11:00 to 11:25 AM · Wednesday, May 13, 2026
words of wisdoms, or at least those observations, I'd like to invite Phil first, back up on stage, and with Phil today, this is going to be Manish Sharma. Manish is the Chief Strategy and Services Officer at Accenture, and, as you see from the big banner that we have, a former rock star that joined Phil in his podcast. So, please join me in welcoming Phil and Manish to the stage.
Thank you. We're going to do this. All righty. Not bad. Four minutes behind. Just get a couple more people back in and we'll get moving. Okay.
We're barking.
We're hurting. We're hurting. He's making me nervous by seeing some papers.
Yeah. Trying to figure out how to introduce.
And I don't know. He's never been scripted, so I'm not sure what he's looking at.
Okay, I'll get to it. So I think I've known Manish for, since about 2010, something like that. So about the time we founded HFS actually, and Manish was, I think he was the COO of the old BPO. Yeah, or delivery lead that time. Delivery lead in India. And you know, he's been a real voice of energy and vision within Accenture. Anyone who has worked closely with Accenture, particularly big clients on big deals, I'm sure Manish is part of the conversation, and he has been elevated to the chief services officer at Accenture, basically running the whole portfolio of business transformation, AI and delivery for an 800,000-person company. So there couldn't be a better person to talk to about how to deliver service value at scale and go through the biggest pivot we've ever seen in our industry at the same time and maybe can share some of the challenges and opportunities ahead. But it's great to have you here, Manish.
It's always fun to be here. And always, Saurabh is here, Phil is here, and I think we always have a very, I would say, very tense relationships. It's not all easy because we are all the time arguing about stuff, which is always fun, right? With Phil.
Yeah, yeah, Accenture has to be top of every analyst report, otherwise we'll get a rather irate call from Manish or somebody, usually happens. But let's get started, you know, reinvention versus transformation, where has it broken down?
First, I just, you know, this event is great. I always like to come for this because just meeting all the leaders from different places is incredible. And this time I had a bonus. And a very, I would say I'm just coming from a very emotional meeting of 10 minutes with my ex-boss, which was Debbie, you know, and it was so good to meet her here and I see so many colleagues, ex-colleagues here. So Phil, thanks for having me here. I think let's go on this reinvention stuff. We all talk a good game on all these things. Right now we talk a big game around AI and stuff around that. The reality, it is very, very tough. Debbie and I were just chatting that we were together when we did our first RPA 2013 and we were together in Mumbai when we were doing it. And now it is 2026. Even the promise of RPA has still not delivered. You know, so let's kind of get extremely real about some of these pieces. Has it been kind of fully deployed by all my clients across the patch? No, not yet. Now, AI has come, we're talking again, massive stuff, and you know, the hyper frenzied headlines everywhere about how the world is changing. Technology is great, access is, you know, great. The cost is coming down every day of that. But guess what? The operating model is still not tuned towards it. People talk about reinvention, is it permanent? I think it's permanent stuff. You cannot be just going and doing something for a few months or a few days and kind of hope that it is gone. We put a model, people will make a video, they will put a LinkedIn profile of what great stuff has happened. Where is the ROI? The best part which I always hear is how great the token usage is. Token usage is equal to activity, but it has got to be linked to outcome. And nobody's talking about an outcome. And I'll give a client example where we were just asked to come and say, help us put AI and just do some few pilots. This is what I love. I published a paper which is from pilot to profits. So we went there, we kind of explained them the whole stuff and they said, oh, we got it, you know, and it is great, and they deployed that whole thing in their stores, in 50 stores, and then they got the bill. They stopped it. But we said, what is the ROI? So I think the reinvention part is failing because people are still not changing the operating model to kind of really take advantage of it. Despite great technology, great excess cost coming down, the operating model has to be fixed to kind of get there.
Right, so what is working for the few who are breaking through versus not working for the many who are maybe struggling? What are the few who are breaking through doing to actually fix the mindset from one of inputs to one of how do we actually reinvent the model? Five things.
First is, do you really have your process in the right place? Automating waste is not the best scenario, and I think people continue, whatever you are trying to do, if you've got bad processes, trying to kind of get over that is not going to work. So first is, please fix your processes. Second one, digital core. Is your digital core ready? Is it scalable? So that's the second one. The third one is around data. Is your data in the right place or not? So those are kind of the three things we talk about. And then the two underlying aspects. One is operating model. This is not a functional gig. This is like what I call is a mega process change. And some of the clients which are in the public knowledge is like Ecolab, really focused on large scale mega process transformation. And the last one for me is talent. Do you really have the talent which can really take this and make the change? So if you don't have this five, having a huge amount of token usage does not really help. That is why most of the clients, when they come and they even ask, now does that mean, earlier you always want to kind of help your clients, but our percentage of clients who are going from pilots to going to scale is almost like 15 to 17%. And if you want to be in the 17% and you want to kind of go ahead, you've got to fix the two things. And I think process, nobody ever believed in process. And even today, most of the clients, like if you say I want to do a process smart and I want to re-engineer the process, they hate it. They said, you know, what are you trying to do with this process stuff? Can't you just put AI on top of this? It's like, on what? You know, we are the largest partners of OpenAI, Anthropic, and Gemini. We got accredited 2,000 folks on cloud. Now, the reality is, it's not about just taking a model, which are great, fantastic models, but how do you really deploy this and how do you change the processes? So I think those five things are important and I will just say that please do not think that this is about driving productivity. The best successful examples that we have got and more sustainable ones are the ones in which you are driving revenue through AI. Revenue growth is the big story. Like one of the consumer goods companies that I'm working with, we are doing lots of stuff about classical finance and we're doing S/4. The real impact that we are making is putting AI and actually in a very challenging market for that consumer goods, we are increasing the sales by 6.4%. Now, that is the real story. So I think I will just encourage everyone as we kind of go through this journey that this is not a productivity game. It is about revenue growth, cross-selling, upselling and AI really helping you, that is where you are getting the real, real benefits for our clients.
Right. And you talk about the process, you talk about, you know, the model, but this is like a leadership mindset shift in terms of the type of skills at the leadership level needed to make this work a shifting. You need people who can balance that vision with that ability to understand how to execute, the doer mindset.
It is, for once I agree with you. Right, because it is on the, this has to be a CEO agenda. And the most incredible examples that I have are like when I talked about Ecolab, Christophe, who is the CEO, is himself directing this. It was, that is kind of just the power of when you are trying to do this. We had, I think, Saurabh was there. We had an analyst meet in Bangalore, I think a few days back, and we had a lot of clients who came, it's one of the stories which, I don't know if Saurabh ever told you, because he never will tell any good things about us to you. But, you know, it's about, UPS CHRO was there. They do what I call a surge hiring in the peak period. We use AI to hire 300,000 folks in 12 weeks. It used to take the whole algorithm and all the stuff, it used to take days. Now we do this in hours. And he said that in a public forum, but people leading, Darrell himself is fully involved. Darrell himself is fully involved. It is the CEO agenda, it is not something that somebody is doing in the side job or a hobby job on this.
Yeah. I mean, I think this is the crux of the big conversations happening right now is if you don't get the CEO, the COO, the CFO bought in to the real outcomes here, you're just ending up with a yet another technology, right?
You know, as I said, this technology is there. It is a great technology. Let's not criticize the technology if you're not getting the ROI. I am not seeing the clients saying, like nowadays, even when I go, even Accenture team is kind of showing me the great pilots that we have done. I always open the P&L. Revenue, COGS, SGA, and I've got the line items in each of them. And I say, tell me this one, which line item is it going to impact? What is the current baseline, and what is the impact, and when will I see this? Till the time, if you don't see this, please do not start those pilots. The only companies that are making money are the model companies, as you can see, because there is no ROI still in most of my clients.
Right, I mean, and it feels almost like the model companies are panicking a little bit because they've got lofty goals to hit and their clients aren't moving fast enough. And then launching these enterprise offerings and deployment services companies and things like that. How do you see that playing into the services model as this evolves?
It is, you know, like all the companies, we are the number one ecosystem partner for most of the model companies. As I said, we have the largest scale with most of them. I'll tell you, for the service company, this is my 32nd year with Accenture. You know, worked with Debbie for almost, what, more than 25 years out of that, right? The reality is that this is the biggest tailwind for this industry. I think what we will have this AI industrialization wave which will be there for the next 10 years. I think this has given a new lease of life for all the services company. It means if you even look at some of the stuff that we look at or some of the announcements we had of them, they're all trying to buy some consulting companies. And why are they buying consulting companies? It's 150 people, so it's a small scale, this thing, right? But everybody wants to have access to somebody who can deploy these models. So I think despite everything that is going on in the stock market and everything else that is right now visible to us, the reality I think people are missing out is that this is what next 10 years, the biggest tailwind that you can ever see in this industry. I think it has given a new lease of life to this industry rather than other way around. That is where I kind of just see all my clients. We serve 9,000 clients at $70 billion. Like we are there at the front line. We are what I call the most incredible distribution channel in the world. And what you need is trusted advisors, trust-based relationships who can believe in the model being deployed in the right way. So I think it is just absolutely fascinating for me. I can say that I've never been more excited in my 32 years as now. If I look at all the waves that I've seen, I've seen the digital wave, I've seen the offshoring and everything else, this has been the most incredible stuff. And we are hiring, for example, in the next three quarters, we'll be hiring 10,000 folks for data in AI. So we are going on a massive, we will absolutely have rotation and reskilling of talent, but we are hiring entry level people. We are hiring very senior data in AI folks and we will be like, you will see some big increases in our stuff as we go along.
So how is this changing your thinking around hierarchy and organizational matrix. Are you really flattening out the whole model?
I love this one, right, because this is where we might get into a fist fight, but you know, we'll have to maintain the decorum of this meeting. People are saying that there are two views right now. One view is you don't need entry level people. Because all the work is kind of getting coded away. The second one is you can use the junior people and they can do the middle management work. So there are the two parts of it. I personally believe that you really need the entry level people because I think they're smart, they're AI native, we can train them to be AI native, that is what we are trying to do. So I don't know how many of you saw, but we purchased a company called faculty.ai. In fact, the CEO Marc Warner now has become the CTO of Accenture. He recruits 10% of all the PhDs which are there in Oxford and Cambridge. And we are launching the same fellowship program now in India, in Australia, in Brazil, and a few countries in Europe. AI native talent. Now, where will this go? I personally believe that you will still have the middle management. I think we will be hiring even more aggressively on the end of this year. We hired 13,000 entry level people. That's higher than even the previous year. So I think the roles will change. I think the middle management will be orchestration of agents and some of the teams which is there. I think the entry level people will be doing some more value added jobs on this one. I think they're AI native. And when I say AI native, I want to be very clear. And even when I think about myself, like I have a software engineering background, right? But the ability to not to code, but to figure out how to do AI first is a very different challenge. And I don't know how many people will go through that journey on this. But that is what we are trying to get. So mix of right software engineering skills, AI natives, we call them reinvention deployed engineers because we believe that technology is getting commoditized. It's fine. Models will get commoditized over this thing. But the real challenge for us to solve is the intersection of industry, process, data and AI coming together and trying to solve for clients' business outcomes. So think in that way. And our RDE is going to be a very important element and I think it has never been tried, but can you in one person instill function, process, industry along with technology skillset, just imagine that part. So I think you just need a very different skillset. Will we do the same work with less people? Yes. Will the productivity go higher? Absolutely yes. You know what we used to do, it will go higher. But will the jobs go away? I personally believe it will never. It's like the same, this thing. We said, this was Debbie, I don't know, in 2020, or I think when you were the CEO, we had a big discussion and we said, Accenture will never cross 500,000 mark. And I remember, Pierre was the CEO and Debbie was the CEO. We actually said Accenture will no chance to go and cross 500,000. Today we are 800,000. Now, is there a convergence between, like you have this, Saurabh and you have this famous this thing about services is going to eat software or software is going to eat services. I think they're going to converge. And we're going to create a new category, right? Which, you know, Saurabh and I have been trying to play with some names. He gave me a name, I didn't like the name. But, you know, so we'll come out with something which is totally different category. Because software services all will be required to deliver some outcome. And it's going to be, I think, probably a very pivotal moment for the industry.
Yeah, we were talking about it this morning. I mean, I call it the Doer model. And it was like taking a Formula One engineer and putting him or her into the car itself to drive it and test it. But it, you said technology is commoditizing. So the shift, this opens up a huge opportunity for people with business skills, people skills, data skills, to embrace the commoditized technology because it's now about managing the technology, not using the technology so much. So this is actually opening up a message of opportunities for people.
Correct. You know, it's like, that's why I'm so excited about what we have today, right? Because just go and see every, and this is where, again, and I think we have heard, and many of you know about the Jevons Paradox. The more cheaper it gets, you will just see an explosion of folks. So people say, are you, one of the things which is there is that, oh, Accenture, you've got 780,000 people. How is it going to work in, is AI, AI really has impacted only one place in a massive way, which is on the coding side. Do you know how many coders Accenture has? People who code? 30,000. So you know, it's not like you are doing some real, I will say, you put AI into some of the large, complex legacy organizations, AI will run away with fear. Because it is tough to kind of just go and get alignment and deploy, getting the right test environments, putting them in a very interesting mythos has come. So now we are running, it is creating another massive opportunity in security actually, identifying new vulnerabilities. And now we are seeing an explosion in security. We are seeing operational technology, just imagine nobody has paid any attention on operational technology. And the security in the industrial plants, and again a massive opportunity. If you just look at some of the pieces that we have done, we took, you know, again, I don't know how many of you saw, but we took DLB Associates. So we are creating a network of businesses and DLB Associates we took 65%. What their business case was, they right now, they are double that amount. Faculty.ai frontier platform, you know, killing it. I don't know how many you saw, but the biggest control points is data. And this is where we got a company, people are saying, "Why is Accenture buying Ookla?" You know, on the network side. And that is giving, I'm getting the control point, which is data. You know, and then we are seeing, you know, as we go, we have a partnership with ANSR, which is going on to the GCC side and trying to deploy AI in the GCCs. You know, again, a massive opportunity, you know, which is out there. All the, the 2.6 million people in about 1600 GCCs in India. And a real good play, none of them are AI native. And there is, again, a massive opportunity to kind of transform all of that.
It sounds like you're building the execution layer we've been talking about, what do you call it, like the Bloomberg of network intelligence, or the goal, but the bravery to go out and make parallel acquisitions that are broadening this control layer is where to go, not just buying the same as, or just trying to add 50 consultancies.
So you see, most of the stuff that we bought this year, we spent already 1.5 billion, we're going to spend five billion capital outlay in acquisition of companies. And what you will see is very different IP, outcome-based, very differentiated skill sets, platform-based, that is the pivot of the convergence that you have been talking about for about 10 years now. Finally, whatever you said is coming true.
I'm not very happy to say that, but I have to admit that.
I'm so happy to stop thinking I'm insane.
Great, well, I think one question, and then I think we've used all our time. Do we have a question for Manish? How do you think about the generation of kids in high school?
Is this debate where they should be allowed to use AI to learn or they will cheat on tests? So how do we start changing this next generation right when they are in school, allowing them to use AI for good, how do you change the actual tests? That is not like I don't think the courses have changed for years, right? The professors or whatever give the same test, which it's very easy to use AI. So how do you start changing the foundational system in any country to where kids now have to be able to learn with AI, the testing is different. You need to actually explain what you did as opposed to get the right answer, and for a math problem as an example. Because that's the foundation layer that needs to change for us to be able to cope up with this, right, going forward?
So that's a very tough question, actually. You ask me the toughest question. Like, you know, and I will not give an Accenture view, I'll give a personal view, because I have a strong view on this. I do think that you have to not let kids, like, forget the basic reasoning ability. So you have to control some of the stuff so that they are learning the basic skill sets as they go along. And I think probably at a certain level, probably I will say, this is again, I've not done, but I'm kind of very interested in this subject, but I think around 11th or 12th grade is when they should be allowed to kind of start using AI in a different way. But you have to have a deep foundation on reasoning of ability to think. You cannot forget your ability to think. Because finally it will come down to what will be the only thing that remains in this after it is fully commoditized. It will be about experience. It will be about experience. It will be about trust. Trust in this, all the human qualities actually will come to the fore. So you cannot lose your basic human qualities because that will act as a differentiator. Like I am a big believer that experience is the most important thing in every company that will be there out there in the world because that will be the only differentiator you can have. So yes, control to some point, but afterwards, let them run with it.
Terrific. Do we have time for one more? Atul.
So trying to bring the last panel and this discussion together, look, I think a huge amount of discussion because many are providing IT services or BPO services, the focus is so much on debt, process debt, data debt, all that, and Phil talked about leadership debt, and I wanna focus on that because I think at the pace at which tech is moving, that there is an imagination debt. And the reason I talk about that is if leaders cannot imagine what outcome is possible and what layer do they need to build so that when the tech, when, you know, Mythos gets released, how do you leverage what you have to take the benefit of it? I think that's what's getting lost.
Completely agree. I think, you know, you have to, this is where I said the story has to be revenue growth, not cost cutting. In fact, if you are getting people, can we use those people to get your revenue to be accelerated? That thinking has to be there. That's why this is a CEO agenda and not an agenda which is you're giving to some leaders, in the organization where they are trying to optimize their own piece while they might be sub-optimizing the full piece. One of the clients that we did, and this is a good example on this great leadership, while we reduced, they had like 100,000 folks in a consumer contact center staff. What we really got the answer was a billion dollar of revenue uptake and cost reduction of 500 million. So the revenue line was much bigger on this. Tell me how many places people are talking about revenue growth versus a productivity game.
Only in the small percentage who've gone enterprise-wide with the focus and have the full C-suite engaged. You need the CEOs to be there in the game on this. Terrific. One more. Arindam.
I think I can speak loudly. Hi, Manish. In the former Citi, now working with Phil and Saurabh. So glad to be here. You talked so brilliantly about the people with the process knowledge are really key to make this revolution work. Do you see that the enterprise is starting to see those people? Because they have always been behind the scene, who just made the place run, and kind of not that recognized. Do you see that recognition is coming through to make it happen?
15% of the clients, one five, not five zero, one five. Because they think it's too much of a hard work. And there's also a feeling of lack of trust that this consultant's trying to make some money. You know, because they want to do this process stuff and everything else. The value around that part is still, so if you look at the five things which I mentioned, process, digital core, data, I think digital core, yes, lot of focus. Still not enough on data, still not enough on process, very little on operating model, and very little on talent. That's why it's a 10 year journey where it will be a massive tailwind for this industry.
Terrific. Well, on that, I'm gonna thank you so much.
Thank you. Enjoyed it.
Thank you, Manish. Come up.
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.
