Rohit Kapoor: Sure, Surinder. So look, I think Gen AI and the application of digital is certainly allowing our clients to be able to have greater cost efficiency but at the same time have a much better end customer experience and to be able to target this towards generating growth. What we are finding is because our penetration and our share of wallet within each of our clients is still relatively low, it’s less than 15% to 20%, anytime we are able to deploy Gen AI for our clients’ benefit, we actually gain overall volume, and we gain in terms of revenue with that same client even though we are providing them with greater efficiency, and we are providing them with a greater ability to create impact. In the example that I shared in my prepared remarks, that’s a classic example where we’ve had amazing experience with our clients, and we’ve been able to deploy a proprietary Gen AI-based solution across the enterprise and impact 1,100 customer service agents and 6 million calls a year.
But what’s happened is that the customer has given us more work because they want us to deploy this in other areas, in other business lines and across other providers that they work with as well as on their own captive operations. So the net impact of that is that our business relationship with our clients becomes a lot more strategic and it increases in size and value.
Surinder Thind: That’s helpful. And then just as related to that, as you think about implementing these solutions are more from the perspective of clients what are — how are they getting across or thinking about things like what’s an acceptable error rate for an implementation? It seems like the solutions operate to various extents in terms of the quality of the end result here. Obviously, there are some low-risk cases, there are some higher risk cases. How do we think about the time line on some of these types of projects of going from these proof of concepts to actually the live and what they actually mean from a revenue perspective?
Rohit Kapoor: That’s a great question, Surinder. And that’s — I think at the crux of how clients and prospects are thinking about using Gen AI. So first and foremost, accuracy of the results is critically important. And second, the adoption of that solution is critically important. And the adoption only takes place if the accuracy levels are high enough that they’re providing you with a distinct competitive advantage. Our role is to make sure that we can deliver high-quality service and high accuracy and, at the same time, enable change management so that the adoption of using these tools is a lot better and a lot more sustainable. Now there is a fair amount of risk sharing that we undertake with our clients and align ourselves to the end outcomes.
So there are — there is a fair amount of investment that we are making in terms of having prebuilt accelerators and pre-build solutions that we can bring to bear to our clients. And they don’t need to invest in that early adoption capability. And then there is a lot of fine-tuning and a model change that we need to do to get to those levels of accuracy. So when we start with a new solution, the accuracy levels are very low. They are like 65% to 70%, but we have to drive that up to 90% to 95%, and that is driven by a strong understanding of the domain and understanding of how and what the data is telling us and therefore, the fine-tuning of our algorithms and our ability to embed this into the workflow so that it is actually useful for every single person that is adopting these solutions.
And that’s where we are unique, and that’s where we stand out. And that’s how we’re deploying this. But at the end of the day, our clients are partnering with us to be able to get to those levels of high accuracy and high adoption.
Surinder Thind: Thank you.
Operator: And thank you. And one moment for our next question. And our next question comes from Puneet Jain from JPMorgan. Your line is now open.
Puneet Jain: Hi, thanks for taking my question and thanks for doing the webinar last week. It was very helpful. Quick question on that. Like, are you seeing — like it’s clear like that AI and Gen AI is helping drive some of the new conversations, new use cases for EXLS and you are well positioned. But could it also be driving clients to take longer in deciding project awards like because it’s new like the use cases might be new, like you mentioned, Rohit, just now the accuracy levels required are also high. So could they be taking longer in deciding those project awards when it’s about implementation in their processes?
Rohit Kapoor: Yes, Puneet. I think there is a longer lead time in terms of decision-making by clients because they are looking at a number of proof-of-concepts and they are looking at experimenting as to where these technologies are going to be impactful and where actually the return on investment is not very high. So there is a fair amount of experimentation and doing pilots and proof-of-concepts that is there and the decision-making on enterprise-wide scaled up deployments of these solutions is definitely taking time. We are in a fortunate position because when we do these proof of concepts for our clients, we’re able to very quickly demonstrate our ability to create impact. And as soon as our clients see that, and that’s very visible and transparent to them, then the deployment of that across the enterprise becomes a much easier decision for them and a much quicker decision for them.
And so we are, I think, benefiting from an ability to kind of take some of these initial proof-of-concepts and deploy them across the enterprise. So I would expect in 2024, this is going to be something that is going to result in active decision-making by our clients and our ability to be able to implement and deploy this for them.
Puneet Jain: Got it. No, that’s very helpful. And the prior guidance assumed, I think that macro will remain unpredictable in the first half and potential normalization in second half? Particularly in Analytics. Is that still your assumption? And do you expect marketing analytics to stabilize? When you say that, does that mean it will stabilize on a sequential basis from here on?
Rohit Kapoor: So there are two parts to your question. One, from a macroeconomic standpoint, our viewpoint is that while the macroeconomic environment seems to be somewhat stabilizing, it continues to remain challenged. And therefore, our assumption is that the macroeconomic environment and our clients will remain cautious in terms of how they think about making investments and how they think about investing in new areas. As far as marketing analytics is concerned, that business for us has declined and — but it has a seasonality to it. And the seasonality is that marketing analytics is typically stronger in the first quarter and in the fourth quarter. And in the second and third quarter, it’s actually much lower. So we’re going to see that play out for 2024 as well. But calendar year 2024 over calendar year 2023, it would still be a decline.