Strategies for developing the data-led enterprise: Q&A with EXL’s Ankor Rai & Gaurav Iyer

I recently met with Ankor Rai, the Chief Digital Officer and Gaurav Iyer, SVP and Head of Advanced Digital Solutions at EXL, a data analytics and digital operations and solutions firm. The company recently posted its Q3 2022 results, reporting a sizeable 24.5% YoY revenue increase from last year. CEO Rohit Kapoor noted that demand is coming from data-led solutions. Let’s break down what “data-led solutions” mean for EXL, how Ankor and Gaurav see the industry shifting, and how enterprises could rethink their approach to transformation to start with the data.  Our research shows that fast-growing firms are far more likely to setup centralized analytics COEs and work primarily with third-party partners, showing a correlation between topline growth and focus on data and analytics.

Your operational friction problem is actually your data problem

HFS: Gaurav, would you say EXL’s approach to operations transformation is process-led or data-led?

Gaurav: At a core level, it’s about removing friction from customer experiences and operations. Businesses want to hyper-personalize and remove friction, and they need to automate operations and remove manual processes. Because of this, the front and back offices are converging into one. In EXL’s experience, you can’t remove friction without addressing the issue within operations.

The other learning for us, thinking about operations, is that speed is the unifying currency. Before, if you did transformation at the backend, the business case was about cost. The numerator was cost reduction. What’s worse, trying to do transformation on the CX side, people would say, what does a 2% CX improvement translate to? At a global insurance agency we work with, they have created some standard global service level metrics: respond to every customer complaint within five minutes, all customer information must be available 24/7, and claims must be resolved in 24 hours. Speed becomes a fundamental metric of a transformation program. Therefore, a data-led approach is important; without data, you can’t solve data or ops friction.

A data-led approach differs from traditional transformation programs, starting with discovery rather than narrow business requirements

HFS: Completely agree. In our recent study, we found that the expectations for running operations are changing, and more enterprises expect to redefine the scope of their services with third-party partners around enterprise-wide data flows in the future (Exhibit 1). But we’ve been talking about digital for a decade now. Why have these data initiatives failed so far, and what is EXL’s vision for operationalizing data?

Exhibit 1: More enterprises want to organize BPM services across enterprise-wide customer, employee, partner, supplier data flows

Sample: HFS Pulse Survey, 2022; n=604 Global 2000 enterprises
Source: HFS Research, 2023

Ankor: Our new data-led transformation with clients is a four-step process (Exhibit 2):

• Transformation is not about client requirements but data access. It starts with undigitized metrics— dark data that might be in people’s minds. Before you begin any process, you need access to this data.
• Next, start with the problem you are trying to solve and create insights at speed. An example of this is when we were working with a big asset management firm that unsuccessfully deployed automation for 50 types of customer transactions. The tech was perfect, but the problem was that no one had looked at the data around all those transactions and created an insight out of it. Three types of transactions, including 401 rollovers, accounted for over 80% of the volume. Looking at implementations, we routed the work manually and solved for that 80%. That’s how we turned this program into a success.
• End-to-end adoption—For any program to be successful, you need change management. You can’t just implement technology. You need to change the entire process.
• Continuous evolution—One of our clients is a transportation company. If you’re a large retailer and want something moved from this warehouse to that center, you need to call someone and make an appointment. Our client has been trying to improve the CX of this appointment-setting process for a long time. They were failing before we came to them with an iterative evolution process. We said, “Let’s solve the email channel just for one of your retail clients.” Then we scaled that to another. Then we scaled to their website.

Gaurav: EXL’s approach differs from traditional transformation, which starts with business requirements. The core point is that you can’t begin there without looking at the data. If you’re trying to do process transformation, you must change the whole process together. When doing data-led transformation, you can change some segments. In an analytics process, you do test controls, proofs of concept…today’s transformation germinates from analytics, not technology.

Exhibit 2: EXL’s data-led approach prioritizes access, speed, iteration, and adoption

EXL’s approach to helping clients build the data-led enterprise: Continuous storytelling about access to data, and the importance of domain + data

HFS: A lot of companies aspire to be data-driven. How is EXL helping clients figure out where to start?

Ankor: The problem isn’t that you don’t know what to do, but accessing and organizing data is time-consuming and requires strong methodologies. Further, insight requires domain knowledge, so to find value, you must marry data and insights with domain. This is the new category of operations and using data.

Gaurav: A client in Tennessee had a collections problem. No one buys capability or concepts anymore; you need examples. So, we just shared our experience with another client who found it was taking too long to collect. For that client, we went in and started looking at customer operations. The typical process-led transformation answer is to put more people on the job and call customers more. If you get good at that kind of optimization, you might suggest calling three times in XYZ segments or calling certain customers at, say, 6 pm for best results. Instead, we started with the data and found that it was billing anomalies related to an inability to pay. It wasn’t even a billing mistake but a difference in the bill itself. We proactively sent notes to those customers to say that their bill is changing, and the result was that we reduced collections and associated costs.

The Tennessee CEO, upon hearing this, said that this is precisely my problem… every single CEO and CXO realizes the problem they have isn’t “tech–people–process,” just data they can’t access to power their organization. It’s resonating. It’s not a vision we’re selling but real stories we’re sharing about putting a data lens on operations.

HFS: Enterprises will need to rapidly gain data and analytics capabilities to go on this new journey of data-led operations transformation. In a recent HFS study surveying more than 200 finance leaders, published in partnership with EXL, we found that fast-growing firms (with >15% top-line growth) prioritize setting up analytics centers of excellence and working with external partners to get up to speed (see Exhibit 3).

Exhibit 3: Fast-growing firms are rapidly organizing capabilities, including CoEs and partnerships with external service providers

Sample: HFS Survey, 2022; n=207 finance and accounting executives across Global 2000 enterprises
Source: HFS Research, 2023

The Bottom Line: Becoming a data-led enterprise isn’t an overnight transformation; to get there, clients need to see peer examples more than ever. EXL is set on showing the way.

It will be no easy feat for EXL and its clients to turn over a traditional business services dynamic that rewards cost and productivity measures to one that prioritizes data flows. Ankor and Gaurav see the shift as creating a new category of running operations using data. At HFS, we see it as the evolution of the business services industry. We term it the potential move toward “business data services” from “business process outsourcing.” Regardless of how the terminologies pan out, the idea behind powering operations by better exploring the underlying data is here to stay and one EXL is excited about.