Welcome back. You're watching Corporate Radar. Now, let's talk about the IT space. Remember, IT plummeted on Friday after Accenture cut the guidance. What's the read-through? We have with us Srikanth Velamakanni, cofounder, group CEO at Fractal Analytics. Srikanth, great to have you this morning. You know, I'm going back to that town hall conversation that we had when you came down to our studio and we'd asked you the prescription for Indian IT going ahead. And you said that companies need to take some short-term hits. They need to increase their R&D spend. It's okay if the margins reduce. They need to convince the street about it because that's what it takes for long-term sustenance. And you know, I'm just tying in with what Vishal Sikka, the former Infosys CEO, recently said on Voices of the Valley to CNBC-TV18. He's actually arguing for IT and SaaS companies to consider going private because then they can take those bold, radical calls for innovation. I mean, you, you know, since the last one, two months since we spoke, things are again changing. They're changing rapidly. What is your view on Indian IT from now on? And talk to us about how they need to reinvent themselves.
The first thing to realize is that the demand for software, for tech, is only increasing. Companies are becoming more and more tech-centric, tech-intensive. Therefore, overall, people will spend more on tech than they have spent traditionally. So, which means that the opportunity space is large. The only question is, how will the tech services industry capture it? And they can capture it if they can move from input-based models to output-based models, bring in more productized revenue, and increase the overall R&D investments into what they're building. I think they will be ready to capture that much expanded market in the next couple of years. I foresee that the best days of the tech industry are well ahead of us. It's just a short-term blip that companies have to find ways to encounter and grow from.
You know, tech spends undoubtedly are going up, but it's getting redirected towards the infrastructure side. The services spend seems to be compressing. How long do you think that will last? And do you think the only way for the Indian IT companies to thrive, not survive, is to go beyond services? Or do you think the services spend itself will expand after this initial period of compression?
What you're saying is that overall the world economy is growing by 2 to 2.5% a year, whereas the tech spends are growing by about 13% a year. So, there's a major gap between the growth rate of the world versus the growth rate of tech. That's an expanded opportunity. A lot of that, like you rightly said, is going towards infrastructure, is going towards tokens, and so on. But, what you will see eventually is that every company in the world will need to transform itself through technology, through AI. And therefore, even the service opportunity will expand dramatically in the coming years. The way it will work is that let's say you need 100 units of work to finish a task. Earlier, it was 100 units of people time. Now, it's probably only going to be 20 or 10 units of people time. But then, the total amount of work dramatically expands through what is called a Jevons paradox. So, therefore, the services spend will also expand, but if you are trying to be undifferentiated, you will be struggling. You have to be very differentiated. You have to bring in AI, and you'll certainly build products as well as a way to overall expand your revenues and expand your margins over the next few years.
And that point that Vishal Sikka made that companies must consider or think about going private because that gives you the freedom to innovate without getting tied up in quarter after quarter numbers and reporting. I mean, it's a radical thought, but do you think it's far-fetched or do you think that's what's needed right now?
Well, this is not the first time that would have happened, and any company that needs to, let's say, structurally reinvent itself, it makes sense for them to go private and fix whatever has to be fixed and come back into being public. That works. My sense is that at this point in time there is peak fear. There's peak fear about technology services stocks. You can see with what happened in Accenture over the weekend and so on. The key thing right now is that the market will understand if you're going to say, 'Hey, I'm going to let my margins suffer a little bit. Let me increase my R&D intensity. Let me come back in a different avatar.' They don't have to necessarily go private in order to do that. The market will understand if they're making structural adjustments in order to be much more competitive in the next couple of years.
So, R&D spends. Now, broadly Indian IT as an industry may be spending one, one and a half percent of their overall revenues. You are at six and a half percent and you've spoken publicly about going to 10%, and many of the global companies are, you know, 10, 20, even 30% in some cases. At least to begin with, what should be the number that, you know, by how much should the R&D spends go up by?
So, firstly, let's look at product companies. They spend roughly 14% of revenue on R&D. Services companies are currently spending about 1% or less than 1% of revenue on R&D. Fractal sits in the middle at about 7% of revenues on R&D. I would expect that most of the services companies should go at least to 3% of revenue as R&D. That would give them a really good opportunity to build things that the world needs. AI is really changing the way the world works and it is giving an opportunity to reimagine every business process with AI. So, what it does is that if you build those products, you can certainly be a part of that transformation, drive your licensing revenue and move from input to output-based models. Right now, a lot of the industry is in a commoditized selling-people's-time kind of stage. That is the part that gets compressed very badly and therefore companies have to move away from selling people and time into more output-based, outcome-based, license-driven models.
Should they also—that's of course one part of the pivot, moving from labor arbitrage, input-based to outcome-based. But should they also consider building products?
