Phil Fersht — CEO, HFS Research[00:12]
Good afternoon, it’s good to get some time today with an old friend of mine who’s having some exciting times in the services industry. Obviously, you know me, Phil, first. I’m the CEO of HFS Research. And today, I’m joined with none other than Samir Dhir. I met him when he was leading a lot of the business for Virtusa, and now he’s taken over as the CEO of Sonata Software. So maybe I’ll just ask Samir to give a little bit of background of where he’s at, why maybe he decided to take this new job on, and a little bit about — we’ll talk a bit more about his big plans for the future. So over to you, Samir.
Samir Dhir — CEO, Sonata Software[00:56]
Well, thank you, and thanks for the opportunity. And talking about the old times, I do recall our first meeting, and you’ve been known to ask some tough questions. I’m looking forward to the next 20 minutes of exciting discussions. But I think Sonata, as you know, has been around in the industry for 35 years. It’s close to a billion dollar company. And as I was sort of coming out of Virtusa, I think this presented a great opportunity for me, and I think there are two or three points that sort of stand out for Sonata. Number one, and probably the most formidable one, is that about 70% of the revenue for Sonata comes from clients who have been with Sonata for more than 10 years. Now I’ve been in the IT industry for a long time, but I haven’t seen that metric — that so many clients stay with the company for more than 10 years for such a long time. So it just shows the passion and the energy the company has shown towards the clients that get in with Sonata and then stay with them for a longer term. The second point I think is extraordinary about Sonata, which I don’t think many people realize, is that 88% of the company employees tell us that they love working for Sonata and this is their second home. Now that’s a staggering metric. I have never seen 88% of people say this is a great place to work in any company that I worked in prior times. So I think it’s the two ends of the book — when the customer’s staying with you for more than 10 years, and secondly employees, 88% of employees saying this is a second home — I think are two very, very special metrics, which sort of tilted my decision to join Sonata, and I thought that this will be an exciting journey ahead. And also, the third part is that the offering that Sonata has built over the years is in the modernization space, and there isn’t a client today which is not talking about modernization, and the topical discussion today is on data. I think it sort of leads well into that whole piece that every customer is trying to move their consumers and give them a better experience by modernizing their platform. So I think a combination of all these three factors contribute to it, and I’m very excited about the journey in front of us. I think Sonata can be the next mid-tier rising star. We already are, but I think we can keep punching over our weight and take the market share away from the Generation One companies with whom a lot of fatigue is beginning to set in, at least in the client base.
Phil Fersht — CEO, HFS Research[02:56]
OK, interesting. So what are the key areas — you talk a lot about data modernization. So what are the key areas that you’re talking to clients about in terms of modernizing their infrastructure and their capabilities?
Samir Dhir — CEO, Sonata Software[03:25]
Yeah, so I think there are many parts of this modernization story. So, number one, at the first layered level, I think customers are talking about modernizing their front office, essentially, not so much the back office. I think everybody’s trying to gain the loyalty with their consumers or end customers or B2B customers. So I think that’s a significant first part of the effort. Now if you peel that down further and say, so where else is the effort in that? So I think the effort is in the contact center, the effort is in the sales and marketing side, and effort is in the HR side to provide superior experience to the employees and engineers. So I think those are the few areas that customers are peeling down and saying, what do we need to do to modernize the estate — which is the application estate — for this set of audience to provide them a superior experience. Now underpinning that is the whole data, which really defines the experience layer in that. Now you dovetail on top of it the buzzwords of AI and GenAI, it gets very interesting right there. And that’s really where we’re seeing most of the customers beginning to take shape in terms of their go to market. Let me give you an example, Phil. So one of the travel companies out of Europe — you know, earlier on when you had to book your travel, you’ll go and say from this date to that date, this location within this budget, I want to book my travel. But now it is becoming far more generic. So people want to search for experiences, people want to search — you know, I want to do this during my vacation. So the search in itself for a travel vacation and destination is becoming very open-ended, and that’s where I think the modernization of the platform, both from an infrastructure perspective as well as application perspective, becomes pivotal, because if you don’t have the underlying platform and equation to do that, then you’ll be under a competitive disadvantage where the market is headed for.
Phil Fersht — CEO, HFS Research[05:03]
Interesting. Yeah, I mean, to build on that, we recently did this joint study, and we’ll share some of that with this interview, where we spoke with 500 executives across Europe and North America. These were a split of tech and non-tech senior executives. And of those people, 68% of respondents said ineffective data capture and human error were the top reasons for poor data quality. So what do you guys advise clients to do to fix this human error issue and really improve the quality of data that they need?
Samir Dhir — CEO, Sonata Software[05:49]
Yeah, no, I think there are a few key elements which help customers build confidence on their data. And if you sort of peel back and take a look at it from outside in, the starting point of this is: where do you want to focus on collecting your data? What are the sources of data that you really want to collect and rely upon? That’s a critical part, because not many companies provide the application infrastructure to make sure that they’re collecting the right data — both the real transaction data as well as the alternate data through external interfaces. I think it’s an important, integral part of that architecture; that’s point number one. Point number two is the interface between the various functions within an enterprise is extremely important. So if you look at a single data element of sales commission, the finance team looks at the sales commission in a very different way versus the sales manager looks at the sales commission in a very different way, versus the operations people look at the sales commission in a very different way. So how do you build a data dictionary which is consistent across functions is an important aspect that I think gets overlooked often. And the third, probably the most important these days, is around security. You know, how do you provide an environment which is secure, both in terms of PII data compliance, various regulatory compliances, but also in terms of the overall application architecture and data architecture that you provide, so that within the organization also there’s enough security and layers of security provided — that’s a third dimension to this architecture. I think those are the top three, Phil, that sort of come to my mind in terms of how we look at it and how we advise our clients to construct this whole piece out as they modernize their data platforms.
Phil Fersht — CEO, HFS Research[07:09]
OK, interesting. So let’s move to, I think, some of the topics du jour. And we’ve talked a lot about GenAI, and it’s obviously — I mean, I’ve sat on a bunch of briefings with all the early generation companies you spoke to and some of the big four firms and others, and they’ve all jumped fully on this bandwagon of excitement and, let’s be honest, some hype around the space. But I gather you’ve got a slightly less hyped view of where this is and how this is going to impact the industry.
Samir Dhir — CEO, Sonata Software[08:03]
Yes, I do, and I think as a company we do. So clearly, with anything that comes out new, I think our tech industry has mastered the art of overhyping it, and it takes four or five years, sometimes 10 years, for our technologies to mature. Of course, the timeline is shrinking, but it does take some time for some of these technologies to mature. If you look back at cloud when it started in the 2006, 2007 timeframe, it took us about 8, 10 years before really cloud went mainstream in a very meaningful way. So on GenAI, I believe there is a lot of hype. Having said that, there is a lot of goodness in it as well. So this is not a passing fad. This is definitely going to stay here, and I think it will mature over a period of time like with any other thing. And hence we are taking a very deliberate measure of really positioning Sonata as a responsible-first GenAI company. I think the insights are available a dime a dozen these days, but how do you provide that insight in a very responsible way and also build the infrastructure around it that is meaningful and scalable for the companies? We believe, and I personally believe very strongly about it as well, that in the next two to three years, the technology will mature. I don’t think it is enterprise grade at this point in time. I think it is commercial grade or consumer grade, which doesn’t require a very high level of accuracy. Hence words like hallucinations have become very popular these days. The reason they’ve become popular is because the technology also is not very mature. Until the time it becomes very context-driven, till the time it becomes very predictable, you cannot say this is enterprise grade ready at this point in time. But as you have to make forays into it, and in the next two to three years as the technology matures, maybe sooner, that time you want to be really leveraging it for its power, and the powers are very, very high. The powers are in providing insights, powers are in providing productivity benefits, the powers are in terms of really shortening the time cycle of innovation from a time-to-market perspective. So those are all absolutely high class power that you can leverage upon, but you have to ride out the hype before the real stuff comes out.
Phil Fersht — CEO, HFS Research[10:03]
Very well put, Samir, very well put. I think a lot of people will appreciate that point of view. Because as we think a bit more about the study we did, 17% of respondents felt the majority of their data is still unusable. So how can we work better with partners to improve this — you know, how can we actually start to deal with these elements of errors? Why do we continue to struggle with poor data capture after so many years of talking about this and trying to find the right tools to help us?
Samir Dhir — CEO, Sonata Software[10:42]
Yeah, you know, I have a little provocative point of view on this. This is less of a technology issue. It is more of a business ownership issue. And many customers that I talk to — it sort of comes out in a few minutes that that’s true in many, many enterprises. Let me explain. You consult with many customers yourself; well, your organization does. There isn’t a big company today which doesn’t have either two of the three hyperscalers in any shape and format. Everybody has at least two of the three hyperscalers, in some cases all three. There isn’t a company out there which doesn’t have a Snowflake footprint. There isn’t a company out there that doesn’t have a Databricks footprint. So investments in technology have been pouring in, in the hope that that investment will solve the problem. The reality is, unless you solve the data ownership issue in an enterprise as to who owns what data piece — and that has to come from the CEO down — you’re going to, what I call it, if you automate a mess, you will have an automated mess. And I think that’s what companies have been trying to do, to really automate the fragmented, disparate data systems, to automate them out. So you have automated data mess now in many, many cases. And hence, if you take a step back and say — let’s take an example, the example I was talking about, sales data again. Who owns that sales data, and what are the interface points of that, and has that data dictionary been fully socialized and understood by various entities? The answer is no, because each function decides their own data and then they expect others to follow their definition of data, which doesn’t work for the finance organizations sometimes, it doesn’t work for the chief operating officer organizations many times. And hence the definition of organization around data is really important, and aligning around it is extremely important. Otherwise you’ll keep buying these, you know, buzzword new tools, and there’ll be more inventory of that in any enterprise, and every downturn people will cut the license cost and in any upturn they will increase the license spend. And we’ll go through the sinusoidal motion, but I think the fundamental part is how do you define and establish the ownership of data in an organization, right from the CEO downward — I think is an important part that often gets overlooked.
Phil Fersht — CEO, HFS Research[12:51]
OK, yeah, I think those are wise words. This needs to be a senior-level strategy. And now you’ve got the senior-level folks who are jumping on GenAI, they’re going to realize quite quickly that you can’t enjoy all these potential capabilities and benefits down the road if you haven’t fixed your existing mess, quite simply. And you mentioned that you can’t automate a mess, you can’t clarify a mess — you’ve got to fix it. And you know, we’ve heard about data swamps, and I’ve been hearing about data dumps now, but it really feels like the patience that companies can continue to have around data is really drawing very thin, and the time for action is really becoming very, very critical for many businesses.
Samir Dhir — CEO, Sonata Software[13:38]
Yeah, absolutely. And on the GenAI piece as well, Phil, if you think about it, a lot of people are talking about code generated through GenAI and things like that, but there are lots of nuances behind it. For example, when you generate code through GenAI, who owns that code? And if you’re a bank and you put code generated by GenAI in a production environment, then there is a huge amount of risk that tomorrow somebody can claim that that would be their code and the bank has misused the code. So I think there are a lot of legal and regulatory type issues which are still unsolved. That’s why I keep thinking, you know, some of these hyper- and more mature companies realize that and really unbundle it or unpack it — to the extent there is goodness in it, leverage the goodness and keep moving forward.
Phil Fersht — CEO, HFS Research[13:56]
I like that. So, you know, we’ve come through a fairly tough economy, and you guys seem to be moving on quite aggressively. What do you think we’ll be talking about maybe beginning of next year, midway through next year, if the economy is going to get a little better? Inflation seems to be under control a bit now. What do you think we’re going to be talking about next year? Is it going to be a year of more positivity and excitement and investment, do you feel, or do you think we still have some dark times ahead of us?
Samir Dhir — CEO, Sonata Software[14:50]
Yeah, look, I’m an eternal optimist, Phil, and I think that what we’ll be talking about is still very much how do we accelerate the modernization journey for our customers. I think that will still be on, absolutely no doubt in my mind. I still think that what we’re talking about is the proactive measures many of the high-tech customers have taken in anticipation of a slowdown would have eased out a little bit. They would have started spending it back again. I do hope — and this is the hope, I’m not sure about this — that the data issues that reside in large healthcare and banking customers, they have really gotten around it and really taken to heart some of the discussion we’re having today and some of the findings that your team have come up with. Those can get more mainstream. But you cannot ignore these hard realities — you know, you can’t keep putting the lipstick to the pig. I think at some point in time you have to stare into their head and solve these main issues in a meaningful way. So I think those will be the conversation, but will they — how far they will go, I think only time will tell.
Phil Fersht — CEO, HFS Research[15:59]
Only time will tell, but I feel like the time to make big decisions and move more aggressively feels like it’s the order of the day. And data is at the core of everything now, and I don’t think companies can keep ignoring that. So it’s been fantastic getting some time with you again, Samir. Hopefully have you back on here again sometime soon as well, but I look forward to sharing this interview with our audience and keeping the great conversations going.
Samir Dhir — CEO, Sonata Software[16:17]
All right. Thank you, Phil. Enjoyed talking to you, and thanks for the hard work that your team did with us. Thank you so much. Bye bye.
