All right. Welcome to the Citi TMT Conference, and as a reminder, this session is for Citi clients only. I'm Ryan Potter from Citi Research. I'm glad to be hosting Genpact for this session, and from Genpact, we have CFO Mike Weiner. Mike, thanks for doing this.
Yeah, thanks for having us.
I guess, just as a start, for new investors, can you kind of give them a idea of a short elevator pitch of who Genpact is, who competition is? Like, what leads you guys to win in the market or your competitive advantages?
Sure, sure. Thanks for having us here at the conference. I apologize for being a little hoarse today. We had a number of meetings.
No worries.
Let's talk a little bit about what Genpact is. Genpact originated from GE, right? It was, in many cases, the middle and back office operations, first from GE Capital and then from GE in total. And for many, many years, it was its only client prior to it going public. Still, today, GE is still a client in its various forms of it, but the company has grown dramatically and diversified itself away from GE. We're a little over $4.5 billion of revenue, and the interesting thing is about 75% of our revenue we consider annuitized or repeats year after year, as we do large managed services contracts. We have deep experience in business services and being process-level optimizers in terms of driving value on behalf of our clients.
We have strong partnerships and disproportionately focus our efforts on the Fortune 500 companies. Today, our revenue and really our go-to-market offerings are bifurcated between, we consider our data tech and AI offerings, which about half of those are annuitized, about 45%, and 55% of our business is digital operations, which ostensibly has its roots in what was legacy BPO, with a large focus on using advanced technologies to drive substantial amount of productivity on behalf of our clients. We compete, which is in a unique space, we compete against very large, IT players, right? We're not a big systems integrator, right? We're a, you know, business management and operations company. We have deep domain experience, particularly in data tech and AI, and we offer comprehensive solutions that really drive value on our clients.
We have large service lines, including that of finance and accounting, and arguably, we think about ourselves as either the number one or number two player, as well as in supply chain and procurement offerings to drive that on behalf of it. Our clientele are very large multinational companies, as I said, Fortune 500, and we disproportionately serve the CFO as well as the CIO, in terms of what these offerings are.
Got it. Perfect overview. I guess just shifting gears a little bit, since BK came on as CEO, there's been a focus on execution, and he has his own 3+1 execution framework, includes things like partnerships, Data-Tech-AI, simplification, Client Zero. I guess could you provide further detail in terms of what elements are in each of those and could you highlight progress maybe you've made towards each element of that 3+1 framework?
So we've made progress in every one of those four things that are part of three plus one. So let me explain what it is. It's an execution framework for the company, right? So the first one is partnerships, and in terms of partnerships, what we're doing is we are partnering with, for lack of a better term, large hyperscalers, the likes of the Microsoft, AWS, Google, as well as partners such as ServiceNow and Databricks and Salesforce, to go to market with solutions that solve problems on behalf of our clients, right? We're already seeing strong progress in terms of doing that and driving revenue as well as offerings. The interesting thing is that the industry as a whole has a much higher percentage than we do in terms of what revenue they generate from these partnerships.
Now, you know, we're in the single digits. Others in the industry arguably are 30%, 40%, 50%. That being said, we're not a large systems integrator, but we think there's a huge opportunity for us to have deeper penetration by with these partnerships. So what have we done to date, as you alluded to, in BK's eight or so months of being CEO? We have a new leader in the space who's joined the company, as well as a team that continues to expand and grow, and we've had a tremendous amount of interest and partnership with these large hyperscalers.
The way I'd like to think about it is we disproportionately sit between the client and these tech companies in terms of taking, in many cases, they're on the shelf, very complicated solutions and clients' problems, and trying to adapt them to solve certain things, right? To believe in this model that the industry operates in, in ostensibly is that these large hyperscalers are not in the implementation businesses, right? Or the consulting type businesses. They wanna sell SaaS-based revenue, right? Sell as many of their products, but we need to be there to customize those solutions using our domain-level expertise to solve those problems. So that's partnerships as part of, you know, the three plus one.
The next is data tech and AI, which is to no surprise to anybody. It's really leveraging the power of AI to effectively drive solutions for our clients. We've already seen tremendous impact in expanding our reach to our clients. We've had a tenfold increase in bookings associated with generative AI, and we'll probably talk about it later. AI and generative AI is nothing new to Genpact. We're probably not as known for it as we should be, and early signs are very positive for us to continue the pivot towards becoming a generative AI leader in the space. The next is kind of a simple one, and we call it simplification, right? How do we make ourselves more effectively internally with our go-to-market strategy, right? To be more client-focused.
Today, we have 12 business units that are specifically focused on the 12 areas which we focus on with our clients, everything ranging from high tech and manufacturing to consumer goods and life sciences, and we've organized ourselves there, all centered around meeting the needs of the clients. And then the last one, which is one disproportionately as a functional leader in the organization I'm focused on, is that of what we call Client Zero, right? We need to be our best reference, right? So as we talk about adopting these new technologies to drive things such as finance and accounting to a CFO, we need to show them proof points that we're doing it ourselves. And then the benefit of that is gonna be twofold, right?
Will obviously be that test case, for us to go to market, as well as we think it's gonna drive a lot of productivity, expense savings for us in terms of our organization, which we could plow back into offerings.
Got it. Maybe double-clicking on the simplification element and moving to 12 business units. You mentioned being closer to the client, but was there any other rationale behind making this move? What benefits do you expect to see under 12 business units? And have you started to see proof points of that rationale actually coming to fruition?
Absolutely. We have seen our execution, and I would, you know, attribute our performance to better than our own internal expectations for the first half of the year, a lot towards that. So what does it allow us to do? It's not just re-segmenting our business. I don't want people to have that view at all. It was creating these 12 business units to create clear accountability, primarily on sales and client execution, to drive it. And in that clear accountability is really focused on not just historically on potentially, pipeline and bookings, but that through pipelines to bookings, to revenue, through gross margin in many cases, as well as really the quality of revenue that's associated with just not just digital operations, but data tech and AI, and then now pivoting to the, to generative AI-related things.
The ultimate focus of it is exactly that, a focus. It's enhancing our focus with our clients and providing them with the deep level of experience that we have internally to help them with their needs. So, traditionally, we had gone to market with a much more broader sales force level approach, having sales folks associated with it, without the robust offerings that we have today by being solely focused on these 12 units.
Got it. And your earnings results past few quarters, especially in 2024, have been relatively solid. How much of this would you attribute to this three plus one execution framework? So execution improving versus being more prudent or cautious or conservative in your outlook, or some combination of both. And then also on top of that, I think the outlook raises have been mostly just the outperformance in quarter. You've been flowing that through. Is there any reason you haven't flown that outperformance into more of the second half of the year into future quarters?
Yeah. So if you think about our performance the first half of the year, it's really a combination of the two things that we alluded to. It's improved execution that we've had in our business, all tied to our three plus one strategy, right? We have a prudent guidance philosophy that we continue to move through the organization, right? And we've flown through those results through the second half of the year, right, and maybe a little additional to that. Now, what the ultimate variable in terms of our performance for the second half of this year, it's really gonna be on what ultimately happens with the general business environment that we operate under. We have assumed a relatively benign environment that really started a year ago, right, as the third quarter pivoted to much more, flatter, benign.
We saw that certainly through the fourth quarter and the first two quarters of this year. So we're not anticipating anything dramatically improving or deteriorating for the remainder part of the year. We'll see ultimately where it goes.
Got it. I guess maybe shifting gears to the demand environment, how would you broadly characterize the current state of the demand environment, where you're seeing, I guess, maybe since the end of the quarter? And then given this enterprise spending environment, where enterprise is more focused on costs, has that twofold led to increase in core BPO demand, but also has it led to clients maybe perhaps pushing back for more kind of productivity savings?
It's a combination of a lot of things, right? So if you'd asked me this question a month ago, I would say really two things are affecting the buying behavior of our clients in terms of a cloudy business environment. And if you take these two large macro things, you'd bucket them into two things. One is the general financial and business environment, AKA interest rate environment. And then the second is the macro global environment. So let's do, you know, double-click on each one of those. I think we're all relatively comfortable, particularly our clients, that they have line of sight in where the interest rate environment's going to be, which helps them in terms of CapEx, helps them in terms of really helping them understand where their demand is gonna come from, both on a B2B and B2C side, right?
The other component of it, why we haven't seen dramatic acceleration in spending, is that we hear from our clients we need to understand really, for lack of a better term, where the global environment's gonna be. And if you really talk to them about it, what they're saying is the elections, right? So we'll ultimately see what happens in November, right? But that still is an overhanging factor on our clients, as particularly they move into their budgeting cycles that we're seeing right now. So we're assuming that the environment, particularly in our, as I talked about a minute or so ago, doesn't improve dramatically from their buying behavior or deteriorate. And we'll see what happens.
If we perhaps saw certainty on interest rate reductions, got through election cycle, maybe we're entering next year in a more positive state?
I think so. I think so. I think we've taken a prudent approach for this year. We'll give guidance as we complete the next year, and the interesting part of our business, what I said in my opening remarks, is that 75% of our business is annuitized. So we're, you know, come January first as CFO, and it makes me feel great about the business. The other 25% of our business, we have to go out, and we've converted those bookings into revenue for that period of time, and a lot of that's really all within our data tech and AI business. Some of that work is discretionary, short term. You know, that can be anywhere, you know, from a year or less, and I think if we have a healthier business environment, we hope to see an acceleration of that....
Compounded by the work that we're doing and the investments we're making in both partnerships and data tech and AI.
Got it. And you mentioned earlier that maybe Genpact's not as well known for AI capabilities and investor base, so here's your chance to kind of talk up. What are some of your specialties, Genpact specialties towards AI and ML analytics? Maybe you could talk through some work you've done recently, use cases.
Yeah, so this is a really interesting one. You know, so I joined the company a little over three years ago, right? And when I was meeting with my predecessor, and we were going over the productivity commitments that we make to our clients, I said, "Oh, my gosh, how are we able to do this," right? And these are not small companies that we're going after. And so you have to believe they have inherent internal capabilities in terms of running their operations efficiently, right? And it isn't just geographical arbitrage of where your staff is located. It's our process domination that we know how to run processes using inherent technologies in doing it. So we've been using AI for years, right?
We did an acquisition many years ago called Rage Frameworks, which has helped us implement AI, not generative AI, well, it's relatively new, in November of 2022, into our operations. So you know, a lot of people don't think about Genpact. If you think about 45% of our business is in data tech and AI, and using enhanced AI in terms of the analytical work that we do. So I think we're really well-positioned as we move into the adoption of generative AI into our service offerings, right? So if you think about generative AI, it's just that. It's generating, in many ways, context or answering questions that is gonna be either verbal, textual, or visual, right?
And we think we'll be in a situation to take that and disintermediate some of the offerings that we have and also help drive productivity. But that's all easier said than done, right? So I have a few facts, right? We talked, BK talked about it, is that we've taken this foundational view on generative AI, and what we've said is that we will double invest on it from our perspective in terms of training the whole organization, that we have 130,000 team members in becoming, you know, Gen AI enabled in what we do. So we've done a lot of work, right? We have 100,000 employees that we're actively learning in the space. 70,000 of them have completed level training, 18,000 more advanced work that we've done.
And this doesn't just stop at that. You know, they sent me back to school, just completed my MIT applied AI-related classes. I'll be doing much more. But it's important for us to signal to our clients and to our organization as a whole that this is the way forward, and we will be a leader in this space, both in the short, medium, and long term.
Got it. And in terms of Gen AI in particular, have you seen client work move beyond the experimentation, proof of concept, pilot stages into more production stages yet?
Yeah, so we've done quite a bit of work in terms of proof of concepts, right? We have done some production-level work or institutionalizing generative AI solutions. What I would say is as follows: The work that we've done in proof of concepts using large language models has produced phenomenal results, right? Where, us and others in the industry, I wouldn't say struggle, but look for additional adoption of it is really two things, right? It is still costly to implement these models, right? And the business case, while others will underwrite, some will not. Fair enough. But fundamentally, a lot is done, the unstructured data to implement this on a wholesale basis in many of our clients isn't there.
That is what we consider an existential opportunity for us to help them through our partnership channel, but also to working with them directly, to organize that data in a meaningful way so that we can implement these technologies for them on a go-forward basis. I think there's a grave misunderstanding about how structured the data is, particularly in our large multinational companies that we deal with, right? Born in the cloud companies are completely different, but when you're dealing with, you know, some down-level component companies that we have, their data is very much unstructured, and so we think that's a big opportunity for us.
Got it. Would you say most client work, like, new client work today involves either some type of data management or structuring component or Gen AI component, or is that still starting to slowly flow through the pipeline?
I think there isn't work that we do today, that it does not involve something within our data tech and AI unit, right? People don't come to us to buy digital operations. People don't come to us for human capital arbitrage, right? It all has to evolve around process utilization of technology, right? And that's why we think it's gonna be a huge beneficiary of this. You know, we think of Gen AI and technology as a whole, as a TAM enhancer for us, right? Because in our discussions, particularly with the hyperscalers, that domain-level, keystroke-level information and understanding that we have is key to driving that productivity and that for our clients, and we think we'll be paid for that.
Got it. And I guess early days of Gen AI in early 2023, investors were quick to put BPO in the loser bucket. Is there any way you kind of frame, kind of near term versus medium term versus long term? You mentioned TAM enhancer, but is it... It will not be an even kind of path, I guess, to enhance the TAM from Gen AI.
So yeah. So fundamentally, as I said, it's a TAM enhancer for us. You have to believe that we don't continue to operate the company within a silo. As I alluded to earlier, with some of this, the metrics that I provided with our investment in terms of people training, upskilling, partnerships, so on and so on, that we will continue to pivot to become a leader within the space. So right now, we're in the short term, or short duration work that we've done, and that's really been the proof of concepts, right? Taking the experience that we have with the teams that we have, creating large language models to solve various problems on behalf of clients.
We've had, now it's coming down to institutionalizing them, and the cost we talked about, as well as data security, data cleansing, availability of structured data, has been somewhat of a limiting factor to doing that, right? But we're getting there. I don't think that it is going to stop. I think it will increase at a more rapid pace. You know, then we think about the medium-term adoption of this, and I think about it as, you know, thinking about it from my MIT class perspective, it's human in the loop, right? And then ultimately, generative AI and AI-related technologies will be involved in everything that we do on a long-term basis, right? The question I don't have an answer to is: how long will this take?
What is short, medium, and long term, right? I don't know. I don't think it is months. I think this is a multiple year, you know, process that these companies will go through. I think of cloud migration, right, and how off people are in terms of how long it has taken to get people there, but one thing I can assure you of is that Genpact will not miss the waves that potentially it has missed in the past with either cloud or cybersecurity or other big tech waves. We think this one is real sustainable, and we have a competitive advantage in doing it, all based on that domain-level experience.
Yeah, we also like to point to investors. You mentioned cloud, but also RPA, in terms of disruption fears, and us being the largest distributors of the technology.
Yeah, I smirk when you say that. I, you know, dating back to RPA, the fear of it was just dramatically overblown, right? Not just overblown, it was a net TAM enhancer at that period of time, right? How could we use these technologies to solve problems on behalf of our clients? It all fundamentally gets back to the view that, what do our clients want to do and focus their time and efforts? There's only so many hours in the day. So the way I like to think about it is, our clients want to be in the business of designing and implementing products and services on behalf of their clients and marketing those, right?
If you think about it from our perspective, they want to allocate more time to that and a heck of a lot less time, capital, and resources to the middle back office and to the back office operations that they have, right? I think one thing that's happened, you know, quite nicely, at a benefit, which there were very few of, for COVID, is people's view of remote work, letting work being done halfway around the world. Now with generative AI, in terms of language translation on it, I think people are a lot more comfortable in doing that in an outsource-provided world.
The one I index back to, if you were to tell me five years ago that companies, now, this is not that we don't really operate in this space, will be doing outsourcing of tax work, I would have said, "You're crazy." But that has evolved. So it all supports the hypothesis I alluded to in terms of what our clients really want to focus on. And we see demand still rising. We're at, you know, record level pipeline, and now it's, you know, incumbent upon us to convert that pipeline into bookings and into revenue at substantially higher margins where we are today.
Are there parts of the business to think through or maybe is a little bit more risk in the near term from Gen AI, where maybe cannibalization needs to happen first before you get the TAM expansion?
Listen, I think if you think about where this technology will have its earliest impact, right? It's gonna be in terms of application development work that we do, right? And we think that's. We're not big in that space, and we think we'll actually use that as a productivity tool ourselves, right? CX work and customer service work, we're just not that big in that space. But what we're looking to do, we'd love to be able to cannibalize that space using technology, right? And be able to institutionalize that on a go-forward basis. And then ultimately, where it has a huge amount of benefit now is in content creation, right? And we do do some work with that, and that's gonna allow for enhanced democratization of some of the work that our clients do, right?
We have clients that's, you know, you know, four or five years ago, pre-COVID, that would demand that, you know, certain work being done in certain geos, and certain, you know, ringed-in areas, were required. We're able to now use this technology to drive productivity on those, on those functions or and those tasks quite dramatically by doing work, in many cases, halfway around the world at a fraction of the cost.
In driving productivity, I guess one of the tangential impacts of this is potentially clients may be more willing to adopt more non-FTE commercial models. So I guess just in terms of color on that front, where is Genpact in terms of non-FTE? Is that something you're pushing and clients are also willing to accept more? Because, I mean, throughout history, clients have been wanting to keep all the savings, not wanting to offload any of that.
Yeah, I think it's a few things, right? The answer is yes, yes, yes. You know, the non-FTE-related pricing models, we historically, and as you know, have got a big pushback on. It makes it challenging to compare and contrast offerings and costs among other carriers or other providers. It also makes it more challenging to compare against your internal benchmarks of it, right? So where we have had success thus far is having an FTE-related pricing model upon renewal, converting that into an alternative commercial model that allows us to have that decoupling, right? That decoupling has resulted in, thus far, about 20% of our revenue within that cohort of alternative commercial models, and that excludes our fixed-fee contracts that we have.
The margin on it has been higher from that, and I think it is gonna be required for as we and the others in the industry move towards these generative AI-related offerings, to have that FTE completely being decoupled. You know, and others, including intermediaries in the space, which a lot of people don't talk about, right? In many cases, clients just don't interact with us directly. There's a whole host of intermediaries in this space that act, in many cases, as brokers and advisors, are getting there in terms of advising their clients. You know, last year we did a large booking on, like, a large deal. It's actually revenue that we've done on a SaaS and, you know, non-FTE alternative commercial model basis. So I think we're seeing traction in doing that.
Got it. And have the clients' level of productivity givebacks that they deem as normal change at all with AI or the macro in terms of enterprises being more focused on costs? Like, are clients just assuming on renewals there needs to be more productivity givebacks given all this technology?
So it's hard to put your finger on it, right? Clients always want... And as a buyer of services, I always wanna pay less and get more, right? So we continue to see that, and others in the space have continued to see that. We have not seen any irrational behavior, irrational pricing within the industry as a whole, but it is incumbent upon us to make to do that delivery and to meet those productivity commitments, you know, being on the forefront of utilizing technologies to drive that level. We just can't. It has moved far from just having enough FTEs in low-cost locations, utilizing, in many cases, rudimentary technology or even technologies of the not-too-distant RPA past, to drive that. So, we think we're well-positioned, to evolve that and compete within that space.
We'll ultimately see what happens, but we feel really good about our forecasts for this year and our prospects for many years to come, from a wealth of growth and a margin perspective.
Got it. And this is meant to be interactive, so I'll stop there and see if there's any questions in the crowd. No, I'll keep going. All right, you touched on the opportunity in strategic partnerships in terms of moving that mixed revenue more towards that. I guess, what are you doing on that front to push that? Is it more you're trying to move up partner tiers with all these, the hyperscalers and even RPA providers out there or other software providers, or is there anything else you're doing internally to push teams more towards partnerships?
So yes and yes, right. So we're working aggressively to display and move up within the tiers. And we've had, you know, quite frankly, really nice reception amongst particularly hyperscalers, as well as those other companies that I mentioned earlier. I hate the term hyperscalers. And I think what's so unique about it is why, right? And they're impressed with two things. A, our client list of who those clients are, but our competitors have, you know, brand-name clients, right? But particularly in finance and accounting, supply chain, and procurement, they are really surprised at. We use the term internally, the domain-level experience that we have. I like to think about it as keystroke-level work that we have.
So to drive enhancements in productivity, you have to know where the opportunities lie, where things break, where today there's a human exception in terms of the process, right? Because that has impressed, in my opinion, these hyperscalers to allow us to co-invest in going after clients to utilize that. And that can be everything from copilots to a whole host of other technologies.
Got it. I guess maybe shifting and focusing on data tech and AI in particular, can you comment on where you're seeing most demand, most pressure in this environment? And I guess, historically, what have been some of the faster-growing areas in there? Maybe you could talk about some of your more historical growth rates before you re-segment in terms of supply chain, sales and commercial, financial risk and crime, stuff like that.
Yeah. So if we think about the business as a whole, right? Our data tech and AI business, which has a huge analytics business, a huge data business in it, and also has a customer experience business in it, right? So where that business has done very well has been in the consulting side, it's been in the data, and also in the AI side, but it's small. Where it hasn't grown as much as we would like based on the past, has really been our customers now leaning into growth-focused initiatives and disproportionately focused to the cost take-out agenda, right? So if you think about just the not-too-distant future of 2001 and 2002, and to 2021 and 2022, right?
People had that grow at any, you know, mindset mantra in their organization, and we did a lot of work supporting them in sales growth, operation work that they do. They pulled back from some of that work in terms of everything from digital lead monetization, content creation work. We still do it, right? But, you know, I'm cautiously optimistic that that cycle will improve in the not-too-distant future and will be on fire on all cylinders. But in the absence of it, we'll still continue to invest dramatically and grow our offerings, particularly in data and also our AI offerings that we alluded to earlier. Everything has to be a packaged solution to drive large things. I mean, so at the end of the day, we think about our businesses.
We go into clients, and we solve problems. We design solutions. We blueprint those solutions. We'll implement technologies to effectuate that change. We hope we would also effectuate, and we say, "We'll do all that, and now let us try to run it for you at a lower cost than you can do yourself," so I think the future is quite bright from all of those perspectives.
And then shifting to Digital Operations, where you're seeing the most pressures, most strength. Is there any opportunity to potentially accelerate that growth going forward, especially given this cost-focused demand environment?
We've seen, and that's reflected in our robust pipeline. We've continued to see that business demand really remain consistently strong. Longer lead cycles, large deals, right? You know, there's a great line that one of our competitors had, and I have it on my desk. I won't say the name of the competitor, but he had said, "The CFO's back at the table." What he was alluding to was that buying behavior in the 2021, 2022 vintage was, you know, the sales leader or the CIO was really dictating a lot of buying behavior, and that cost takeout agenda really came back strong. And we have continuously seen that in terms of client demand, right?
And then you think about it. Still today, I was at a CFO event this weekend, and, you know, everyone talks about what they do. There isn't many companies who don't have a blueprint to go forward to utilize some of the services that we have, right?
Mm-hmm.
What was the exception of using a BPO provider if you weren't a large multi-cap company or had your own cap, that is now the norm. So we feel really good about that, but it's still massively under-penetrated. So some are using it for certain functions-
Mm-hmm.
but not others, right?
Mm-hmm.
And we continue to sell into that, and that continues to expand our offerings and our TAM.
Got it, and a theme across BPO in recent years has been on, captive carve-outs a little bit.
Yeah.
I know last year you commented a little bit more on some of the large deals having rebadge elements, a little bit different. In terms of some of the large deals you're seeing today, is rebadge still an element of that?
Very much so. So we're seeing two things, right? We are seeing rebadge, and I would call it rebadge enhancements, right? So a lot of the clients we operate have various... either have multiple vendors, right, BPO vendors, that they want to consolidate, or they have captives, which they've established themselves, that do pieces of the business. We do a lot of work in helping companies establish captives, putting in best practices for it. But at the end of the day, you know, I think and we continuously get captives coming back to us, for us to move them-
Mm-hmm.
for us, in the sense of you need such scale and efficiencies in a captive, in a particular business unit to drive that value that only the industry itself can generate. And, you know, we feel good about it. In addition to it is the retention quality of talent, right? You know, I use the term in many of these meetings is, you know, people don't want to be an accountant working in an accounting firm. They want to be an accountant working on behalf of clients, and we offer that, you know, us and others in the industry, that these people are the front office in what we do, and that's meaningful in terms of the quality of the talent that we have. We're a human capital-intensive business, and we need smart people who want to work here.
Not saying you don't get that in a captive, but it disproportionately, we're able to attract and retain that talent better than others.
Got it, and I guess maybe shifting gears a little bit to margins. We got the CFO. Can you talk about, I guess, 2024, where some of the margin levers, puts and takes, to think through what's driving margins in 2024? And then looking forward, I know, like, what was said at the investor day a couple years ago is maybe not as relevant, but you talked about maybe accelerating margin opportunity over the medium term. Is that still on the table? And I guess, how should we think about the margin expansion and opportunity moving beyond this year?
Yeah. So we'll talk about this year, right? So in terms of what our guide is this year, you know, to the extent that we continue to deliver better than anticipated margins, where we have needs to reinvest that money back in the business, we'll continue to do that all within the parameter, within the three plus one execution framework, disproportionately in terms of what we need to do to enhance our generative AI capabilities and offerings, right? I think and as far as taking that off the table, no, but we'll talk about our guide in the preceding quarters for next year. But we're not thinking about taking off margin expansion on the table, or more importantly, we're not thinking about taking margins back.
Got it. In the past, you've talked about moving more from Tier One cities into Tier Two cities in some of your geographies. Does that come with higher mar-- Is that still a focus for us, and does that come with also higher margins?
By definition, it comes with higher margins. If you couple that, particularly that with non-FTE-related pricing models, and you think about some of the offerings we do in a factory type, you know, conceptual setting, sure to our ability to generate higher returns will inure to us, right? But again, you know, these are open discussions that we have with our clients from that perspective, but that is still something very much we're focused on. But at the end of the day, it's not just about the cost of that talent, it's about the quality of that talent and the accessibility of the talent. Particularly in Tier Two and Tier Three cities, you know, in many places such as India, I think the talent base is massively underestimated of what their skill set is, and what they're able to do.
And if you then layer on top generative AI-related and the ability to democratize education, skills training, and now language, wow, we really have a leading edge in being able to tap that talent base. You know, we were one of the predominant companies and still are, within places like India, in terms of our ability to recruit and retain, and that's really been the secret sauce of the company every year. I often get the question: "You know, you guys have a 25% or a 20% retention or attrition rate. You have 130,000 employees. Are you hiring and training and moving through your pipeline that many people?" And the answer is yes. We do it very, very effectively.
We manage a very effective pyramid through our organization, and as we grow, we look to decouple that revenue, that FTE growth within our revenue by using different technologies to really focus on that decoupling.
... Got it. And in terms of investments, I would assume most of your investment focus is geared towards the three plus one execution framework, but just broadly, where are you making the most incremental investments?
Yeah, so we have a much more disciplined approach under BK's financial management stewardship of the organization to look at the ROI of our investments that we do and triage them accordingly, right? We just can't do everything that everybody wants. They all have to fit within the three plus one framework, that being partnerships. And we've hired and brought on a whole bunch of people, data, tech, and AI, both from a leadership and from a training perspective. Client zero has to show both a return for us internally and, you know, making us that proof point, eat our own lunch. And simplification, while it doesn't have a tremendous amount of investments to it, it is a focus of ours.
So I would disproportionately say it's gonna be on our partnership investments, as well as our data tech and AI, where the vast majority of our investment focus has been.
Got it. And just generally, on capital allocation, could you walk through Genpact's strategy? And I guess M&A, you haven't been as active in M&A as maybe you've done in the past. So how is that pipeline looking?
Sure. So what we've done, certainly under my stewardship as the CFO of the company, is put together a fairly simplistic capital allocation strategy. The wonderful thing about the company is that, you know, not just that it's got 75% of annuity revenue, it's generating a tremendous amount of cash. Our cash flow forecast for this year, free cash flow from operations, about $525 million. We have said that we will return half of our free cash flow from operations through a combination of dividends and share buybacks. And to the extent... And the other remaining piece will be used primarily to fund inorganic growth of the company.
And to the extent we don't use that to fund M&A, and I'll address M&A in a second, we'll return that additional capital to the extent we can in a most tax-efficient way to our investors. Last year, we did a little over 65% return of that free cash flow. As far as M&A, we see every deal out there. We are focused wholeheartedly on acquiring capabilities that we don't have. We don't buy businesses or buy revenue that we only use just to synergize it. So we will continue to seek out predominantly tech-related offerings associated with generative AI inorganically. And we have more to come on that in the future.
Got it. We're running up on time, so I guess I'll just end with a what keeps you up at night question in terms of focus on execution. What are some of the biggest risks that you've thought through, and how you plan on mitigating through them? And also, just in terms of macro uncertainty still out there, how are you thinking through that?
Yeah, so black swan events always keep us up at night, right? But we mitigate them to the best extent possible. But what keeps me up at night is that we have to just continue to focus on our internal execution and our execution strategy and keep the 130,000 team members that we have engaged every single day and be client-focused.
Perfect. All right. Thanks again, Mike.
No, thank you, and thank you, Citi, for having us today.