Achieving CX Excellence: An Ongoing Journey with Ankur Agrawal

Speaker A: Welcome to Humans of CX, a podcast

Speaker B: powered by ozone 10.

Speaker A: We share the latest insights and customer experience from industry experts to help you humanize your approach, placing empathy at the center of the customer experience. I'm your host, Karam.

Speaker B: Thank you so much, Ankur, for joining me today. Welcome to the Humans of CX podcast.

Speaker C: Thanks Garima. Glad to be here, Ankur.

Speaker B: Before we dive into the details, let's understand your journey and what really inspired you to get into the space of customer experience.

Speaker C: Yeah, so I actually started my career in tech. I spent a few years in tech, then branched into product management and business leadership, especially after my startup and a travel guru. And I've done business operations, so I have experience across functions. Right. Customer experience actually happened by chance. Right. So after Next Education, when I talk to people. Right. Then, um, one of those was Rohit, who's the co founder at Stamp Deal. Right. And they were looking for somebody who has cross functional experience to come in and set up this horizontal function focused on improving brand advocacy and customer retention. So, so essentially a uh, broad strategy, customer experience role and in my experience running the, let's say the flights behind the travel group or business operations at Next Education. Right. What I realized is that there's a huge amount of growth in long term customer value that you can unlock by focusing on brand experience. But most of the growth firms in India don't even have a clue, right. So they're not even trying to. So people are not aware. Right. So the two focus on the transactional behavior. Right. So I acquired so, uh, spent so much money on acquiring the customer and spent so much money on the transaction, people don't have a view on the ntv. Right. So, and it was also very clear that it requires, uh, cross functional problem solving. And the levers, especially in the Internet forms is typically the levers are productive. So found it very interesting. I said, you know, why not? That's how it started. Right. And after that, over the last three, four stints, it just become, once I took up that role, I realized how much of a huge growth impact true experience can have.

Speaker B: Right.

Speaker C: Ankur?

Speaker B: Um, when we say customer experience and you are someone who's really passionate about this space you've been here for so long, what is it exactly that you have been trying to improve for the customer? And when we ask this question, we also go back to, you know, where exactly are the gaps and how the customer has evolved over the years, you know, in terms of their priorities, in terms of their expectations and you know,

Speaker C: behavioral, uh, patterns so there are two different questions, right? I'll address the first one. Right. What do I mean by experience? So very broadly speaking, things that cause customers to not do all their business with a particular brand. And on the flip side, things that we do that cause people to become loyal, that is at a very high level what I mean by customer experience. And typically when I've done it in the multiple brands that have done this, now we do this through this tool called Promoter Score. There are obviously the devil isn't always in the details. Lots of firms use nps, but a lot of them use it wrongly. Some do NPS survey at the end of a customer support interaction. Even though the NPS is a, the NPS question is a brand question, it is not a transaction. Some companies do it at multiple stages of the customer interaction and then they decide how to stitch it together. But done right, and I don't think we'll have time to go into how to do it right. But done right, that is the code bench. So NPS is essentially telling you how many of your customers are uh, actively saying bad things about you in a social setting and how many are promoting you in a social setting. So essentially that is a very strong indicator of advocacy as well as retention of ship. So that is broadly what I mean by customer experience. And your second question was how have I seen expectations evolve over time? So most of my career has been in growth businesses. So uh, Internet driven, B2C businesses that have been growing very fast because they've come up with a new offering that didn't exist. So most of the time people have been okay just getting that new product or service. Right? Because it didn't exist. So the initial part, when going very, very fast, when the business is growing like 200% year on year, it's growing because it's bringing a product that didn't exist. So till the growth starts slowing down, customers are happy with the base offering. But once there's a strong competition for the customer or the market starts saturating, then you start seeing what the customers are expecting. So like for instance, let's say the ride aging space, OLA and Uber competing, right? There was huge growth for a long time, but in about 2018 it started segmenting so the growth slowed down. Then it started becoming important. Customers started saying, okay, Uber is better than O lines these things or OLA is better in these things, so I will go here or there, right? So it started informing choice, but broadly the expectation. My sense is customers don't expect, they don't want Anything dramatic from brands. So a lot of people talk about the uh, delight. I think that's wrong. People are not looking for delight. People are saying, you have a brand promise to me, just fulfill that promise. I am happy. So don't do bad things, just do what you told me you'll do. And people are broadly very happy with this. So like for this sexy company, be reliable. I need reliability from you, right? So if I book a cab, I'm going to, I have to go to my office, right? In 10 minutes, give me a gap and don't cancel it and that's it, right? Their expectations are broadly very basic, right. They don't really care for, let's say you need uniform, the driver or uh, great smell in the car. Obviously bad smell is not good, right? So the delight elements, I think a lot of businesses think of you. People don't care about it. People just say, just do what you promised to do well and you'll get my loyalty.

Speaker B: Right? And when you uh, rightly pointed this, and just in that context, there's uh, like one example from the customer standpoint when we are trying to book a cab through Uber and if I feel that there is a delay and the driver is not moving, I immediately switch to Ola for that, right? I have a feeling that, okay, let Uber take its own time. Meanwhile, I'll just book something via Ola and if I get a, uh, ride through Ola, I'll cancel the Uber ride right? Now in this uh, scenario, how do you look at this? Because as you said, the customer expectations is very basic, right? I need a ride, help me get a ride. That's all. So how should brands look at it? Like this is not exactly competition. There's not something dramatic that Uber is trying to do or Ola is trying to do. But it's just simple basic things that we are not able to get that right. So how are you looking at this?

Speaker C: So I start with very, very strong bias free survey. So you have to, let's say an NPS or a similar survey, which you do at the right point in time with the right trigger ask. You have to spend a lot of time in survey. But it's all first principle. You don't have to do uh, a course in marketing or customer survey to do it right. So you do it right. Customers will tell you what they care about. So I'll give one example because examples illustrate much easier than abstract points. I'd make my trip when I set up the NPS program and we ran it and we did different for hotels and holidays and flights in hotels, 70% of our detraction. Detraction is people who say, no, I am not happy with you. Right. I will not recommend. 70% of the detraction in the hotels business was nothing to do with make m my trip per se. It was because customers had a bad experience at the hotel. Now, going in, the general wisdom at make matter was we are in the business of helping people book tickets or, uh, book hotel rooms. But the fact that 70% of your detraction was because of a bad stay at the hotel, you have to internalize that and accept that. So basically what the customer is telling you, no, you're not in the business of booking tickets. You are in the business of giving me a great travel experience. Whether you believe that or not is up to you. If you will not believe it and solve for it, uh, I will go somewhere else. So the first step is figuring out and listening to what the customer is seeing at scale. Luckily, B2C Internet phones, we have scale. We have thousands of customers. Right. PAYTM has crowds of customers. So we can get a lot of input. So the first step is that figuring out what they want and then figuring out how to give them that. Obviously there's a lot of problem solving and structural thinking and so on that comes in.

Speaker B: Right. Truly listening to customers. I think that's what we are looking at and in this context only. Ankur and you also mentioned customer surveys, right? How reliable you would say that these, uh, surveys are and why it's important to really analyze them, assess them and then come to a conclusion or design a strategy accordingly.

Speaker C: So there are two kinds of surveys brought, right? So there's the proactive one. NPSs should typically be proactive. It doesn't wait for a trigger. You do it periodically. And uh, that is reactive. So reactive is, for instance, customer support score, CSAT scores and CSAT surveys. So one is reactive. We should not draw brand level proactive conclusions from reactive service. So customer support driven surveys, for instance, are useful only within the CS context. So that's one very important thing that reactive surveys will never give you the brand picture. The second part is surveys exist, right? So going in. So I've done this across four or five brands. Now when I went in, whatever surveying was happening had lots of issues. They were triggering issues. We were triggered at the wrong time. They were bias issues for because of the design, there was an inherent bias which people were not doing it intentionally, but they were not realizing the questions you're asking are wrong. For instance, most brands, the CSAT question is, how was your experience with the agent? That question you'll get an answer to the question you ask. So you know there will be scenarios where the agent was very helpful, very polite, but if they didn't really solve my problem, it doesn't matter. So you'll get great score. A lot of brands, CSET scores are like 80, 90% meaningless because the reality is never that good. So surveys are. I think customer service, customer voice is extremely, extremely useful. In my view, that should form one of the core elements of a business strategy. Unfortunately, most places they do it wrong. So that's one point. The second point is most businesses, whatever voice of customer they get, they look at it in isolation. Whereas the true insights and the focus comes by joining these scores or the voice with other data that you have about the customer. So in transactional ecom businesses, for instance, you look at CSET scores or NPS scores when you combine your transactional data. So for instance, just to take a random example, I'll go back to, let's say make my trip and say five star hotels, what is the NPS versus three star hotels, what is the NPs? What is the NPS in business cities versus le cities and so on and so forth. You do that, then you get actionable insights independent of what the text or the actual what the customer said. Just the score itself. You combine and break that into cohorts, you'll get actionable insights. And the VOC itself, what the customer is saying. Typing, you have to be very careful because people will respond to what you ask. Right? Having the right questions is extremely. Again, I'll take one example and at any point, Garima, interrupt me if I'm going to look.

Speaker B: No, no, no.

Speaker C: So like I struggled with my teams a lot in support. I said, when you get a negative C said, I don't want the quality team to tell me why we got it. I want you to call the customer and ask them what was the result. We had to iterate a lot before we got to the right question. So people would say, you gave us a negative cset. What can we improve? It's a wrong question because when you say what can we improve? People will give a list of stuff that you can improve. What you really need to ask. And it takes a lot of skill in probing is when you called us, you had a certain problem, you wanted us to do something, we did something else. Please tell us what you were expecting versus what we did. Once you ask that question, then the answer is very useful. Otherwise people will just Take what are things for improved. So, very long answer. But I think voice of customer and customer scoring is very critical, especially at scale.

Speaker B: Right. I think that's a brilliant point. And just to give a basic example from my own experience, you know, if I order something through swiggies, suppose the food quality is bad, or, you know, I get a different order and then there's a help section. I try to contact support. What they do is they'll try to contact the restaurant. They'll try to tell me that, okay, can we give you a coupon for this and this amount? What you don't understand is you actually wasted my time. Right. I waited 40 minutes for that order to get delivered. I was hungry. I was waiting for my food. And after like 50 or 55 minutes, you're sending me something which is not good. And then you ask me, how was your conversation with the support? The person was polite. Of course, I will say it was 10 on 10. But you are right, it didn't solve my problem. Right. That one hour is gone.

Speaker A: Yeah.

Speaker C: And you're not getting the insight. Right. You're thinking, my customer support is doing a great job because you're focused on the customer support agent.

Speaker B: Right.

Speaker C: Your question is to evaluate them, whereas your question should be to evaluate the brand. Right.

Speaker B: Also moving on, Ankur, we are looking at so many technologies. We're talking about generative AI, not exactly knowing how and in what ways it will improve our operations. We are still discussing a lot of things, but amidst all of this, there is a fear amongst people that AI will take up our jobs. And how is it going to be? How are you looking at it? Because technology is something at the core. People also expect it to be very inclusive and accommodative. So how are we going to navigate through this is a question that I think senior leaders like you can answer because you've been in the field for so long. You know, when a, uh, technology comes into the picture, there are, like, hundred steps that you need to follow. Right. In order to implement that. So how would you explain this?

Speaker C: So I think, see, while generative AI is a, uh, pretty revolutionary technology, basically we are talking large language models. But I think there is always. This is a tech cycle. Right. Every time something new comes in, there's a lot of hype around it. This is, in my view, this is similar. There will be significant improvements, um, to how we do CS based on generative AI, But I don't think the big movements will come from where the people are currently thinking they will come. So if you look at the majority chatter around AI and customer support, it is, oh we'll do so much automation and um, adcount will come down. This is not the first time self service automation technology has come. We've had bots before. We have self service platforms like freshdesk and others which are essentially where a customer support team without using technology, people who can come in and create their workflows. So AI is not a dramatically better automation technology. So therefore I don't think it will make a dramatic difference to the share of self service because headcount impact will only happen when share of self service goes up significantly. Uh, product agent productivity goes up dramatically. Otherwise why will there be an agent headcount impact? So I don't think that's the big thing, especially in the Internet forms. I've been mostly working in Internet forms. Right. So we have tech, we all have significant self service already. AI is not going to change that. And um, even outside of the Internet world I don't think there are any brands today that have very little self service. So it's not like you're moving from a pure contact center driven support suddenly to automation by. Right. I think the main benefit of AI, the biggest challenges in CS today from a customer experience point of view are not on automation, they are elsewhere. Like one of the challenges people faces with brands is I don't even know how to get support. So you put all sorts of AI if they don't even know how to get support, you're not going to solve anything or the quality of the solution is bad. The quality of solution, even AI is dependent on what you feed it. So even if your question is knowledge based driven, so let's say you know, I'm not getting a sound from my tv either agent or AI, they are limited to the knowledge base article that you've given them. But I don't think there may be impact. So like reduction in escalations is a huge use case. So we have so many cases especially in B2C. So somebody writes on Twitter. Yeah, no, I reached out to their support. The agent was useless. They did not do anything. I've been trying, I've called them 10 times. Right. What nonsense. Right? And these are very damaging to brands. But this is so easy to basically use LLMs to basically proactively identify cases that could potentially escalate and proactively handle. So I think that is like the one major use case. Another big area where I think AI will impact is in generating actionable insights at a structure. So based on all the conversations that you have, all the knowledge base you have, figuring out what to focus on what is potentially wrong is relatively easier with AI. And um, the third is in general self service becomes easier because you can actually generate code also from text. So you need your developer productivity increases massively. So whoever is writing the platform or the self service independent of writing a bot, just generating the self service itself becomes much cheaper and therefore people will do a better job. That's my sense. I don't think there is a direct impact on job. If anything it will help support the folks do their job better.

Speaker B: Right? And also Ankur, we are talking support, we are talking experience and contact centers are ah at the heart of this conversation. Right? So if we talk about contact centers from your experience, uh, from your journey so far, how have you seen the role of contact centers evolving traditionally or earlier? Previously they were referred to as cost centers, but now we are looking at contact centers as treasure trove of conversations, right? Customer conversations and get as much insights as possible. So how are you looking at the role of contact centers in modern context?

Speaker C: So uh, I'll answer that in two different ways. How am I seeing it evolve and how do I think it should evolve? So the first question, the answer is that unfortunately I have not seen any change. See a lot of people say this is like critical and this is brand impacting and so on. But the practice is not. It continues to be a cost center. The CFO keeps pushing the contact center head of the sport head every year to say boss, I don't care how much business is growing. M right. I will not give you one money one rupee more than last year. Even if the business is increasing 60%. So 60% more customers who transactions mean there will be 60% more support means. So it is largely still considered a person. The focus or the way people run contact center has not changed. It is per minute billing or per resource billing. So they keep thinking about going to tier two, tier three cities. So I am actually very disappointed with how things have been moving fundamentally because people don't understand the value even if they say so. Where I think it should move is fundamentally you have to start thinking of support and contact centers outside in and customer first rather than inside out and process the most meaningful brand experiences. And I'm sure you're a consumer, right? You have brands right? In your interaction you realize that brands which are more customer focused, their support agents will give you much better support answers and resolve your problems. But most brands continue to be process driven, right? So somebody will Write down if this is the issue customer comes with. This is the process that you have to follow. And the entire ecosystem is built around following that process. Right. So the Q1 team is scoring you on process adherence, and your schools are process driven. Your everything is process. Whereas no amount of process can actually take care of everything that the customer might face. And typically, things go wrong. When they go wrong, that means your process is broken. So I think that is the one thing that has to change. And, um, like, Amex is very considered, like, the gold standard in customer support world today, but most people don't take the right takeaways. So they'll say, oh, we'll hire 90,000 rupees agents, and they'll. They'll change everything. The big change that Amex did was exactly this change. We focus on the customer, not on the process. I think fundamentally that's the only major change that contact centers. And then obviously there's other things that you do on top of them, so self service and channel choice and all of that. But within the contact center, I don't think fundamentally anything should change beyond just this.

Speaker B: That's very striking, Ankur. And I'm, uh, sure, like, people will, uh, like to dig deeper into this and try and understand where are the challenges, where are the gaps? Exactly. And since you have led customer experience and so many brands that are like household names, like paytm, Karo is what we say, Right. That's part of our life. So when we talk about a product like paytm, what does it take to, uh, you know, really design an experience? What do you have in mind when you're thinking about a product that will capture, like, it's not like 10,000 people or 20,000 people you're looking at.

Speaker C: Right.

Speaker B: It's an entire nation that you're talking about people. PAYTM Guru is like, a part of our, uh, everyday life. So how do you look at this and what is your message for people who are coming up with products like these? Brands like these? And they try to capture the market, but what is it that they should keep at the core of their initial inspiration, idea, strategy?

Speaker C: Yes. It's no real rocket science here, honestly. I'll go back to just saying, understand the customer, listen to the customer. So listen to the customer. Doesn't mean that you just do what they tell you, because people will tell you solutions to their problems. You have to understand what their problem is and devise the solution. So essentially, look at what customer is trying to achieve. What is their goal, and how can you make it more easier and easier. People use paytm fundamentally to make payments. So how can you make it easier and easier for them, both in innovation? Because there will be things that you think of that people have not thought of. But also when they say this is my problem with payments, how do you solve it? So I'll just take one or two examples because you talked about btm. So it's largely UPI driven payments today. One of the challenges people have with UPI payments and starting PayTM specific problem with a UPI problem. Let's say you finished a tech seed I and you've tried to make a UPI and then the response you get is your payment is pending. M Please wait. Now the customer doesn't know what has happened, the recipient doesn't know what has happened and they're stuck because let's say it's on a taxi. Right. The taxi driver will go away, you will not get a chance to talk again. Right. The contract has to be closed there and then. But if the system is not going to be able to tell you for 48 hours what happened with your money, what do you do? If you make another payment then and that the first payment has succeeded, then you've paid them twice and if you don't make another payment then the taxi driver might never get paid at all. So now this is a huge problem. Huge meaning in percentage terms it is small. But because of the volume of UPI payments and the frequency that we do, we've all seen this, we see this, I see this like multiple times a month because I do like 100 UPI transactions. So you have to solve for that. So whoever is doing this, like ptm, this was one of the big items for us in upi. See how do we fix this? Because this is one of the things, if we do this well and we do this better than others, then the amount of loyalty we can generate is just phenomenal. Right. If I solve for this without the UPI ecosystem solving for this and I market this well, then I for one would never go anywhere else to my upi. So I think it's basically just looking at that. What is the customer trying to achieve? What are the key attributes for them? And then how do you solve for it?

Speaker B: Right.

Speaker C: Pretty extended product management thing.

Speaker B: Yeah. And I think PayTM Soundbox was really uh, one step ahead in that direction. Right. Merchants were confused whether the payment has been made or not. And it's like sure shot confirmation that your payment has been received. And it's, I don't know, like whenever I make a Payment and there is a sound box and it says like, PAYTM pay chalice, uh, receive hue. Hai. I see that smile on that vendor's or merchant's face and I'm like, okay, that's how happy he is with that announcement. Right. He doesn't have to look at his phone or anything and it's just done. Yeah. And even as a Paytm user, I think it's like an entire ecosystem that they have built. It's an experience. You can book anything, like movie tickets, travel tickets, everything is there. Like, I don't need to go on five different apps to, uh, do anything when we are talking about payments. Right. So that does make my life easier to a large extent. And I'm sure there are many more developments on the way, so I think that's pretty exciting. And Ankur, just to conclude, I would like to understand what does it mean to be a human in cx, to belong to a community that is constantly trying to help improve something for the customer and help the brands that are trying to deliver personalized experiences, hyper personalized experience for the customers every single day. How do you feel about this journey so far?

Speaker C: It has been really fulfilling for me and largely because I've seen this make an impact on the business and the experience. So what drives me personally is the ability to make a dramatic difference in whatever area you do. So once you put in the right metrics and then you see the metrics move and you know that this will impact the business growth. So ultimately businesses are exist because revenue and profit, but you can do short term revenue and profit by not doing the right thing, but structurally wrong long term, you do the right thing for the consumer. That's the best way of generating long term growth. And that has been the fulfilling part, fixing things, helping businesses generate this growth.

Speaker B: Right, Right, Absolutely. And just one last question, Ankur. Now since, uh, we talked about your journey, your experiences, what message do you have for someone who wants to build a career in customer experience? Because you yourself started from a very different background. Right. As you said, that was not the core, uh, dream like getting into customer experience. You've been an engineer and you've been at Infosys, and how has that been for you? And what message would you have for people who want to have a career in customer experience? And why I'm asking this question to you is because I love the fact that you're drawing very basic examples and telling us that fix the basics, fix the fundamental, and then talk about high level stuff. Right. So this is again a very Basic question that why should they pursue a career in cx? And if at all they want to pursue a career in cx, what is it that they should pay attention to?

Speaker C: Right. So I think it depends on who we are talking to. So there is CX and CS as we've said in the beginning. So customer experience is a strategic hunt. So the people would probably even think about CX are people who probably coming from a marketing background, a product management background or a business background or uh, strategy background, those are the people that should be considering and why should they get excited about it is because of the ability to create sustained profit and growth for the business. And when you talk about CS though, customer support, that's a very different set of people. And somebody, let's say who's starting their career, say as a contact center executive, you know, their ability to influence how support is done in their OR is very minimal. I don't think they will be able to even if they have the right minds. So for them I would say think carefully about which brand do you want to join? Because if you join a uh, process driven brand, then you will not really get a fulfilling experiences. But if you join a brand that is fundamentally focused on customer experience, then you'll get great fulfillment. You know, your day will be good, right? You'll be happy at the end of the day and if you do the right things then your growth will be fast. So that would be my advice.

Speaker B: Absolutely. Thank you so much, Ankur. Thank you so much for the great insights. It's been lovely listening to you and learning from you and I'm sure our uh, audience and listeners would like to connect with you on LinkedIn because I see that you're very active on that platform and we would uh, love to share your uh, LinkedIn profile as well in our show notes. So thank you so much once again for joining me today on Humans of cx.

Speaker C: Thanks for having me. Gareen.

Speaker A: Thank you for listening to Humans of cx, a podcast brought to you by Ozonetel. If you enjoyed today's show, visit ozonetel ah.com to learn more about how our robust omnichannel communications platform makes it the industry leader within the customer Experience experience space. You can find Humans of CX on Apple, Spotify, Google Podcasts and other platforms that are featuring podcasts. Don't forget to subscribe and share. Thank you so much for listening.