Globant S.A. (GLOB) Earnings Call Transcript
February 26, 2026
Earnings Call Speaker Segments
Good afternoon, everyone, and thanks for joining us for Globant's conference call on the results of the fourth quarter of 202. This is Arturo Langa speaking — I handle Investor Relations here at Globant. <<operator_script>> I'd like to point out that this session is being recorded and is streaming live on YouTube. You ought to have gotten the earnings release by this point. Should that not be the case, you can find one on our site, investors.globant.com. To kick things off, we'll hear remarks from Martin Migoya, our Chief Executive Officer; Diego Tartara, our Chief Technology Officer; and Juan Urthiague, our Chief Financial Officer. After that comes a Q&A, and our Chief Revenue Officer, Fernando Matzkin, will join them for it. Before we get started, let me remind everyone that certain comments made on today's call might count as forward-looking statements. That covers our business and financial outlook as well as responses to a number of your questions. Statements of that kind carry the risks and uncertainties laid out in the company's earnings release and its other filings with the SEC. Keep in mind that our financial statements follow IFRS accounting rules. On today's call, we'll also report non-IFRS, or adjusted, measures — the way we track performance internally, and the simplest method for comparing Globant against others in the industry. A reconciliation between IFRS and non-IFRS measures appears at the end of the press release we put out on our Investor Relations website announcing the results for this quarter. And now, I'll hand the call over to Martin Migoya.
Hi, everyone, and thanks for joining us again. For two decades, Globant has assisted the top companies around the globe in creating and overhauling their technology, building strong engineering skills, genuine sector knowledge, and lasting relationships with clients throughout that time. That remains our bedrock, and it always will be. During the last 12 months, we've been layering something new on top of that. What I'd like to discuss now is how that additional layer is already reshaping where we're headed. Large organizations are shifting away from trying out AI and toward putting it into action. Following a stretch of heavy spending that yielded modest results, our customers are making decisions more decisively; they grasp what AI can offer, and they want allies capable of producing tangible results — not merely tests, but solutions ready for production, crafted with an understanding of their sector, their infrastructure, and how their operations work; that's precisely what we've created through our AI pods, which operate atop our AI studios tailored to specific industries; and that's the reason we see Globant as the AI native technology solutions company. Clients are picking this partner to bridge the distance between what they invest in AI and what it actually delivers; we introduced our AI Pods 9 months back, and it's already demonstrating genuine traction among our customers. During 2025, we hit our largest revenue ever along with our best free cash flow generation, all while reorganizing our delivery organization and overhauling how we deliver. In Q4, we recorded the biggest quarterly bookings of the year, climbing 32.4% compared with the same period a year earlier. Our pipeline stays strong, sitting at $3.4 billion. I'd like to use this call to take you through our results, our approach, and the particular metrics showing why we feel confident about what lies ahead. The IT professional services sector is undergoing a fundamental transformation. Capital for technology is pouring predominantly into AI infrastructure, and Gartner forecasts IT services will expand only 4.4% in 2026 — under half the pace of total IT spending. Yet the big 4 hyperscalers are nearing $700 billion in combined 2026 CapEx, almost 3 times what it was merely 2 years prior. The magnitude of that spending is remarkable, though it has also generated an enormous implementation gap; during 2025, MIT research revealed that the majority of enterprise AI pilots hadn't yet produced measurable P&L impact, and a considerable portion of firms halted or reorganized their AI efforts over the last year. Meanwhile, according to HFS Research, the accumulated technical debt across the Forbes Global 2000 sits somewhere in the range of $1.5 trillion to $2 trillion. On top of that, Forrester's data shows U.S. customer experience quality has hit an all-time low, after declining for 4 consecutive years. The takeaway here is not that AI is falling apart — what it actually means is that we've entered the execution phase for this sector. Following an 18-month stretch of experimentation, enterprises have developed a clear understanding of what AI can do for their operations, and they're actively seeking the means to roll it out at scale. That shift from exploration into execution is precisely what's driving our record bookings right now. We're in the middle of a generational transition. Think back to the moment AWS showed up. It wasn't just about cheaper servers—it created a whole new sector. Cloud-native companies, today's SaaS, the entire startup world that sprang up over the past 15 years—none of that existed until AWS made elastic, easy-to-reach infrastructure real. That's exactly where technology services stand right now. AI native delivery, with intelligent agents supervised by domain experts and powered by a token subscription model, isn't a better way of handling what we already handle. It's the foundation for a sector that hasn't fully come into being yet. Globant was the first company to establish what AI native technology services actually are, and 2026 is the year the market begins confirming that wager. Our foundational business — extensive software engineering, digital transformation, and domain knowledge built up across 2 decades — remains firmly in place. For many years to come, enterprises will continue requiring that capability, and we will continue expanding it. What we're doing at present is adding something fresh and potent on top of that base, an AI native offering that grows in step with the AI opportunity itself. Previously, a company's digital products functioned as its moat, and creating differentiated software required hundreds of elite engineers plus hundreds of millions of dollars. AI has sped up and simplified the process of creating software, which actually fuels greater demand across the entire industry. When every organization is able to produce software with greater efficiency, the differentiator is no longer your ability to build — it becomes the volume you produce, the pace at which you iterate, and how consistently you keep evolving. We're heading into a period defined by a massive surge in software creation alongside far quicker competitive cycles. Our substantial engineering depth and 20 years of domain expertise, now amplified by AI, put us in an ideal position to serve that demand. With that as our context, we identify 4 distinct and expanding sources of demand. The first is Agentic Workflow Orchestration. Companies require AI agents that operate independently and work together across intricate environments. They don't want isolated fixes — they want complete processes that genuinely advance how the business runs. Next comes updating foundational systems at the speed AI makes possible. Among the Global 2000, technical debt has piled up to between $1.5 trillion and $2 trillion, weighing heavily on their ability to innovate. By delivering in an AI native way, we can chip away at that accumulated backlog far faster than anyone once believed achievable, giving our clients the operational flexibility they must have to stay competitive and succeed. Last, bespoke software is winning back territory from SaaS. For a long time, SaaS was the go-to option whenever an enterprise needed software. Now, delivering in an AI native way broadens the scope of what companies can affordably create, which makes deeply tailored software realistic for scenarios where previously only packaged platforms made sense. The goal isn't to do away with SaaS. It's to give enterprises greater choice, plus more say over their data, their workflows, and what sets them apart from competitors. SaaS and custom software are becoming more and more complementary, and we're in a unique spot to provide both. Fourth, AI governance and corporate sovereignty. As enterprises roll out agents from a range of vendors across departments, data gets scattered and control slips away. To keep everything in check and make sure every interaction remains under their control, they need an orchestration partner they can rely on. Central to this approach are our alliances with NVIDIA, OpenAI, AWS, Salesforce, SAP, Oracle, Microsoft, Google, Adobe, and others. We act as the AI native orchestration layer that produces results for our clients. AI pods function as service units that AI powers, tailored to specific tasks and industries. Through the software inside these pods, technology gets built and pushed forward. AI agent workflows, overseen by specialists at Globant, generate working software artifacts under a token subscription approach. AIPodOps brings automation to business processes once they're live in production, and every token used adds to a growing store of institutional knowledge. The customer holds ownership of everything — there are no seats involved, just usage. Unlike conventional approaches, our AI Pods operate on a capacity model built around subscriptions. What clients sign up for is a dedicated tier of orchestrated output, capped at a set level of token consumption. Our proprietary platform, Globant Enterprise AI, serves as the delivery engine driving both of those, and it consists of four interconnected hubs. There's the enterprise hub, which connects securely to every corporate system; the AI hub, which intelligently routes across more than 140 LLMs while keeping full data sovereignty intact; the agent hub, where we create and publish industry-specific agents that encode 20 years of domain expertise; and the AI Pods hub, where clients subscribe and scale up. I want to be completely clear on one point: this platform wasn't built overnight.
For years now, we've been putting money into Globant Enterprise AI—creating an actual product, genuine orchestration infrastructure, plus real security and compliance architecture. That spending sits inside our operating expenses and shows up in the EBIT margin we have right now. Put differently, the margin picture you're looking at today already includes what it costs to build a proprietary AI platform. Twelve months back, AI Pods revenue stood at 0; during 2025, we got to an exit rate ARR of $20.6 million, with gross margins ranging from 45% to 60%, against a blended gross margin of 38% for us. This isn't some trial. It's an actual business. When it comes to 2026, our target sits between $60 million and $100 million for AI Pods exit rate ARR. Beyond that, we anticipate that margin profile getting even better as the subscription model grows and cost per token keeps falling. That marks a basic change to the DNA of our structural profitability. As AI pods grow as a portion of revenue, they should widen our overall margin profile. Our AI Pods pipeline hit $283 million during Q4, a 34% increase versus Q3, and it now makes up 8% of the whole pipeline, whereas in Q2 it was only 3%. More than 60 AI Pods are running for clients around the world, and 24 new subscription offerings were closed in the last quarter by itself. A number of our top 10 clients have finished demanding security and procurement approvals and are currently running AI pods on the platform. That pipeline is converting. Revenue is coming through, and we've only just begun. Given the record bookings we're reporting today, the quickening adoption of AI Pods throughout our client base, and pipeline conversion trends that keep improving, we have a clear line of sight to getting back to positive year-over-year organic revenue growth by mid-2026. This isn't wishful thinking. It rests on bookings we've already signed and a pipeline that's converting. Our 100 squared accounts generated 73% of total bookings this quarter, plainly showing how the market is shifting toward high-value, long-term transformations. Beneath those record bookings lies our reorganization around AI studios by industry. The record bookings we're reporting today directly mirror that organizational transformation we carried out last year. A number of our top clients have already gone beyond the pilot phase and are scaling AI Pods across all of their operations. Let me give you a few examples. We're working with Employbridge, an Apollo-backed portfolio company, pushing AI-led transformation via our AI Pod subscription model. Following a successful pilot phase, Employbridge chose AI Pods as their core operating layer, speeding up delivery and driving fast adoption throughout the business. We're also working with Banco Galicia, among the most prominent banks in Latin America. After the pilot phase with our AI Pods, they ran an assessment to measure how efficient the model was among other vendors and similar teams. In almost every category, our AI Pods came out on top, which led the organization to decide to enter a scaled phase together with YPF, the state oil company of Argentina that is a century old. Using our AI agents under human supervision, we built a resource orchestration platform that assists YPF in coordinating its intricate supply chain more effectively, connecting with more than 5,000 providers. Already, our solution has assisted them in lowering the contract process cycle requirement by 30% to 40%, and in raising the productivity of their supply buyers by as much as 50%. By applying AI on Globant's orchestrated platform, we support their inventory optimization, allowing YPF's managers to secure the most suitable products for the job at hand prior to placing orders for new inventory. We enjoy a long-standing relationship with FIFA, assisting them in enriching their fan engagement channels in the digital age. Thanks to the deployment of AI Pods, we could go past traditional consulting services and reach a major financial milestone for the organization, cutting costs by 20% while not sacrificing the velocity or quality of our engineering output. Our work with LaLiga shows how AI Pods swiftly transform a whole ecosystem; within just 3 months, we went from concept to execution, deploying AI agents across critical functions such as budget preparation, contract analysis and audience data. The outcome is an enormous leap in institutional productivity. By shifting from traditional services to AI native solutions, we enable LaLiga to move new functionality at a speed once thought impossible. We likewise brought our AI Pods model to our long-standing partnership with Santander to power their new digital payment platform, Santander Pay. By deploying a specialized product definition AI agent inside the pod, we halved the projected time for the app's product definition. This AI native approach produced a 50% rise in the client team's overall productivity; in summary, it clearly showed how we can speed up the software development life cycle for one of the world's leading financial institutions. The professional services industry is being restructured at this very moment. The companies that possess the orchestration, the domain expertise and the talent to supervise AI at scale will determine what comes next; we will be relentless in delivering value for our clients, our partners and our shareholders; we will be disciplined in how we invest, and we are determined to build what we believe is the defining AI-native technology services company of the next decade. Globant has spent 20 years building the foundation for this moment; we have the platform, we have the people, we have the offering. And with that, I'll hand it over to Diego Tartara, our CTO. Thank you very much.
Thanks, Martin. Hi, everyone. Pleasure to be here. Building on what Martin said about where the industry stands, we continue to carry out our own transformation and our clients' with real commitment, paying attention to what customers need, assisting them in seeing where the gaps are, and putting together bespoke solutions that generate genuine business value. That extends past reducing costs and gaining efficiencies, reaching into strategic territory like growing market share or making customers more satisfied. To achieve this, Globant has reworked how we deliver, making sure the quality of what we deliver centers on both technology and the client. Groups that used to sit within delivery and operational functions are now part of the technology organization. As a result, our teams work free of competing siloed agendas and with greater unity between the quality of the solutions we offer and getting results out on schedule. What has come of this is technology-driven solutions for our clients that rest on stronger operational foundations. Let me walk through several examples. We're engaged with a top North American bank that is undertaking a strategic, enterprise-wide modernization of its credit and debit card platform, shifting from Gen 2 to Gen 3 accounts on AWS. Globant was chosen as the strategic partner to steer that migration, providing a next-generation cloud blueprint that boosts performance, speeds up delivery, and sets this line of business up for ongoing innovation at scale. This effort demonstrates our capability in supporting financial institutions already on the cloud and already leading on innovation as they keep breaking new ground in the industry. We've also been collaborating with Trafilea, a worldwide e-commerce group that creates and grows direct-to-consumer brands, which needed to quickly move new client stores onto their Trafilea platform. We developed an AI-driven solution that handles the whole process automatically, producing a migration that is 40x faster. That spared Trafilea considerable time and resources, while also letting new customers come on board more quickly. Over in pharma, we're working alongside PharmaMar, the global leader in discovering, developing, and commercializing marine-derived anticancer drugs, to speed up oncology research using AI. Using Globant Enterprise AI, the two of us built a multi-agent AI system that achieves over 90% accuracy in retrieving complex data and cuts the time to insights by as much as 15-fold, which helps scientists pick high-potential drug candidates for clinical development in a fraction of what that used to take. This intelligent system pulls together information from internal databases, scientific publications, and regulators including the FDA and EMA, which lets PharmaMar's teams spot promising treatment combinations and reach better-informed, quicker decisions; we also teamed up with TOURISE to build the groundwork for the world's first universal Agentic protocol for tourism. AWS, Salesforce, Amadeus, Red Sea Global, and Riyadh Air, among others, are involved in the initiative too. We showcased it at Davos in Switzerland before more than 30 global CEOs, and it's picking up serious momentum as the standard for how AI provides seamless, personalized traveler experiences at scale. GUT had a landmark 2025. The agency wrapped up the year thanks to breakthrough campaigns crafted for a number of the most prominent brands on the planet, among them a fully integrated 360-degree effort, Renaissance of Snacking, which took over the Las Vegas Sphere and introduced the Cheetos and Doritos Simply NKD product line. GUT stands as a true competitive differentiator, and its creative momentum keeps building. A key priority remains making our partnerships with leading AI model developers, enterprise platforms and hyperscalers stronger. Globant keeps presenting its strategic partnership with OpenAI to top clients across its key markets. A few weeks ago, we hosted their first multi-industry event in Spain so we could discuss opportunities with more than 60 current and potential clients in that region. In December, AWS awarded us competency certifications in both financial services and media and entertainment, which further cements the autonomy and quality of solutions from our AI studios; we also earned the SAP Excellence Award 2025 for delivery quality in Latin America, making us the most certified SAP partner in that region. Our Salesforce ecosystem capabilities also grew substantially, reaching expert level implementation distinctions across MuleSoft Anypoint, Data Cloud and Agentforce, plus top-tier partnership status across multiple Salesforce clouds. Our teams will take the stage at the NVIDIA GTC in March to explain how LaLiga is transforming its business via the most ambitious AI program in global sport, using Agentic AI to build connected intelligence across operations, competition management, content, marketing, sporting performance, broadcast and fan engagement. During such a disruptive year, we felt it was especially important to share our perspective with the global business community; in Q4, we published industry reports on retail, games and our annual tech trends outlook. All of them can be downloaded at reports.globant.com. Although AI continues to dominate many conversations, execution will be the real differentiator in 2026. Companies that want to remain relevant must accelerate their transformation journeys. Over the past year, we've evolved Globant to be the partner of choice for organizations ready to act and set the pace for the next decade. Thank you very much.
Good afternoon, everyone. I'm glad to be here to walk you through how we performed in the fourth quarter. It's encouraging to see our top line stabilize and to notice clients sounding more upbeat — that's a real step up from the kinds of discussions we were having 9 months ago. We finished the year with a strong quarter on the operational discipline front, beating our original expectations on revenue, operating margin, and free cash flow. Revenue for Q4 came to $612.5 million, which topped the $605 million we had guided to. That works out to a 4.7% drop versus a year earlier, and it includes a favorable FX effect of 180 basis points. Let's talk profitability now. Adjusted gross profit margin landed at 37.6% for the quarter. Gross margins took a modest hit from USD softness against LatAm currencies and, to a smaller degree, from mandated cost rises in 2 of our key delivery locations, Colombia and India. Even so, adjusted operating margin held at 15.5% for the quarter, unchanged from the prior quarter. We did a good job refining our delivery pyramid and keeping tight control over SG&A, which let us safeguard the bottom line even as we push to speed up growth. Our effective tax rate for the quarter was 23.5%, and adjusted net income came in at $68.9 million, which is an adjusted net income margin of 11.3%. Adjusted diluted EPS was $1.54, right in line with the profitability targets we've set. What makes me especially proud this quarter is how our cash generation worked. In Q4, free cash flow totaled $152.8 million — the biggest quarterly number in our company's history — and that gave us a free cash flow to adjusted net income ratio of 221.6% for the fourth quarter, or 355.3% on an IFRS basis. Looking at the full year, free cash flow hit a record $211.7 million, which equals 76.6% of adjusted net income and 203.6% on an IFRS basis. In Q4, we put $50 million toward buying back shares under the plan we announced in October 2025, and we intend to keep carrying out that share repurchase program. A big gain in our days sales outstanding, along with efficiencies in working capital and CapEx, helped boost our liquidity. We closed the year holding $250.3 million in cash and Short-term investments, up almost $83.3 million from the prior quarter. With total net debt at a modest $116.4 million, our balance sheet is still strong, which gives us room to keep following our disciplined capital allocation strategy, share repurchases included. Now let's turn to the outlook. I'll begin with our guidance for the full year 2026. Given where the market stands today, we're guiding to revenue between $2.460 billion and $2.510 billion, which implies year-over-year revenue growth of 0.2% to 2.2% with roughly 100 basis points of FX tailwind. We've treated the bottom of that range as a conservative starting point. The top of the range mirrors the conversion trends already showing up in our pipeline and growing uptake of AI Pods among our clients. On the profitability side, our expectation for adjusted operating margin falls in the 14% to 15% band, and baked into that band are the effects of a weaker USD along with mandated cost hikes in Colombia and India. We treat the bottom of that band as a stress test case, since it bakes in local currencies strengthening even more than they have at today's spot rates. The top of the band assumes currencies move in a friendlier direction, plus the gains from what we keep doing on SG&A dilution and better utilization. Operational discipline remains our priority so we can counter these headwinds and push toward the upper portion of our margin target. For 2026, the IFRS effective income tax rate should land somewhere in the 21% to 23% span. And last, our guidance for adjusted diluted EPS is $6.10 to $6.50, based on an average of 44.2 million diluted shares. The bottom of that range folds in the cautious margin assumptions I described earlier — in particular, the chance that USD weakness persists. Meanwhile, the top of the range captures the operating leverage we anticipate as we grow. Looking at Q1 2026, we're expecting revenues of $598 million to $604 million. That marks an improvement versus earlier years, when the Q1 drop was steeper. At the midpoint, our Q1 year-over-year guidance works out to a 300 basis points gain compared with how we performed year over year in Q4. For Q1, adjusted operating margins should come in between 14% and 15%. Gross margins will take a modest hit from USD weakness and from some statutory cost increases in Colombia and India, which I noted earlier. The IFRS effective income tax rate should be in the 22% to 24% span, while adjusted diluted EPS for Q1 should land between $1.44 to $1.54, based on an average of 43.7 million diluted shares. To wrap up, 2025 was a year of consolidation and evolution. We broadened where our revenue comes from, changed our go-to-market approach, made our operations leaner and reinforced our financial base; we head into 2026 with a solid pipeline, a delivery model that's more efficient and embeds AI across every project, and the financial strength to seize what's in front of us. Thanks for your ongoing support.
Our next question is going to come from Bryan Bergin with TD Cowen, so let's go to that line.
Two questions, then. I'll put both out there right away. First, a quick clarification on growth for the year at the high end. I believe you said it presumes a strong pod demand pattern, the one you've observed in Q4. But does it additionally call for some degree of macro or wider demand enhancement, or is it simply the same macro environment? Then my second one concerns the pod model across your GenAI offerings. Regarding the customers making use of those pods, are these discrete pieces of work, or larger engagements? Could you describe where it's specifically being applied, and then the net effect of a shift from old to new, if you're able to walk us through that.
Thanks, Bryan. On the first portion of your question, the high end of the guidance presumes we keep executing very well with our pods, along with a bit of betterment in the broader market. The midpoint, as always, is the most probable case, in which we essentially see roughly the same, more or less. That is, nothing significant shifts on the macro side, nothing shifts, nothing significant shifts across the business as a whole. That's the way we constructed the revenue guidance for the year. On the second portion, I'll hand it to the team here.
Absolutely. What we're observing among 7 of our 10 biggest clients is that they're genuinely enthusiastic about it. And by enthusiastic, I mean they truly want to shift away from billing by hours or other engagement formats toward this output-based approach, where naturally we bill for tokens, yet there's always a business outcome tied to it. So what's occurring is that at times we're moving that work from our existing engagement style into this fresh engagement style. With certain customers, modest pilot programs are beginning to emerge. With others, we're moving straight from pilots into full scale with no additional request along the way, since the results are truly outstanding, as I demonstrated through the examples I shared. So that's essentially transforming the entire dynamic regarding where this company is headed, correct? At this point we can not only grow our teams by adding personnel, but we can also link directly to everything taking place across the AI landscape. There's an entirely new market we're building, referred to as AI native technology services companies. And those AI native technology services companies need to discover a method for delivering their offerings, using agents that carry out specific processes repeatedly to guarantee that whatever gets produced meets enterprise-class standards with proper security, with the appropriate attributes for whatever needs to be produced, and then people overseeing those assets being generated. That shift was, 12 months back, in fact, 9 months back, this offering was not available. And today we find ourselves, in 2025, with over $20 million in ARR. And now we're expanding with major clients such as the one I referenced, such as FIFA, such as Santander, such as LaLiga, such as Employbridge and numerous others. So I believe this transformation is quite beneficial. It puts us in a distinct position. And naturally, all of it rests upon what we already possess: over 800 relationships with premier corporations, 28,000 individuals prepared to oversee every sort of product we're able to generate using those agents, the proper technology platform for delivering those services, and a commercial model that is entirely unlike anything previously seen. And the greatest part is that it isn't merely a forecast, but an actual business too. So we're incredibly pleased about that. I'm not sure whether that addresses your question.
Okay, I suppose that's partially true. What I'm attempting to understand is this: you did say that gross margin is extremely elevated compared to your typical level, correct, within these pod structures. But I'm attempting to consider how revenue shifts. So beginning fresh is fine for an engagement. Yet what happens when you begin with a client who already had an engagement — what does that revenue look like? You're becoming more efficient, after all. Does that create some netting effect on revenue?
No. I'm completely pleased to trade off that revenue. What I mean is we're essentially taking the teams we had at that customer and converting them into AI Pods, which brings a quite different revenue proposition and a distinct revenue value proposition. So this transition is occurring gradually, yet it is occurring. And at times there are new customers. At times there are customers engaged with us at a fixed price whom we now serve through this new approach. So that shift is beginning to take place, and we anticipate it will build momentum as the year moves forward.
Absolutely, Bryan — so with some of our customers, the extra efficiency we bring can end up helping them chip away at all that technical debt you usually see inside organizations. In other instances, it might be tied to one particular project that you can perhaps price more cost-effectively. So you're going to see plenty of scenarios, correct? What ties them together, though, is that with so much technical debt sitting with many of our customers, we're now able to be more productive and offer to handle essentially a portion of that added work using our AI too.
The next question comes from the line of Maggie Nolan from William Blair.
I was wondering if you might share your outlook regarding Latin America for 2026, especially in light of some of the fresh uncertainty that has come up around tariffs.
Certainly. As you'll recall, when it came to Latin America, we ran into some challenges at the start of '25, and for a few quarters the region was posting negative growth. Then, as we moved into the second half of the year, a recovery got underway, and we finished in a really solid position, with Latin America being the quarter's fastest region. Conditions differ — as you mentioned, situations vary from one country to another. Argentina and Chile, 2 of our main operations, are performing very well. Brazil is okay. We're essentially delivering in line with what we expected; and at this point, naturally, we have to wait and see what unfolds in Mexico, which is somewhat of an unknown right now. Still, the key countries are doing well. Regarding the recovery we managed in the second part of the year, when we examine which customers are behind it, the majority are in Argentina. So we don't see — we are not — no headwind is coming our way from Latin America, in our view.
All right, that's helpful. You also came across as fairly bullish on turning the pipeline into revenue, yet I noticed something in the prepared remarks — perhaps you anticipate customers seeking bigger scale or engagements that run longer. My guess is that would alter how fast the pipeline converts and shift how revenue ramps over time. Could you walk us through the way that shows up in the guidance, and whether it differs from what we've seen historically?
Sure, Maggie. What we're observing is that smaller deals close in less time, while the larger ones continue to trail somewhat, moving more slowly than we'd prefer when it comes to signing and scaling up. Still, building on the tremendous quarter we delivered in — the tremendous quarter we delivered in Q4, plus Q3 as well, we anticipate kicking off onboarding and turning it into revenue quite rapidly in Q2 and even in H2. So yes, when it comes to really big investments, customers remain hesitant and take their time before deciding. Even so, the pipeline stays robust. The caliber of the deals remains very strong. The Henry Square are doing extremely, extremely well, and that's where most of the bookings originate — as Martin mentioned, 73% in Q4. So I feel quite confident that this mix will let us proceed with a great deal of confidence.
The next question comes from the line of Puneet Jain from JPMorgan.
Given everything we've heard over the past 1 or 2 months regarding how Agentic AI is developing, what does that imply for spending on IT services? Martin, you said it's time for some of those AI investments to shift into execution. With all the news flow across the last 1 or 2 months, are you noticing greater urgency from your clients about adopting Agentic AI?
Yes; over the last few months, we've observed that businesses are starting to take action in this area. The paths include: how do I speed up my technical depth? How do I swap out certain Software-as-a-Service solutions that aren't very deep? How do I use AI to automate my processes? How do I replace workflows from agentic AI processes that existed previously? Naturally, humans need to oversee them. And I think those 3 paths plus a fourth one — how do I enhance my customer experience? That research genuinely struck me when I read it, regarding the notion of consumer happiness. Yes, consumer happiness concerning how interfaces and experiences are changing has declined for the last 4 years in a row. So there's a substantial technical depth of $1.5 trillion to $2 trillion, yet there's also a substantial consumer experience depth. So an additional source of demand is asking, okay, how do I bring all these interfaces up to the next generation of interfaces? So all these paths are generating a great deal of demand for AI. I think the method for delivering those next-generation services, those AI native services, has to be completely unlike what we did previously. And picture that each of these AI Pods, Puneet, is similar to a recipe or a set of instructions, like a process we've been honing for years and years; it contains various steps to produce enterprise-ready, security-ready kinds of solutions. And what we're generating with those tools is far more scalable than previously and far quicker than previously. So customers are noticing that. Now if you simply handed AI tools to people, those results don't appear. And that's why it's crucial to emphasize that this new industry is the means to create — is the means to create the savings you anticipate or, if savings aren't what you want, is the means to create the productivity you anticipate from these AI teams. So that's why I think the AI pods are truly gaining traction; it's a fairly straightforward way to grasp how to make those savings real rather than merely continuing to hand out licenses for AI tools for people to use. I'm not certain they'll use them properly; and once more, orchestrating and supervising a group of agents producing software is far more different, and that's genuine productivity, than simply handing AI tools to people. The two items differ by an order of magnitude. That is precisely what our AI pods are focused on. So yes, momentum is visible to me, and it will continue to build. It will continue to build.
And then when it comes to all this AI-related expenditure — for core modernization, for consumer experience, for AI Pods — do you view it as extra IT services spending? Or are those budgets going to come out of reductions in other areas of discretionary spend?
No. Look, I mean, I think humanity is going to produce 100x more software than it did previously, full stop. And for us, that's purely expansionary. So I don't believe that this will — oh, well, now we're satisfied with this modest gain in productivity and this modest gain in what our product can do. Companies these days are delivering features live at a pace much faster than before — you see it and you hear about it. I ran into quite a few instances of this during the past several weeks, which leads me to think it's only going to keep expanding. The more capacity you gain, the more you end up using — that's how it has always worked, right? Across every case — if we can turn out software at higher volume and with greater speed, then our software usage will rise; we'll ask for additional functionality, and we'll anticipate that a larger number of customers will be happy. So it isn't a passing trend. So sometimes, when I check out what analysts are saying or go through their reports, it strikes me that the scope they see is kind of narrow. What I really want to convey to you is this: the scope isn't actually narrow at all. Take just the technical side — that alone could stand as a whole industry the size of ours, just that technical depth, you know? And then if you layer in everything on the consumer experience side, plus all the new stuff — there's simply no scenario where the software required stays at the same level as before. It's going to be 100x more software. All of this is going to add up to improved offerings across a wider set of platforms, a greater number of AI Pods, and a more robust pipeline. During my remarks, I noted that over a span of four years — actually, for roughly two to three years at this point — the bulk of spending has been directed toward AI infrastructure, which does not automatically convert into appetite for professional services. Prior to that period, those same dollars were flowing into improved cloud capabilities, which in turn produced stronger Software as a Service results and additional implementation services. That cycle now has to return. That's the reason I emphasized both technical depth and the depth of the consumer experience, since eventually those areas have to catch up. If they don't, the consumer experience index is just going to decline year after year after year, which is completely illogical. Right now, while we're able to deliver the better and the best experience to our customers, satisfaction among customers when they interact with companies is dropping. What's the explanation for that? So in one way or another, companies got distracted by putting money into AI, pushing AI at people, and now the moment has arrived to take it seriously. There's nothing more I can add, buddy.
The next question comes from the line of Bryan Keane from Citi.
I suppose, just at a big-picture level, Globant has consistently been a company expanding at a double-digit pace through organic means; and this particular year was somewhat of a transitional one, with 2% growth across the year and clearly a 5% decline in the fourth quarter. What would you identify as, say, specifically taking place this year that may not happen again in future years? Was it merely some client consolidation? Was there any AI pricing pressure that got factored into the model? Like what precisely is the difference that occurred this year that necessarily won't repeat as we move ahead?
You're referring to this year or 2025, correct?
Yeah, 2025 compared to — yeah, moving ahead.
Right. My view is that 2025 overall was marked by unpredictability. In a lot of instances, businesses pulled back their spending plans. I'd say it was a period where macroeconomic ambiguity proved very difficult to get past for a large number of the clients we serve, and that hit us hard. At this point, though, things look somewhat clearer on that front. That's why I've raised what I expect — we should see a more typical year. After that sort of downward compounding off of last year's revenue, that figure hit its low point, and we anticipate returning to growth on a year-over-year basis around the midpoint of this year. As we get closer to the close of this year, the exit rate should return to a fairly solid growth level. So what you're observing in the year-over-year is essentially this: it was 12 months of readjustment, reorganization, client hesitancy, and similar things. The entire sector expanded more slowly, which really hurts; at present we're making up ground and beginning to expand once more. And by the time we near the end of the year, the exit rate is going to be considerably stronger than what's showing up right now. One thing about that year-over-year figure you're looking at — it already reflects something stable, correct, tied to where we are in the calendar. And it shows an enormous gain versus what we delivered during the same period a year ago. I'm not sure whether you or Juan picked up on that.
Martin is talking about Q1 versus the same quarter a year ago. So the start of this year is clearly stronger than the previous year, yet the way last year's quarter was paced is somehow weighing on the growth rate for 2026. If you examine the 2026 exit rates, they look closer to mid-single digit. And assuming we continue compounding, that ought to position us better for '27. Now obviously, an industry situation has emerged. I mean, looking at the overwhelming majority of the players, all of them sit somewhere between 3%, 4%, 5%. So growth across the sector has been weaker following the enormous investments made during COVID times and around then. There is a bit of getting — moving beyond that stretch of enormous investments. But the needs remain. The pipeline indicates that customers have been piling up debt, technical debt, and at some point that has to begin converting. Naturally, a stronger macro backdrop, a healthy U.S. economy should eventually be a tailwind. I believe we are emerging from 2 years of considerable uncertainty around the globe, and that has not been helpful. But overall, to sum up, I think Q4 marks a bottom when it comes to year-over-year numbers. Q1 is already delivering a better result compared with Q4, and that is expected to carry on through the rest of the year.
All right, a fast second question for you, Juan — how should we think about the way headcount expands and the revenue we get from each employee during this year? And does that approach shift in any way when we bring in more AI Pods?
Absolutely, yes. What we're finding is that we're able to deliver somewhat larger figures, and we can keep lifting the revenue we generate per person. That happens even while our number of employees stays flat or drops, since the AI Pod approach inherently calls for fewer individuals. AI Pods are agents overseen by a small number of staff, so the demand for talent is reduced. My view is that this will begin to shift things somewhat — not only at Globant but across the industry — regarding the pattern of staffing and revenue that has held for the last 20 years. Without a doubt, as we succeed in pushing AI Pods further into our client base, the proportion of AI Pods versus the remainder of the business climbs, and that ought to help both revenue per head and margins.
Arvind Ramani with Trust Securities is up next with a question. It looks like Arvind's line is having a problem, so we'll move on to the following question. Jim Schneider with Goldman Sachs is up next.
Could you possibly speak to the AI Pods segment and what it would take to reach the higher end of that $100 million run rate ARR target for that unit? What's needed to get us there? How many additional bookings must come in? What portion is already backed by the AI Pods pipeline you have today? And maybe you could just share your general view or how confident you feel about hitting the top of that range.
Excellent question. Thanks, Jim. Reaching the upper portion of that range wouldn't require a large number of major customers adopting this model. Still, let's watch how it goes; that's exactly why we remain careful; the momentum we're seeing makes us very enthusiastic. At this point we're witnessing deals worth $20 million, $18 million, and $15 million shifting into this type of arrangement, which we find highly encouraging. So we anticipate hitting those figures. That said, I don't want to sound — I mean, this is our first time providing guidance on them. I also don't plan to give guidance on those figures each quarter, but I'm attempting to stay measured here. So — even so, I feel quite positive regarding the chances of getting to that top line — the top guidance we issued at the close of 2026.
To build on what Martin said, how the pipeline is behaving with regard to AI Pods is quite heartening. A favorable pattern is emerging, along with extremely rapid expansion. Beyond that, there's the willingness of our biggest customers to begin piloting — yes, piloting — and to begin scaling up. Looking over the roster of clients now starting to ramp up with this new technology, it's genuinely uplifting. So we have a lot of confidence, and we believe we will land quite near the range we provided for AI Pods, yes.
That gives us useful context. Could you also walk through what gross margin trends might look like across the whole company over the course of this year? Juan, I believe you brought up certain FX-related and regional cost pressures that weighed on Q4. Is it reasonable to assume gross margins have bottomed out and will climb steadily as the year goes on? Or what's the right way to frame how this develops?
Thanks, Jim. So yes, the weaker U.S. dollar has definitely hit us. When you consider the Colombian peso, Mexican peso, Chilean peso, and Brazilian real — nearly every currency in the Latin American markets we serve saw meaningful appreciation during 2025. That's what dragged down the first — pardon me, the final quarter of last year, and it's carrying over into the start of this year as well. My sense is the dollar is now averaging a level that's quite low by historical standards. So some effect will continue from that; but equally important, we should stay focused not only on currency movements but on steering the business toward AI Pods, since that's where we get productivity gains and better margins; and the more we run those models, the more efficiently we can operate them; so I believe pricing will hold up fine. That said, pricing across the general business won't deliver huge growth this year. Still, there's clearly room to grow our AI Pods share, and by doing so, sustain or lift our gross margins as that business scales.
The next question comes from the line of Jonathan Lee from Guggenheim.
What I'd like to understand is this — when you talk with customers during January and February, what makes you feel confident about the timing of conversions that your outlook assumes, especially since you've flagged certain client hesitancy and certain historical difficulties with conversions?
The thing is, compared with before, customers now seem more willing to pick up discussions about major agreements again. We're also noticing that the volatility and uncertainty are settling down to lower levels in those discussions. And here's another point worth factoring in, Jonathan: as we built out the figures for 2026, we managed to incorporate several quite significant deals that we signed in Q3 and Q4, correct? Plus additional deals currently in progress that we hope to finalize ahead of Q1's close. So with part of that volatility fading, more clients showing openness, and those signed deals now being onboarded and ramped up, we have confidence that the path ahead will look different.
Great. It's good to hear that; and as a quick follow-up, could you break down what you anticipate for each of your verticals throughout the year? And do you see any of them slowing down rather than speeding up compared with what you've observed?
Looking at our various industries for last year, financial services enjoyed a solid year, with growth of roughly 13%; consumer retail and manufacturing have been performing extremely well. We still expect that pattern to hold in that particular industry; up to now, we haven't witnessed a rebound in professional services, which was somewhat of a drag during 2025. Technology will bounce back. We're beginning to see some large deals taking shape with our tech customers, which was another sector that didn't perform the way we wanted last year. But certainly, looking at Q4 and some of the expectations ahead, that will be fine. And last, health care. Health care and gaming, correct? Those are the 2 that are major deals already signed that are currently ramping up and that help explain the sequential growth we ought to see for the remainder of the year; so that's broadly speaking, I attempt to cover all the industries as we report them. So hopefully, that's helpful.
The next question comes from the line of Sean Kennedy from Mizuho .
Regarding the increase in bookings and the momentum we're seeing in the business — it's great to see that. So I was curious about AI Pods and what your customers are saying in those conversations. Are procurement teams getting more comfortable with the AI Pods business model compared with the legacy approach?
That's an excellent question. Thank you; this has been among the hardest items. But as this thing builds — as the concept builds momentum across the industry, among analysts, among all of you, procurement teams are becoming more at ease. And beyond that, I think the fact that what we're describing is something incredibly solid and is, I'd say, 10x more transparent than the traditional model, procurement teams adore it. So when you're able to link any asset being generated to the quantity of tokens and grasp that, that connection between what you're paying and those tokens is far better than claiming we used this many hours to do whatever; so in my view the AI Pods offering is incredibly solid. Naturally, a lengthy path lies ahead in convincing more people about this. The more all of you help us, the more we're able to do it. So we value any kind of explanation in your reports; and analysts from Forrester, from IDC, from McKinsey, from Bain are already describing this way of working. And 70% of people, as I read in a report the other day, 70% of people purchasing technology anticipate a shift in how engagements take place. And what answers that shift is either a monthly subscription, a quantity of tokens or some sort of combination of the two, but it has to be on that code. So procurement teams are replying quite favorably to that. That said, naturally, it's always complex, but it isn't impossible. And the business is pressing extremely hard for that.
Also, at the beginning of this effort, we paired the AI Pods with a capacity equal to what a team of a certain number of people could handle, correct? And that kind of tool is something procurement is accustomed to; these days we've grown far more sophisticated, and in fact we're discussing and linking usage together with the subscription to results to some extent — we also offer complete visibility; so additionally — we've become increasingly adept at explaining and demonstrating the way an AI Pod delivered, and the data has also put procurement teams much more at ease.
Understood. Then for my second question, I believe you mentioned that the upper end of the guidance assumes the conversion rates currently being observed stay steady. So I was simply curious about how that has been moving across the past several months.
Are you asking about AI Pods specifically, or more broadly?
Sorry to interrupt.
Are you referring to AI Pods, or speaking more broadly?
No, no — just broadly speaking, overall. Well, they're — I mean, the guidance got put together by us a few weeks back. Thus far, what we're observing with conversion rates, plus certain large agreements we finalized in the prior year that are now scaling up, gives us confidence to land at the middle of the range we provided. Our pipeline, along with an anticipation that things will get somewhat better as the year progresses, could potentially push us toward the top end. Yet certainly, given where conversion rates stand right now — or the present rates plus everything already accomplished in Q4 and early this year — that ought to bring us to the center point of the guidance we gave. However, once more, that — that — the midpoint itself factors in no significant enhancement or substantial shift regarding the broader conditions.
The next question comes from the line of Arvind Ramani from Trust Securities.
Nice results. There's a ton of AI-related questions, so I suppose I'll follow that same theme. I believe you mentioned AI Pods produced roughly $21 million in ARR during this quarter, which is quite impressive considering how early things still are. Even so, that represents under 1% of total revenue, so it remains fairly modest. When you examine this model, though, it's essentially built to accomplish more through tokens and less through people. As AI Pods grow, what's your approach to keeping them from eating into your main seat-based revenue? Then, second, what does your internal modeling indicate regarding the crossover point for revenue — the point where token-based revenue starts bringing in more than headcount-based revenue?
Preventing cannibalization isn't something I'm trying to do. I actually want that shift to take place. That positions us on the correct side. It also means that as AI expands, we'll continue expanding too. The method we use to deliver services through these AI Pods is extremely scalable by a wide margin. It's somewhat the style built over many years of experience, along with the configuration files we rely on for every one of those AI Pods and how the agents need to execute the process — it's genuinely very, very impressive watching how those small recipes used to reach the assets our customers hold are truly transforming how we apply AI and the models we currently have, since we're embedding into those configuration files all the experience accumulated within the company. So this represents our north star for delivering technology and solutions going forward; and — and naturally, some customers will be at ease with the hours and the existing business we share with them, and so on. I'm very happy to continue doing that. That said, we're strongly urging our customers to shift toward this new model, since we trust in the advantages around transparency, productivity, and the long-term relationship we share with them; and AI Pods aren't necessarily less expensive — they're more productive; so we foresee a transition year during which we'll be moving from one type of business to the other type of business. It'll therefore resemble a sensible migration more than anything else. So I'm not sure whether that answers your question.
Now, regarding the second part, Arvind: as we shift the business over to AI Pods, and those AI Pods begin picking up share — ideally by year-end, based on the current forecast in our internal projections — the figures Martin referenced regarding ARR at year-end; so without question, we moved from 0 to a $20 million run rate within merely 2 quarters. The model has undergone considerable evolution, considerable testing alongside customers, considerable internal effort to ensure it delivers a differentiator for Globant that positions us more strongly against other players or other models that could exist. And at this point it is beginning to accelerate. As we talked about earlier, when we examine which customers are now coming aboard, which is the size of certain deals. Now it isn't merely, let's experiment with something small here, but certain customers are in fact discussing $10 million, $15 million, $20 million being shifted to the new model; so I believe we are beginning to see that acceleration following a couple of quarters of understanding the customer, gathering feedback. We only introduced this in Q3 last year. So it's merely a few quarters that theithas been around. It's already generating interesting revenues, generating plenty of momentum with customers. I would say that by year-end, it will hold more relevance relative to our old pie. But without question, it ought to be next year when the curves begin drawing nearer.
Yeah, and when it comes to your models and everything you folks are tracking, I believe we've got to recognize that this industry is changing — and this new sector around AI native services will essentially be unlike the way things were before. AI native services, naturally, draw on knowledge, but also on repeatable processes and capabilities we never had previously; similarly, Amazon built the cloud computing sector when they introduced Amazon Web Services. Obviously, at the scale we operate at — and I'm not trying to draw a comparison with Amazon — but at our scale, we're essentially carrying out this vision of AI native technology solutions. And how you model that, how you go about it, is completely unlike the way things were before. That's the reason I began discussing ARR, since it's a form of recurring revenue that isn't tied to headcount. At present, we rely on people to oversee what the agents produce. Each type of asset requires a specific amount of time for that oversight to happen. So we're able to work out how many people are required for it. But our belief is that revenue has no connection to how many people we place there. Revenue is connected to how many assets we build and how enterprise-ready those assets are, correct? You might use Codex, yet you won't obtain all the discipline and the rigorous approach that enterprise software demands; so what we're building here are processes for creating assets that are enterprise-ready, security ready, possessing the scalability, repeatability, and maintainability that you'll need going forward. So it truly is a different way of looking at the industry itself. There's a $2 trillion industry that has to shift to a new model, and this is just the very start of that.
Great. That really helps a lot. I just have one quick follow-up. I just have one quick follow-up. When it comes to the token, the token costs you a specific sum of money, and then you bill your clients for it — are the margins you earn on that higher or lower than what the company makes on average?
Like we said in our report, those margins fall between 45% and 60%, and it depends on the AI Pod, it depends on the customer, and it depends on what we need to build. As time goes on, we anticipate that margin will grow, since we become more efficient with technology, we become more efficient at supervising, and the technology used for supervision keeps getting better as well. So that's the type of model we have in mind. When we can oversee more assets, using the same number of people, using fewer people, or using better technology, we'll require less supervision, and we aim to raise our margins; and that's a virtuous cycle that takes place here. And that's not all — each conversation and each token gets stored on our enterprise AI platform; and those tokens can be leveraged to enhance the processes and to preserve corporate sovereignty over the company's processes. So this is a sort of productivity explosion, and it will reshape the entire industry, and not only for software development — it will also happen for process operation. Right now, we have AI Pods plus AI Pods software, and we have a further AI Pods offering called AI Pods operations; and those AI Pods of operations, they handle that kind of work for running certain processes of companies, and we also bill for it based on consumption. So it's a completely different way of thinking about how professional services and how services will be delivered going forward. There's plenty of value to add for companies like Globant, and there's plenty of mindset change that's required to grasp this new industry. I could talk about this endlessly, but.
Much appreciated, Arvind. Regrettably, we've reached the end of the time allotted for today's question-and-answer portion. So, at this point, I'll turn things over to Martin for a few closing remarks. Martin, you're unmuted now.
Much appreciated, Arturo, and thanks to all of you for the backing, for the assistance, and for joining us right now. So long. Catch you in the following quarter. Thanks.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete Globant S.A. transcript - plus 132,000+ transcripts from 7,000+ companies, speaker segments and full-text search - through the EarningsAPI REST API or hosted MCP server.
Get an API key View API docs →For developers and AI pipelines
Programmatic access to Globant S.A. earnings transcripts and 132,000+ others is available through the
EarningsAPI REST API and the hosted MCP server.
Quarterly plans from $162 - full transcripts, speaker segments, full-text search,
and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.