Third Bridge专家讨论生成式AI对埃森哲及IT服务外包业务的长期冲击

[[00:01:36]]
Accenture has been described as the best-in-class in the IT services sector. The company is often a good barometer for where the broader sector itself is positioned or heading towards. Could you provide an overview of its business, focusing on its positioning within the broader IT services sector?

[[00:02:09]]
It is known that Accenture is a skilled leader in the IT services management consulting and IT service provider. Accenture has had a strategy as part of its scope of work over the years, but it's really known for management consulting and IT consulting and the top stack from technology consulting, app development management and operations with their BPO work. Accenture is differentiated just by scale. Accenture is one of the large, I'll call it, management consulting companies that sits around. It's probably tagging right now 700,000-something people. Its scale and breadth gives it a large opportunity and more of a positioning across the board because of its different portfolios of work that they do. When one look at it, we talk about it more from a barometer perspective in the sector. When you talk of any other company, where they say where Accenture goes, the mid-tier providers follow. Everybody looks at Accenture as almost a barometer from every capability in the management consulting and IT consulting space.

[[00:03:19]]
Accenture's Q3 FY26 results noted 6% revenue growth and a 9% increase in earnings per share. The company saw incremental improvements in margins, generating USD 3.6bn in cash flow. Bookings were one of the weaker optics. There was some strength in consulting bookings where there was 11% growth, which is the highest that we've seen over the last four-year period. Despite revenue and those financial metrics, P&L performance was in line with the Street's guidance. There are some positive inflection points to glean from Accenture's results. The share price slumped pretty significantly, declining in the double-digit-teen range between 16% and 18%, and it has not recovered since. Where do you think Accenture sits today? Do you view the company as being in a positive or negative position?

[[00:04:50]]
I think if you look at the debt metrics that you just talked about and you look at some of the numbers that we shared as part of those earnings, the metrics tell us, from my perspective, two stories. The discretionary and consulting side is generally soft. If you can think about reflecting on the macro conditions that were there, well-managed services and gen AI are the good engine for carrying the quarter. That's what Accenture positioned itself at. Then you look at consulting bookings, book-to-bill, which I think was about 1x or something like that. It's a number that you have to look at because it signals clients, they are not green-lighting large, discretionary transformation. If you think about the way clients are looking at right now, most clients are not looking at doing USD 100m transformation programmes. They're looking at, where do I leverage, in this case, more agentic capabilities or maybe managed services to be able to improve productivity and reduce costs? They're not looking at the same large scale. That, you would tell, is impacting. The gen AI ramp-up in some sense is real.

It's fast in some areas, specifically in the managed services sector where you're looking at gen AI, in an essence, taking over some of the work that you had RPA capabilities doing. It's not yet big enough. It's not where you're going to say, "Now, I'm going to sign this large USD 100m to do a gen AI roll-out or an RPO automation roll-out." Companies are not just going to put discretionary spend towards that. There's a genuine point on that. I think when investors look at that and they look at Accenture's books, not just the numbers they reported during that earnings call, their books overall does have a significant amount of managed services and BPO work. If you start thinking about it, looking at that slide, I think people are thinking about how would AI cannibalise that Accenture backbone. Most companies are going to be looking at, "Hey, if I signed a five-year, USD 100m now we're two years in and there's all this gen AI stuff, I'm not going to pay you the same fee because I'm anticipating that you show some efficiencies." That's where I think investors might be looking at their backlog of what they already booked previous years and what that will return relative to revenue and would they see other companies looking at renegotiating their contracting, which, at that point, definitely will impact Accenture's bookings or Accenture's books overall.

[[00:07:29]]
Several sell-side firms, third-party research platforms and providers have discussed the belief that, over the next decade, increased demand for generative AI and AI capabilities will largely mirror what we saw in terms of the demand curve with the cloud technology cycle over the past decade. What is your perception of this AI technology cycle, particularly when we compare it to the previous ones such as cloud or mobile?

[[00:08:16]]
That's a good point because I think there's a difference. If you think about the cloud, the cloud had primarily, I would say, three big shifts. One of them was the ad migration from hosting your own capabilities to looking at shared data centre spaces like co-location. That was one. When that shift happened, companies looked at how much they can reduce spend from, call it, real estate spend from data centre and hardware spend. That was genuine dollars that you see. You actually saw the dollars and those made sense. On the fact of shifting from now not just co-location and you're moving things to the cloud, so you're migrating on-prem apps to a cloud infrastructure, that was genuine spend as well that you see on your books because now you reduce your spend from actually co-locating spaces and hardware and now to a managed service, SaaS or IaaS or PaaS-type environment. You could see the money. Third was around just transforming back to cloud-native. When you started building things with containers and things like that, again, you see the move from one to another. From an ROI perspective, you can actually relatively see that return on the investment.

When you think about gen AI, I think that's where, from my viewpoint and looking at it, the challenge you get for that is you've got all these things now in the cloud. You've got the data on the cloud. You've got whether it's just your big data capabilities or your ERPs or everything is SaaS or cloud-driven with workflow orchestration all in the cloud. The problem with agentic capabilities is that we already had RPA capabilities that automated the day-to-day, the mundane things that people did on a day-to-day basis. If you start going gen AI and leveraging it for day-to-day work beyond just prompting like you use ChatGPT or Claude or all that, now with those gen AI providers or providers now charging you per transaction, that's a completely different thing because the level that you're being charged for that transaction is different from just your one-time cost and savings of migrating capabilities to the cloud and being able to just pay for utilisation costs. Now, you're paying for utilisation cost of your data on the cloud, but now, you're also paying for transaction cost for every prompt or every transaction you pass through agentic capability.

That makes a big difference. I think until the providers that are providing large scale agentic capabilities have actually fine-tuned their pricing, so that pricing is either by user or in a certain way that companies see a significant return on investment and actually allowing developers and others to leverage it in a large scale, you're not going to see that adoption because the biggest thing will be… I think relatively in the news, we've seen Microsoft, Meta and others provide feedback around how much your developers are spending just for using tokens. That's something that you cannot justify. You can't justify a developer that you're paying USD 200,000 a year and then that developer is spending USD 500,000 as part of leveraging tokenisation to be able to be effective in your job. What is the return on investment there? For me, it doesn't show the value. I think the adoption is going to be slow just because of the fact of how providers are monetising their agentic capabilities and what companies are going to see us a return on investment on those capabilities as well.

[[00:11:30]]
Accenture's average price-to-sales ratio is around 1.5x, which is less than half of the S&P 500 average of 3.7x. However, the business is generating billions of dollars in annual revenue. The company saw a little over USD 3.5bn in free cash flow in Q3 FY26. Wall Street is valuing firms such as Accenture and other IT service providers as companies whose advantages in the market are rapidly becoming irrelevant due to AI. On the flip side, I've read several research papers and reports that note that one in 10 pilots make it to production.

There's tremendous enthusiasm from enterprise leaders, chief information officers, chief technology officers and executive members of organisations to adopt AI, but they often lack a coherent AI strategy. They're buried under technology, data, process and talent debt. These are all areas where you would imagine a player such as Accenture can step in, given its pedigree to help resolve those challenges. It would be supportive of the future outlook, not diminutive or taking away from it. I read that the technology debt is around USD 18tn across the Global 2000 alone. How are you reconciling the way Wall Street is positioning companies such as Accenture and thinking about the future of Accenture and the broader IT services sector vs the reality on the ground?

[[00:13:52]]
I think investors are going to look at Accenture and any other consulting company. I think that's publicly traded in a certain way, but it also would depend on, I'll call it, their book of business and what is in the backlog because I think the challenge with Accenture, like I mentioned earlier, is Accenture has a significant, heavy backlog that is based on large, I'll call it, BPO operations deals and managed service deals, not the large transformation deals. I think when investors look at that, they start looking at the disruptive part of gen AI, the disruptive part of AI in general, where you need to be able to transform. We're not talking about a large-scale thing where you need a lot of tokenisation or, I'll call it, advanced development to be able to be faster in implementation. That wouldn't be it. If you had to look at it as an investor, you look at, okay, if I wanted to compare Accenture with any of the other traded maybe IPPs as well, where do their books of business fall? If you look at a standard strategy and consulting company or a standard IT advisory company, they won't have a heavy book of business in managed services or basically BPO, but Accenture has a significant amount of that. That is basically prone for disruption by gen AI.

With that in mind, I think investors are looking at, okay, if Accenture gets disrupted in that space and if Accenture's clients are looking at renegotiating some of those contracts because they believe they can get benefits from it, then the Accenture value and proposition actually goes down in terms of scale because if Accenture was saying, "Hey, I have the scale, I have the breadth and I have the knowledge," the scale becomes an issue because that scale becomes impacted by the agent capabilities you can put to take over those BPO work where you said, "Hey, Instead of you using two of your people, I'll bring 10 at low cost from India or Philippines or wherever to do that work for you and do it faster vs you paying one month," which was the proposition of low cost labour. Offshoring, that helps to be able to do that. I think that's where, in my view, investors are looking at Accenture's books and looking at why it's changing in terms of the, I would say, stock drive. It's just its book is heavy on legacy, meaning legacy capabilities tied to managed services and BPO.

The traditional SI work that you would price, I think you would price that today because if you went today, Accenture went to do a USD 100m transformation for an ERP or for any other modernisation capability, it's easier for them to price gen AI in because that's usually straightforward. I won't call it straightforward. It's complex, but that's almost really at a fixed fee, really at a multiyear like 5-6 years BPO or managed services. More of those don't really have, I'll call it, tech debt because most of those are either migration from one platform to the other. Many of the client has already rolled out and looking at the legacy applications they're moving from or the legacy infrastructure they're moving from. I think that's really when you look at it, comparing Accenture with the others is when Accenture has so much in terms of managed services and BPO in its books, that is basically a base of its revenue. That becomes a point as an investor to look at. Is that going to have an impact as gen AI becomes more prevalent in those areas where you can easily do those things without needing that many people offshore or nearshore to do that work?

[[00:17:29]]
Accenture's managed services bookings in total for Q3 FY26 were a little over USD 19bn. Consulting bookings were over USD 10bn, and USD 9bn came from managed services. What do you think that will look like as we think about the evolution or maturation of the AI theme? I've seen some rough research that points to managed services, BPO [business process outsourcing], ITO [IT outsourcing] and IT consulting type of work around global technology services, which is a USD 1.5tn TAM. Previously, this market was forecast to grow at a low-to-mid-single-digit rate. You mentioned the theme of cannibalisation.

Do you expect to see the cannibalisation of global technology services that are lower in the software stack, such as managed services, BPO, customer care outsourcing and IT infrastructure outsourcing? If so, what does that look like? I've seen some estimates of negative 3-5% CAGR declines through 2035 as a result of that cannibalisation. If you were to expect cannibalisation, would that be a fair way of framing it?

[[00:19:14]]
I would expect to see it because I think at some point, when you think about it, I think from a technology perspective, if I can go back before I even bring out the AI cannibalisation, think about today. When you think about, I'll call it, what do you do today relative to capabilities, when I turn and I want to ask a question to something, before, every search you had, you went to Google. You did a Google search. You went to Google and you did a search and that was your total. When you talk to a kid 5-6 years ago, it's like find something, Google it. That was everything that everybody looked at. Now, the conversation is different. The point is talk to ChatGPT or talk to Claude or talk to Gemini or what else. Yes, the search engines are still there, but people are pivoting towards the thing they feel is more convenient. I think there is going to be cannibalisation, specifically in that managed services and BPO space because a lot of that companies like Accenture have rotated in basically gaining value and improving revenue by doing a lot of RPA work to transform those BPO procure-to-pay type work, revenue cycle type things and leverage a lot of RPA to drive down actual cost and improve the revenue and profit.

Now, with everything that's already been RPA, with everything that you have the knowledge management existing, if you put gen AI on top of it, maybe after the RPA, you had 10 people that were still doing the work, you might end up with five people now or even two. The flip side of that is, how do you monetise those additional people that are now being displaced because you've implemented some gen AI agents or capabilities to replace that? Do you find an opportunity within that same client to have them do more high-value work, meaning be able to find more insights in the data that the agents are doing? Maybe find an opportunity to say, "Even though I have 10 people doing this work, now, I have three agents doing it. Those three agents might be costing me this much for the tokens that they're using. If I can leverage the other headcount or resources that were previously doing that work to find high value and do actual high-value consulting work or high-value business analysis for the client, I might improve my revenue that way."

That might be the only way. You can say you might be trying to be almost neutral if you can still see the cannibalisation and still repurposing the people. That is going to be a 50/50 depending on the client you talk about because the clients are going to look at it as you can justify putting an agent and the agent is going to cost you X and the client will probably say, "Yes, I'm okay to compensate you for the agent and the effort you're paying for the tokenisation." In that sense, you might go back and say, "Okay, what are you talking to, Anthropic? Are you talking to Google or to OpenAI or to Microsoft on the AI capabilities?" You're leveraging to do some better pricing on the end on how the tokens cost you, so you can make some revenue. Those are things that all companies doing managed services, BPO work, leveraging tech to do that, will have to factor in into their pricing and the work they do.

If somebody tells you that the cannibalisation would not lead to any negative revenue impact, I think it will be a watch if you can find new opportunities, but I think there's going to be a negative impact on revenue because whether that's negative 2% up to 5% or 10%, it's going to be business to business. Heavy businesses, that really depend. For example, if I took take some of the IPPs like the Tata Consultants [Tata Consultancy Services] or the Tech Mahindras and all that, they might get more heavily impacted than Accenture because a lot of them do a lot of infrastructure outsourcing work and they do a lot of that. Those are easily prone to agentic capability and some of the code that is being built today with leveraging Claude and all that. They might have more impact. They might see the negative 10%. Accenture might see the negative 3% or negative 2%. There is going to be some negative impact that results from the AI cannibalisation of work.

[[00:23:20]]
For decades, IT service providers such as Accenture, Tech Mahindra, Indian pure plays and the sector as a whole have priced effort by the hour, tying revenue directly to headcount in FTE [full-time equivalent] models. Revenue per employee was once a key metric that was tracked across this industry. AI fundamentally breaks that relationship because the value no longer scales with labour.

I've seen the next generation of services and I want to talk about services-as-software – a terminology that was coined by a third-party research firm called Horses for Sources. That firm talked about how the next generation of services will be increasingly priced around separate sources, such as business value, performance, personalisation, prediction and productivity. As we step into this agentified world, or as we see this AI theme mature, which monetisation models could we expect to see across the industry? Do you think it will fall along the lines of things that are valued in terms of performance, personalisation, prediction and productivity? How do you see the monetisation or the existing industry model shifting as a result of AI?

[[00:25:12]]
That's a good point because I think the productivity component of it has already been one that a lot of the IT consulting, including Accenture, has done. When you've gone into a shop to be able to take over an IT shop, you tell the CIO, "What I'm going to do is I'm going to take this 200 people onshore and take them offshore and I'll give you 150 of those people instead of 100 at a lower cost. (1) Your cost barrier goes down. (2) I'm guaranteeing your productivity over time of X." That, you're basically saying you have the scale, you have the skills, you have this, I don't have it, you're taking them over. It's going to be cheaper level for me in terms of my overhead count. You're going to be heavily productive and you're going to be able to meet my needs. That's people. Now, replace those people with AI. From a productivity perspective, would you go back and say, "Okay, I promised you 30% productivity YoY for the next five years. Now, through AI, it's going to be 50%."? The answer is going to be no because that's the challenge that people have.

From a productivity perspective, the agents are going to be as productive as the systems they run on. That's the one thing that people tend to forget is the agents that you deploy and the agent capabilities you run are based on the platforms that they run on, the CPU consumptions or the GPU consumptions that they have, the tokenisation or the tokens they're going to leverage. How do you monitor them? Who monitors them? Who audits them? There's going to be a lot of other things, factors on top of it because once you turn a bunch of agents that are doing capabilities, especially in banking and finance where it's heavily regulated and there's a heavy focus on making sure data is correct, accuracy and all, there's always going to be a human that monitors the agents, makes sure the agents are doing what they expect it to do, makes sure the activities are being done properly. That also increases cost. That also comes into security. That also includes ethical compliance. There's a little bit of different things that once you add those up, the savings that you were going to get from productivity gets washed and you start looking at the other additional capabilities you need to put on.

It doesn't mean that you don't see savings, but it's not going to change the productivity that you already talked about saying, "I was going to give you this with these many people offshore vs now I'm going to use agents." On top of those agents, I need to actually have highly skilled people that want to understand what the agents are doing, can program the agent, can intervene, can audit them, can secure them. Now, those are people now have different skills than the low-cost people you had offshore. You might have to bring that onshore, so the cost becomes almost, in terms of productivity perspective, cost neutral. I think the flip side of that is value. I think that's where you can price. Value-based deals will actually be much more profitable at that point because you can do more of a value share, where in an essence, say, on a BPO deal where you're telling a bank or insurance company that you're doing processing for them and saying, "If I can process these many claims…" Maybe today, you're processing, I'll say, 100 claims a day and they're going out the door.

Now, you can say, "You know what, if I can process the many claims from 100 to 3,000 claims per month and 3,000 claims per month yield this much, if I'm able to meet that value for you where you basically your payables are going up, I want a share of that. If the time to pay claims and the time to be able to interject into a certain transaction reduces your cost and increases your profit, I want a share on that profit." That becomes a value. I think that's where you then get much value from leveraging agentic capabilities to do that because at that point, you can tell the client at that point, "Hey, you're getting value because I'm doing what. I want a share in that gain." There's almost a gain share of value-based where the return is actually helping you grow more. I think that's one I will see or more anticipate more pricing around value-based deals and gain share vs trying to be able to look at more from a perspective of optimising productivity as well.

[[00:29:33]]
To effectively pivot away from FTE-based delivery to value-based models, Accenture requires a fundamental shift in its organisation structure and culture. To this end, the company has made great efforts to reconceptualise its business, going through several restructurings. It adopted reinvention. That said, Accenture is an organisation that still has 800,000-plus employees today. How well-positioned is the company in terms of its ability to transition its business to one that benefits from AI tailwinds or the AI theme? Does the sheer size of the business make this AI transition more complex and lengthy? Based on the C-suite and pedigree of Accenture's business, do you have confidence that it can make that pivot and be an effective leader in the industry, as it relates to the AI theme?

[[00:31:12]]
My perspective is long term. Long term, Accenture is going to be a net beneficiary of AI. Just its skill, its client base, its delivery muscle already adopted AI before gen AI, so that has been something that is put in this business. Near term, no, because the near term is that point that we just talked about. From an investor perspective, it's that fear. That fear that the cannibalisation of Accenture's current books by AI is impacting its margins, it's impacting its revenue and all that. That definitely has an impact. Long term, Accenture can pivot because the point is near term, the contracts they have that are now 5-10 years out or so, those need to go to conclusion to any, if you want to think about now, for them to significantly pivot. In terms of a headcount perspective, that's the thing that you have to be able to do because, one, if you've had a commitment to clients where you've shifted a lot of work doing those contracts, those BPO and managed service contracts that you had just said, "Hey, I'm just moving from onshore to offshore and just putting a lot of people and committing to this productivity," if you want to cannibalise that with AI, it does two things to you that Accenture has to look at.

One is, yes, your revenue has to shrink because the client is going to look at it and say, "I'm not justifying you leveraging agents and paying you the same way I was paying a person, so why? Why don't just keep the person? I'm not going to pay you more. I want to pay you less." If you do that, then you're impacting your near-term revenue, which, again, that's where the issue is. I think near term, Accenture is going to be hard-pressed. They're going to be impacted by it, which again is shown with the stock price and investor sentiment. Long term, I think Accenture is well-positioned for long term. As they have new contracts, new, I'll call it, consulting deals, new managed services deals, they can bake gen AI into it. They can look at it from a perspective of, "We already factored it into our pricing. What we've got is already baked into our margins." At that point, as they're transforming, they can easily do that. With the scale then, Accenture is going to have to shed a few. That's the reality of the business.

That's the reality of it. If Accenture is still 700,000-something people in five years, I think that's not a sustainable business from an Accenture perspective. Accenture is going to be looking at long term. The shift from onshore to offshore would actually shift. It will probably be a more heavy onshore presence with highly skilled, agentic and AI skills to be able to serve clients vs the offshore low-cost labour because that is not going to add value because some of that now is going to be replaced by agentic agents and capabilities like that. In my view, short term, yes, Accenture is going to be impacted. The AI cannibalisation is going to impact their margins and all their revenue definitely will see a decline. It will probably stay flat if possible. Long term, I think they are well-positioned because of the scale and the client base they already have. If they can transform with their clients, I think that puts them in a better position to do more high-value deals, which are now true consulting and managed service deals and leveraging AI to be able to develop those things faster and get more return and maybe do more high-value profit sharing or value-based deals as well.

[[00:34:38]]
In 2025, a third-party research platform called Horses for Sources introduced a really interesting concept known as services-as-software, which I'll define as AI-driven operating models where traditional labour-intensive human services or FTE- and consulting-based models are replaced by autonomous, cognitive and outcome-focused platforms. What are your thoughts on services-as-software?

We noted that Wall Street and investors are focused on value destruction and ultimate contraction as a result of AI in areas such as BPO and global technology services. Do you think investors are ignoring the potential revenue upside and value creation that could come through things such as services-as-software, where we see agentic AI becoming more embedded in the core business functions that it hadn't previously touched, thus expanding the addressable market beyond traditional technology budgets? It gives an organisation such as Accenture more enterprise spend to capture. Do you think that value creation is being ignored by investors? If so, does the opportunity in value creation outweigh some of the value destruction that we noted in areas such as BPO and traditional managed services?

[[00:36:39]]
I think it's twofold. I think investors are looking at, I'll call it, the quick pivot. We just call it consulting as a service or services as software, whatever term you want to put on it, basically packaging your services and just selling them as a bundle like you would go sell a SaaS product, too. How do you commoditise that? Remember, FTE count has been what we've always looked at traditionally. You go into every work and you say, "Hey, we're going to deliver this over 10 years. We're going to need 10 FTEs and this we're going to price it by FTE account and we're going to put a premium. If it's a fix fee, we're going to do a risk premium on it or not. If it's a T&M, you're going to pay for the hours of people work." Now, you want to go do, I'll call it, consulting as a service or service as software. You have to be able to put a value to it. Now, as a client, I'll put myself in their shoes.

Why would I want to pay you for something that is… If you brought a human, I'll pay you money USD 150 an hour and you come to me and say, "I'm going to sell you a package. You don't know how many people are in that package, but they're going to just do this as a service and they will just be there to do your work. This is going to cost you USD 100,000. If you do the maths, it might cost you still USD 150,000." As a client, I'm like, "No, it doesn't add to me value." For me, the value proposition is there, but it's actually putting the marketability or the actual profitability or the actual… How do you actually approach clients to sell it? The struggle, too, that you can also see beyond the consulting services as service is the companies like Salesforce, ServiceNow and Microsoft. The challenge is what, in most of their SKUs when they go to license their software as a service, they started by licensing the agentic capabilities separately than their standard capabilities.

If you look at them, they say, "I'm going to give you Azure capabilities and I'm going to give you Azure cloud or I'm going to give you CRM services and I'm going to give you the agentic capabilities separately. Oh, I'm going to give you this." They realise, these companies, "Why am I paying you basically two fees for the same thing? If I just buy your ITSM services with your capabilities, for example, ServiceNow with Now Assist, why do I have to go buy your ITSM services plus Now Assist plus your agentic capabilities? Why do I need that?" People like ServiceNow have pivoted to the point where they've bundled that in the SKU and said, "Hey, once you get this, you get everything." It's easier for them because then that what clients don't really see the difference is just buying a SKU, yes, but the prices are higher. Accenture can do the same thing from consulting as a service or services as a software. You can do that.

The point is you have to be able to… Clients are going to have to see what value do they get relative to from your traditional headcount type approach. Do I do staff augmentation? Do I do your consulting as a service or service as a software? What value do I see from it? Do I get this type of return on investment I expect? Now, if you bundle that and part of those services are agentic AI agents, which you've already negotiated the tokenisation use and their consumption to a point that it actually gives you much more value because at Accenture scale, Accenture can license those things in a much larger scale than most smaller consulting companies could be. That's where Accenture can make money. I think it's actually helping investors understand what revenue that brings relative to the standard T&M or fixed fee model where you're still based on headcount. You need to be able to help them understand what that is. With time, there is value to it.

Every investor looks at Accenture or any other consulting company differently depending on what they're looking at. I think really what is really impacting Accenture's book today and my view would be even that pivot to consulting as a service or service as a software, it is still not going to yield the same amount of revenue that you're anticipating when you do traditional today T&M or fixed fee bids that are resource or headcount-driven because you're going to have to account for some other things like consumption for the agents, if you do an AI. The resource utilisation and the people that are going to be there to oversight, that also erodes some of, I'll call it, the margins you might have seen on it, but it actually reduces the revenue because now, you're going to be selling it for less and your margin is probably going to not be as much as you were looking for and then that just means less revenue and less margin.

[[00:41:25]]
Did you find that the go-to-market talk track at Accenture started to shift away from selling labour and business outcomes to clients? How were business outcomes best measured for one of the AI pilots, for example? What were the strongest approaches for measuring business outcomes for AI initiatives?

[[00:42:13]]
Not even before I left, I think just Accenture was trying to pivot into the reinvention services that they have. They already started the services, not making it part of the whole business, but there was a significant amount of push to expand that across the businesses and not making it now where it's an organisation or it's basically an entity on its own where everything falls underneath it. At that point, it was just part of it. There's been that push. Accenture has always had that push to go to value-based deals. Accenture has always looked at that where it looked at outcomes and find ways to be able to basically earn more, be more profitable, get more revenue in, add more value to clients through value-based apps. Those deals were far or near, because, one, when you start doing large transformation as an IT consulting company, it's hard for you to put value.

When you start looking at healthcare transformations and looking at more how you deliver services from a provider perspective or payer perspective or even a bank, that becomes where the value add, where you start looking and saying, "Okay, if I can help you and cut down your… I use a claims example, your claim processing time by X, the revenue you're getting or the increased profit you're getting from that, I want to have a share of it, 10%, 5%." That actually helps because you are not incentivised to increase that value for the client. You are not incentivised to make the client see more return on your investment because you're now sharing that profit. There's a difference when, I'm not saying you don't do the same, but again, you have more incentives to find. Go above and beyond your traditional delivery because you now have to look at different things. There's always been that with Accenture to go to value-based deals. They've always been that. The messaging has shifted as AI has transformed how we look at consulting in general and why we're positioned.

That has given them opportunity to start looking at that. I think that pivot has been there. It's just trying to pivot their business to be there. Accenture's biggest problem is not the pivot. Like as I said before, Accenture's problem is their book of business to that. A significant amount of the business, book of business is still sitting in the old legacy, active contracts that are running for multiple years. That's where the challenge is. It's not really that pivot. They would pivot, but it will take time. That takes time. You have to shed the garbage, you have to shed all the clunk you have and try to be able to get to that point, so there will be some spend. There will be some investments, but it's a long time. It's a 2-3-year thing for us to start seeing this value and things and for investors to actually recognise that when they look at Accenture's book of it. I know you said a second part of that and I think I might have not addressed it.

[[00:45:08]]
Looking at Accenture's Q3 FY26 backlog or bookings pipeline, the company noted around USD 19bn in new bookings. Is all of that convertible? Would you expect to see only 60% of that ultimately converting into revenue? What is the conversion timing? If we look at USD 10bn in consulting bookings, does that generally convert over 18-24 months, or is it shorter? What does the general process look like?

[[00:46:08]]
I think traditional consulting bookings convert faster than BPO or managed services because typically, they will either be large T&M deals, that will come through on and they get high quarter-in-quarter revenue. Those will convert faster. You would anticipate that those deals that we have, unless a few of those might be long-term transformation projects that go 2-3 years, but those are far… The net-new ones are very minor. You wouldn't find that. You're thinking about this revenue, you're talking about probably the next 18 months because companies are not going to… They're not signing massive, "Now, you need to transform my ERP for the next three years." What the next three years look like, they don't know because the platform they're going to, you tag AI on top of it, it will probably look different in two months. That's the thing. When you're doing that thing, it doesn't mean that large transformation projects are taking place.

They're just taking place in a different scale and a more phased-out approach and a more or less, call it, legacy doing, which just go in and you say, "Okay, we're doing design for six months. We're doing development for six months. We're doing implementation for two months and we go live." That standard is evolving relatively. Companies are looking at, how do I get quick wins? If you can find those wins, yes, they get the value. If you can get what you have, legacy or not, can I put something in front of it that can actually leverage the existing infrastructure and I don't have to do anything in the back end that actually provides me the value I need quicker vs actually ripping holding off and putting something?

[[00:47:51]]
Regarding the mid-market opportunity, traditional delivery models in T&M [time and materials], FTE or fixed price made it uneconomical to serve the companies that Accenture defines within the mid-market space as having between USD 300m and USD 3bn in revenue. More recently, the company has made a notable shift in its go-to-market strategy, where it's going to emphasise companies within the space. Accenture views it as an area where there is a USD 240bn addressable market for the company and it could grow in the high-single-digit range. It launched a new business called Accenture Edge to specifically target this area of the market. What are your thoughts on the mid-market opportunity? Given that it has historically not been an emphasis for Accenture, how much confidence do you have in the company's ability to capture the addressable market that it has defined?

[[00:49:01]]
I think the mid-market does have a significant amount of addressable market. I think Accenture's pivot to it is, I'll call it, a neutral strategy in my view. Why I say neutral? Because I don't think it's large because in that mid-market, you already have mid-tier consulting companies that are serving them. The question you're going to ask as a mid-tier company that now Accenture knocks on your door and said they've got this shiny thing to sell you or to be able to move you to them, what are they going to do differently that those mid-sized consulting companies you've been working with for years haven't done because they are also pivoting? They're not just staying still. A lot of them can pivot faster because they don't have the skill and the legacy baggage that Accenture has on its books, most of them can make those quick investments and more. Accenture can't because Accenture needs to shed that legacy baggage that they have from IT money services and BPO.

Accenture has a tough hill to battle in my view, because, yes, there is an addressable market, but the market is also saturated because, one, when you think about the Accentures, there are very few of the Accenture. In the mid-market, there are many of those mid-market consulting companies that have dominated their markets and especially locally. You take the slow loans and all that. They built their market local. They are there in the doors of their clients. They're talking to the mom-and-pop shops that they've been the markets that Accenture is looking at. They built the relationships. Accenture doesn't have Accenture started from scrap. In essence, is the market there for Accenture. Yes. Would Accenture get a percentage of it. Maybe in a couple of years, but it's a long shot because that has never been their focus from what Accenture has always been.

Accenture has never been a mid-market consulting company. They've never looked at that sector. They've never paid attention to it. They've never invested in the relationships. In consulting, you buy based on relationships that you build over time. For somebody to pivot from a mid-sized mom-and-pop person, the consulting company that they meet every weekend and play golf or do this to an Accenture where it just came yesterday takes time. Accenture has to invest and build a relationship. Can they penetrate it and get some revenue out of it? Absolutely. Is it going to be a big share of their business? Absolutely not, because there are companies that are targeting that and have already been doing it for years.

[[00:51:39]]
Cybersecurity has great potential across enterprises, mid-market and all segments of the space. It's probably one of the most attractive growth vectors, if you look at this through the lens of Accenture's business. In 2025, it made up almost 15% of the company's total revenue. It's a business that has grown to roughly USD 10bn in 2025 from USD 700m in 2016 – almost a 35% CAGR rate over a decade.

It also underscores one of the things that Accenture has brought up more recently, which is the shift in its focus from IT to OT [operational technology], supporting power grids, manufacturing facilities and pipelines that are as critical as an AI-enabler, at least in the eyes of the company. What do you think Accenture emphasised in terms of that IT to OT shift? Do you think the company's efforts in the cybersecurity space and in growing that business really emphasised that shift to OT security, allowing it to target the OT addressable market share?

[[00:53:33]]
I think IT and OT are areas where Accenture going more into OT is just adding more, I'll call it, into its portfolio of services and support. OT is, as you mentioned, industrial technology, operational technology is something that it is growing, especially as manufacturing is coming back into the west when you think in the US or in Europe, where it's less in low-cost places like China or India or all that. As you're coming back where you're leveraging tech to be able to reduce manufacturing costs and be able to make products quicker to market where, again, same thing, you're not looking at cheap labour, you're looking at technology to drive that. That introduces challenges to companies. They have to find a way to secure them. They have to find a way to run them efficiently. They have to look at the platforms that those things are built on because the technology and the operating systems are different.

They're different from our traditional PC, server-based, cloud-based type of technologies. Those are different for supporting those type of capabilities. The shift to that is something that Accenture has pivoted. Accenture had Industry X that did that, which focused on that, which is a good thing. It's getting into a market where it's leveraging Accenture skill, depth skills of its people to leverage and go into that space. I think it's a good strategy and I think it does help to provide Accenture some other avenue to increase revenue and to be able to expand its client base. That pivot, I think it's good, but you can't just pivot to OT without looking at IT. IT is still going to be the dominant. That's where majority of the business in the world still runs is on IT infrastructure. OT runs operational tech in manufacturing and all that and that's still an area to advance. That's good. From a cyber perspective, when you think about even just in our homes today, in terms of what we do, most of us have 2-3 and sometimes 20 IoT devices.

That's individuals. Think about companies. When you think about companies talking about IT infrastructure, those are edge computing capabilities that they need to be able to look at. How do you manage them? How do you secure them? Just a simple ransomware of any of those capabilities can bring a company down. That is an opportunity for Accenture Security to take advantage of or any other security consulting company that's looking at that space. When you start going to OT, now that you're trying to be efficient in these manufacturing capabilities or industrial technologies as well to be able to expand them and expand operations and make them more efficient, you're exposing those things. Those things where they were more on isolated networks now become on the internet or more bigger scale networks. They expose themselves to more risk. They expose themselves to more attacks.

Securing those systems and capabilities is also essential for those companies because they used to build that and manage them on a point in place solution tied to the device and for you to be able to hack it or for you to be able to do it and you have to be in the factory or you have to be in front of the device to be able to do something. Now, when those devices are connected to private networks, they're a little bit secure, but then again, you have to be able to be in the network. Now, when they have to leverage the internet because now you want to be able to leverage agentic capabilities, you want to expose them to learning capabilities, you want to make it solve autonomous and all that, that increases risk. I think Accenture putting more emphasis on OT security is a good pivot as well. It's something that from a business perspective, as we mentioned, the security business of Accenture has grown. The one thing that every technologist understands is security is not a secondary thought.

If you go into any consulting gig or you implement any system and your technologies and all you're thinking about is putting a system in and all that without thinking about how you secure the system, how do you mitigate risk from a cyber perspective, then you know the technology because security is supposed to be embedded in everything we do as at technologist or a technology consultant. I think having that mindset once you go into OT is needed and it helps you to be able to find opportunities to increase revenue because now you're saying, "Hey, we put in this, but we're finding this. We can help you fix it." Most of those firms don't have the skills and expertise like the Accentures and others we have within their arsenal to be able to bring in to help them with it.

[[00:57:50]]
Regarding cybersecurity, you noted that there's no linkage between FTE revenue rates and non-FTE revenue vector through platform- and software-based offerings. Do you think that should bode well in the eyes of investors as we see the continued growth of the cybersecurity business at Accenture, considering the recent acquisitions that the company has made in the space?

[[00:58:18]]
I think it will go well because as Accenture assimilates some of those things and look at how you price them, a lot of these things are all pricing. When you're doing SOC capabilities for any company, you're managing your security operations centres. If they are paying you per call or per incident, not the FTE, it's the value you get for those number of incidents that you're preventing and all that. It comes down to those type of things. If you're providing services to them based on the capabilities you acquired where on the edge, you're providing security capabilities to protect their IoT devices and all that, yes, great. What's the value you get from them and what you do? It's almost like a subscription model.

You have to be able to make sure that that subscription model compared to at the end of the day comes to two things. They're going to be people, they're going to be systems and they're going to be licences. You'll be able to price them where you have a margin that significantly covers your overhead costs for making it profitable. Revenue is not the problem. You can generate revenue from that, but is it profitable revenue? It's the question because if you price it as a service and at the end of the day is net-net zero, then yes, your revenue went up, but you didn't make any profit from it because you're still spending more on cost vs how much you want to get from it.

[[00:59:34]]
Accenture is guiding for low-to-mid-single-digit revenue growth in FY26. The company saw incremental margin improvements, with margins sitting in the 17% range right now. How are you thinking about the revenue and the top- and bottom-line profile of the business going forwards? Could it see revenue growth going back towards high-single or low-double digits, given the promises of AI? What does the margin profile look like for the business as it shifts away from the FTE model towards a non-FTE model?

[[01:00:36]]
I'll give you my bet and my bet would just be just as we talked about already, I just think Accenture carries too much legacy garbage from its past contracts, heavy on BPO and managed services. I think the current guidance, I think it's, I would say, more ambitious than I would think because I think it's likely negative growth in my view for the next 2-3 years. When you start thinking about high-single digits or low-double digits, that's in the range of 4-5 years from now. When they have shedded a significant amount of those large BPO managed service contracts, multimillion, multiyear, low cost driven by optimisation and efficiencies and all that, productivity, when they start shedding those and those are coming for renewal, that is where I think you can see a significant amount of, I'll call it, both top line and bottom line in terms of profitability. I think from a profitability perspective, Accenture carries a lot of cost. Again, I know every company does. There's a bench time. People get let go and all that stuff.

One of the things Accenture is good at right now and just think about where most other IT companies are looking at is sharing high cost, high-cost employees, whereas at the higher levels of the senior managing directors or managing directors or even senior managers at a high cost or people being on the bench for too long, I think you probably cannot afford to hold that. Again, you have to find everything to show that you're managing your overall cost. Your cost profile needs to be low compared to your revenue, so your profitability can really get to where you're saying your growth is. You can also manage that through revenue as well, if you can book more and get more revenue in. Revenue at high cost, profitability becomes low. I think from my perspective, I think the guidance right now is, I think it's optimistic. Is it achievable? Absolutely. Probably, yes, because it's still in the low-low digits or even negative. You might come from a growth perspective. I'm thinking if you're thinking about high-single digits and double digits, you're thinking about the next 4-5 years. Like I said, you don't see that Accenture ever getting to that point in the next 2-3 years.

[[01:03:03]]
Accenture's CEO Julie Sweet has led the organisation through technology iterations, exogenous market challenges and the pandemic. What we're facing today in terms of AI is interesting. It's somewhat of a double-edged sword, where it is an existential threat for companies such as Accenture and the broader industry, but at the same time, it's a technology cycle or trend that holds tremendous promise that can be realised through successful execution and re-conceptualisation of a provider's business. What are your impressions of Julie Sweet and the management team at Accenture? Do you think this current iteration of leadership is the right team to help navigate this technology cycle and the changes it has brought?

[[01:04:06]]
My personal opinion is that I think the current leadership is struggling, in my viewpoint, in two ways. I think from a Julie perspective, I think Julie did not understand. She came into Accenture as a lawyer from not having been in consulting. She didn't grow in consulting. Does she have good leadership skills to be able to run another business? Yes. To run a consulting business of this size and magnitude and to be able to pivot from a strategy perspective, if you compare her to Pierre Nanterme that was before her, they're completely two different opposite. Pierre grew up at Accenture, Pierre knew Accenture, Pierre had the values of Accenture, knew how to motivate consulting. They had to pivot at the right times, got people to be able to go to work and be able to be excited about their job. That's different from a Julie. Julie is not the same way. Again, I'm not a personal psychologist for her and I'm not saying she's not a great leader. She's a great leader for a different type of business, but I don't think it's consulting work suffice.

I think that has been the struggle for Accenture to her reign as a CEO. You can see the downward trend of the stock. Even though Accenture, yes, COVID impacted Accenture, like everybody, but if you look at during the COVID times, Accenture actually grew revenue despite the fact that Accenture let a bunch of people go and everything else when COVID started. If you look at it from that point in the rebound and all that, yes, some of the COVID money helped as we help states and start put things together, but there was some growth. It wasn't like it was just a downward trend. Post-COVID and looking at it today, if you look at the barometer being the stock market, the stock has dominated. It's not because people are not seeing that Accenture has a brand name, the scale and all that. It's because investor confidence that can Accenture actually the current leadership pivot Accenture to take advantage of the actual what is seen from a gen AI perspective, AI or how the market is pivoting, I think that's the investor sentiment. I think that's where it is.

If you look at the folks internally at Accenture, you just need to go to Fishbowl and all the other threads of your place. You just got to go see employee sentiment. Is the employee sentiment still compared to five years ago to where it is? Absolutely, my opinion is no. I think that's where it is. Again, as a leader, you don't always got to go by employees mad about some decision you make. You can make one decision and people will be mad about it. If you consistently make decisions that people are not happy about, that means you're not listening to them. I think from a Julie perspective, I just think her background as a lawyer and attorney and coming from where she is, she's gained an understanding of how consulting around in our roles at Accenture, but I think just looking at the way she's run the firm and what the firm is right now, I think she's not putting the right strategy or the correct pivot because it's one pivot to one. Accenture has gone from a cloud first. Accenture has gone to merging Song. Accenture has gone to regional model. Accenture has gone back to industry model. Now, they do reinvention services in the span of how many years.

That's in the span of less than five years. We're talking about that continuous switch and change in strategy. That means, what's the long-term focus? The reinvention services, what is that going to change? It's basically taking something people say you take a pig and wrap it in a different blanket and you call it something else. Is that going to change? No. It's how you deliver the work and how the clients actually see that you're coming to market and how you're doing. Accenture has some really great people. It's a great company, great legacy, great people to work with, have lots of friends that continue to be close friends today. It's not the point of the company. I just think the leadership direction has to be one they're looking at not just quick strategies that hit the headlines but something that actually shows investors and the people that work there because Accenture salespeople give the confidence that what you put in as a strategy is actually something that's going to get Accenture to a point where it's growing at that high-single-digits, double-digits perspective. That's my personal opinion.

[[01:08:20]]
I think that's a great opinion to end the Interview on. Let me close by saying thank you so much to for not only taking the time to speak with us today but for the wonderful inputs and insights. Thank you, clients, as always for choosing Third Bridge. We do hope you enjoyed today's content. Lastly, a big thank you to all the team members who have put this discussion together today. Goodbye.

Transcription ends at 01:08:36 of the recorded material