to the bottom right of your screen to start a conversation with a member of our team.
