Editor’s Note: This transcript was created using AI transcription and formatting tools. While we’ve reviewed it for accuracy, some errors may remain. If anything seems unclear, we recommend referring back to the episode above.
Introduction & Guest Overview
Niti Sharma: Yeah. Yeah, absolutely. So, I mean, so well said. I think it, I think it also perpetuates so much courage within the teams, uh, right When you hear that somebody else made a mistake, they failed, and yet they are sharing it in an open podium. I mean, it gives everyone the confidence that they are also allowed to, uh, experiment and try out, and like, not just push their ideas under the carpet. So yeah, I think that’s a brilliant culture. And again, it also brings me to this thing of how important like culture is, uh, when you’re trying to, uh, experiment, right? Like you need to have a supportive culture, um, in the company for experimentation. Awesome. Yeah. Awesome.
Key CRO Metrics at Cashfree
Niti Sharma: Um, and, uh, one, one more question is like, when it comes to measuring CRO success, um, do you feel like there are two or three metrics, uh, at Cashfree, which are most important, uh, for the team, and, uh, why would it be so?
Vinayak Purshan: So, uh, the main agenda of CRO is to get conversion. Okay? And within conversion, you have a lot of funnels, right? So, for example, a person. So, uh, we, the first, uh, touch point of any successful CRO campaign is, did sign up the phone. Okay. So when she signed up the form, when he enters the journey, is there any drop off rates in the funnel stage? So this will give us a brief understanding of where are the, uh, where we should be pinpoint of friction points. Okay. Because, uh, as I mentioned in the previous, uh, uh, uh, this thing, uh, that friction points are not always bad, but it’s always good also. So if there are unnecessary friction points, we need to take it out. Friction points is good when it’s adding value to customer, when it’s giving trust to the customer. So that is the second thing that we follow.
And the final metric would be the acquisition metrics. For example, if a person has signed up, yes, clear all the funnel stages, but easy actually activating is account because just signups for how many transactions. Because in FinTech activation is where value created. It’s not just. That’s, that’s the ultimate goal for any business. So are the three main points that we, uh, into consideration.
Mindset and Experimentation Culture
Niti Sharma: Okay. Yeah, that’s, uh, that clarifies a lot of, uh, things. Um, do you feel like, uh, we were talking just a little bit about mindset before we spoke about metrics. Just going back a little on that. Um, you know, we have, uh, one or two eBooks that we had published, which have done really well. Like, these are really popular eBooks. Uh, we call them small test, big wins. Um, so, and they talk about, uh, tests which are seemingly very simple, but yet led to, uh, great results for different companies.
And, uh, uh, you know, uh, from whatever I have experienced, uh, so far, it’s like, uh, we either talk about things like this where simple tests lead to big wins, or we talk about like, you know, really big tests where like complex changes were tested. Uh, maybe a very important landing page was redesigned and put to tests and all that. But I feel like 70, 80% of experimentation would be still lying in between these two, uh, extremes. Like right, it cannot be like it’s always like this or, uh, like this and this, uh, to be able to sustain this, uh, maximum, like this majority of experimentation that’s done between these two extremes, you do require that kind of, uh, uh, encouraging, uh, mindset where you’re learning from failures and you’re, you’re approaching each test like, you know.
Uh, with a fresh, uh, mindset while also keeping, uh, your previous, uh, learnings in mind. Do you feel like there is some other necessary, uh, attitudes or mindset shifts that teams need to, uh, you know, get better at experimentation, but then these are actually overlooked very often. Do you, do you feel like, is there something like that that you’ve?
Vinayak Purshan: Yes. Uh, so when we talk about the mindsets for a successful experimentation there, two things comes to my mind. One is that we should always have a three for mindset. That we should have this in our mind, saying that not every test direct because the moment you think that, okay, all my tests are gonna pass, uh, so you will not be having the free hand in, uh, running any critical experience because, you know, you have fear in your mind that it might fail. And in this case, Cashfree has been very supportive. Uh, management has been very supportive that, uh, the moment we explain them. This is what we are trying to do. This is a success metric, and I’m thinking that 80% of the time my campaign might fail. I’m not sure about it. Our, our senior leaders are just back saying that, yes, go for it.
Niti Sharma: Hmm.
Vinayak Purshan: A failed thing, you’ll get a learning. We get it. So that is the first step. So we always have a pay for mindset. That is what is giving us an extra cushion, saying that, yes, we have a backup from our management saying that, okay, if something goes wrong. They are able to, they’re able to help us out. So that’s why. And second thing is we always very customer constraints. Sometimes marketer will take some metrics and forget that’s human on the other side, who are we trying to sell to? We’re trying to sell to the humans. That person on the other side of the table decision, right? Human not an OR bot who just runs fancy experiments. Some, uh, campaign, they’ll get promoted. So we, we, we generally make sure that we have an experiments because we have experiments often. So are the two main things that we generally follow.
Niti Sharma: Hmm. Okay. Yeah. Makes a lot of sense. Empathy led, uh, experiments. Wow. That’s, that’s fantastic.
VWO Platform Features and Benefits
Niti Sharma: Um, and other than, uh, mindset, uh, the, the other important things that I would say is the tools that you are using, right? The platforms that you are using. Um, how would you say like VWO, as an experimentation platform has so supported you in the way you run and scale experiment? Like, have there been, uh, specific features that you have found, uh, helpful or, uh, how has it enabled, uh, you know, the way you experiment at Cashfree?
Vinayak Purshan: Yeah. Uh, like VWO has been in this case, like it to run without having dependency is huge. Uh, because organization like we are on the tech part, any that we do go through a lot of testing because we should make sure that, uh, no therefore, or nothing wrong messaging should go to our users. So in this case, uh, we have a huge investment from our development team. Azure, at the end of the day, uh, me being a person, I know that, okay, this campaign would work. I can’t er and say that, okay, you’ll just change it because at the end of the day, you also need to have what they’re doing and why they’re doing it.
Which with the features that VWO is providing, say for example, heat session reporting or something like that. So, uh, I’ll do a use case that I have a button, uh, in my landing page, which is uh, which is a button, but it’s not looking like a button, so people are not actually taking on it. So I can’t go directly to a developer support. They will make the changes, but at the end of the day, they also need to get convinced, right. What are we trying to do? Because it’s a learning thing for them as welling purpose. It’s not. It also explains you the flaws for the, uh, the messaging that we need to follow or the guidelines that we follow as a developer as, so, uh, with this, uh, I can give them a proper sheet or a proper data saying that ac uh, these are the total number users, number sessions I got. These are the, I’m getting, people are not considering this as a, so they do learning for them well, and they’ll also be in spite to get things done. Right.
So this way, uh, we have made sure that like features of map session recordings has, you know, evidence and documentation to show and, and help align team understand what we’re actually trying to do. So yes, VWO has been enable for in getting this done. And yes, uh, also, uh, we, we, like you might be knowing that, uh, we, a small change on the landing page also needs to go through a lot of these processes, right? Which takes like unit waste, the pipeline waste like, uh, the UAT, that thing comes into everything. You make a small change, which I’m not sure about that I may mentioned that uh, CRO is all about like, we should have a physical, right? Like we, we dunno if it’s gonna be a successful event or not with this mindset I can’t invest my developers time into because they’ll bes. So the two functionality that VWO is providing that without a help of a can make changes on the realtime, uh, webpage and launch it to say 2020 5% of the user. And then if it is a successful thing, then I can ask my.
Niti Sharma: Hmm. Right, right. Absolutely brilliant. I think, uh, what you said about, um, the session recordings, heat maps and all also feeding into, uh, UI UX, uh, improvement actions, uh, improvement strategies is also, it’s, it’s really good, right? Because, um, uh, like you were saying earlier, like on the other side is, is a human and they come with a set of behaviors and, uh, it’s so important to listen to, uh, the feedback that they are giving to us through, uh, these features through heat maps, through even through rage clicks, like whatever they’re trying to tell us is so important that we have an open, um, ear for these things. Um, so amazing.
Multichannel Strategy and SEO Insights
Niti Sharma: I wanted to actually ask you like, um, uh, you have built like multichannel strategies and you take insights from various channels, I believe, right? Whether it’s keyword trends or content performance or, uh, you know, campaign data, and then you, uh, feed them, uh, into website testing, right? In a way that it drives, uh, innovation. Uh, how, how does this happen practically? Like, how do you take insights from all these channels and then weave it into experimentation?
Vinayak Purshan: Uh, we generally look every channel as a listening post. Say for example, I’m reading the right, so SEO tells me what people are actually searching for. Uh, uh, campaign the value proposition that can work. Similarly, when we talk about email or phone thing, that data tells us the language that people are actually responding to. So these three are, uh, giving us like valuable insights in its own spective.
So say for example, when we talk about SEO, I know that, uh, okay, there is a keyword, for instance, at mind, which is fighting in popularity. I could see the volume is searching, people are searching for it. Then we need to make sure that we use this trends, this data points on our actual testing purpose, right? So say for example, if instance is spiking popularity, I would make sure that, okay, I use this keyword on my homepage because I know the people are looking for it. And if that particular specific keyword is not on my homepage, I’m using on.
Niti Sharma: Yeah.
Vinayak Purshan: So this way similarly for paid also, so we know the value prop that work, right? Uh, and, and this is the way that we generally use. And also sometimes in FinTech, uh, we, we tend to go very super formal in terms of the language that we use and sometimes we might not work because it’s not always takes call on the other side. There are some small businesses also, uh, who wants a communicate. So this helps us understand, like say for, we have a number of blogs. Some blogs we write in a super formal language. Some we write in a normal language, uh, which help, which we can help user understand the concept in. And the moment we see that, okay, search traffic, people are actually interacting with that blog, we try to incorporate that with our landing pages messaging as, so that is how we, all the channels that we.
SEO in the Age of AI
Niti Sharma: Right. Okay. Okay. Very helpful. Uh, super helpful to know. Um, one more thing, um, uh, because we are on the topic of SEO and you’ve spent such a significant time there, um, with all of the AI now getting integrated into search engines and all the AEO and all of this coming into the picture, uh, do you have any advice for companies out there, like how can they, uh, still continue to get the best out of, uh, SEO in the current landscape?
Vinayak Purshan: So if you’ve heard new. And I’ve been doing for the last two years that I haven’t seen you going away anywhere. It’s coming. Okay. So I will say the nature of search is shipping past. Yes. With AI assistant coming in, the ranking are less visible. So because in today’s search for any given keyword on Google, uh, technically if you’re in the number one question, also you need to scroll one or two scrolls to get to the number one organic visit. Uh, the real estate for is very low, but it doesn’t mean that is completely okay.
So whatever topics that we say, right, the LLM models, everything is optimized around the principles of the moment. You have a solid, strong SEO uh, set up for yourself. Let it build back links. Let it be a site structure, let it be a technical structure. So these things are the one that contributes to the AI. So I would say stick to your of SEO, but have an I what’s happening in the market in terms of AI. Okay. Uh, when you want to rank for AI use, right, for models. But the trend that is happening is the more the band mentions you have. Pick up the, and what actually the brand mention, brand mention is the technical nothing. But in terms we call the more, the more you get from our brand side, a more side and your AI will start picking up and backing.
Coming from SEO and it has been in the industry for the last 10 years. So the biggest thing is just the shift mindset. Yes. The only one thing that I would say is uh, the intent from a keyword search to a transactional intent or a conversational intent keyword, that is the, so say for example, if I’m looking for a payment gateway, I might not rank because the payment key might have a 44, but at the end of the day, I don’t know what is the intent of the user, but if I’m actually for payment service, the only. The volume will be 10. The volume will be very minimal. Not only 1% of the actual payment gateway volume, but the intake is super high. If I’m searching for a Conversional intake, you would say, how do I set up a payment gateway website? That means I’m actually looking for a solution, and here is what the game is. You need to be, make making sure that your website is present. Not that, uh, if I’m asking for how set up gateway, you don’t explain what is gateway, how it works in the background, nothing. You just need to be on a pinpoint proper data with proper on website. You need to explain what is that you need, what is that, and how implement that that is.
Staying Updated and Continuous Learning
Niti Sharma: Yeah. Yeah, true. So well explained. Uh, Vinayak, I thank you for this. Um, you know, you said like so many things have changed and yet nothing has changed. And like SEO fundamentals remain the same. And if that is solid, like you can still win at, uh, SEO, uh, but with whatever, uh, keeps shifting at such a accelerated pace, how do you keep yourself, uh, you know, up to date with everything that’s going on in the industry, like is there some specific resources that you follow, some, uh, you know, some uh, rituals that you like to do on a weekly or quarterly basis?
Vinayak Purshan: Generally what with marketers generally do is we have a huge connect of networks, right? So, so nowadays people are more on sharing data, I would say sharing their knowledge and experiences that we have. We have a lot of forums. We have a lot of groups where people invite us and subject experts and they will try to explain what they have done. In fact, those things are like an actual learning, uh, mind that you have. So, so for example, I get enrolled in multiple digital marketing tools where, uh, a company X has run some campaigns and they will be, uh, very happy to share the success that they have done, the success campaign that they have created and also what went wrong.
These are some of the learnings that we need to take. And also there are multiple, uh, resources available on the nowadays, right. So where there’s a article for journal, everything, like multiple LinkedIn post where subject matter expertise are posting their experiences. So, other article, uh, it’s a constant learning thing. You should always spend at least minimum 30 minutes to one hour to learn new things, read about articles, what’s happening in the market. And they can be like in your as well, like in like instead of just around or doing something.
AI’s Role in Marketing and Experimentation
Niti Sharma: Hmm. Hmm. Awesome. Thank you so much for sharing that. Uh, Vinayak um, and I think, uh, I think one last, uh, question on this, uh, topic, uh, would be that. Payments is a dynamic space, and AI is influencing marketing and experimentation very dynamically. Uh, the pace is only likely to accelerate, right? The pace of experimentation, uh, with, um, AI coming in more and more in this, uh, landscape. So how do you see, uh, marketing or experimentation within marketing evolving, uh, over the next few years?
Vinayak Purshan: So yes, AI has been very helpful. Right, because earlier wanted to do any sort of experimentation, we needed a team of three, four people where one person who do the research, where one person who draft on a campaign strategy, everything with AI coming in, those things are done within minutes. Like AI will.
Niti Sharma: Hmm.
Vinayak Purshan: Simulations and even personalize experiments in a real.
Niti Sharma: Hmm.
Vinayak Purshan: Where we had uh, two, like running near experimentation. So I just use AI to help me understand what is the study I need to do. Like this is the vision, I have the results I wanna achieve, and AI will help you out on that. But along with that, it’s not completely board driven. It’s not completely, uh, AI driven. You also need to have a trust layer marketing in specifically where communication will need to enforce that. AI personalization, but a human more than because AI doesn’t understand human sentiment, right? Messaging, but end need to understand because you know your audience better than AI or something like that. So that is how experimenting at a speed of AI is one. And also layer marketing where you should not hundred percent rely AI, but you should always have one pitch add.
Niti Sharma: Wow. Yeah, that’s, it’s. It’s tricky, and yet it’s so important, uh, right, because it’s so easy to get carried away into this AI wave and I, you know, improve your testing velocity and so much promises coming from AI, but, uh, to be able to do that, leverage that ride, that wave, but also keep that, uh, trust layer intact, like you said. Uh, that’s fantastic.
Personal Learning and Final Advice
Niti Sharma: Um, and, uh, is there something you’re currently learning Vinayak, like whether work related or on a personal, uh, front?
Vinayak Purshan: So every day is a learning for me because, uh, now you mentioned like I’ve doing for the. We try to leverage multiple things. Say for example, uh, how do I get my visibility improved on LLM models? So for example, in CRO also, so we, we, like I’ve been for the last five years in my previous organization also, but the things that used to work during that time is not working right now. So how, uh, my experiments are working fine. So every time a learning thing, learning through us, like, yeah.
Niti Sharma: Okay. And, uh, any, uh, final thoughts or message that you would like to, you know, share with the audience, uh, before we conclude this CRO part of, uh, the podcast? Any, uh, like final advice or thoughts that you wanna share?
Vinayak Purshan: I would say that never hesitate from doing any, uh, big chart experiments on CRO because nothing is a paid that CRO part. So you campaign either you win, you get good business from it, there is no loss in it. So I would say never shy out on running any experiment. But yes, take a cautious, okay. Uh, don’t roll.
Niti Sharma: Hmm.
Vinayak Purshan: Out percent of your audience into it. If you are 90% sure that, uh, this might not work, use the VWO’s audience segmentation, uh, feature you have, right? So say you 400% user, just say you cannot source running only 10% of the user add guard rails because guard rail, one of the features that has, right? Because an extra of running critical experiments without impacting to a higher level.
Niti Sharma: Hmm. Yeah. Yeah. Awesome. That’s awesome. Uh, advice.
Personal Segment
Niti Sharma: And as we get to the tail end of the, this podcast, I wanted to ask you like, what is your favorite way to, uh, unwind after a long day to end, uh, like a long, tiring day at work or, uh, is there something you do for that?
Vinayak Purshan: Yeah, like I love riding, right? So after like eight, nine hours of my, like, people generally say that, okay, commuting in Bangalore hassle.
Niti Sharma: Yeah.
Vinayak Purshan: But I enjoy riding my bike.
Niti Sharma: Oh.
Vinayak Purshan: That helps you relax, right? Like, uh, you, you forget everything. You are not connected to your digital assets. Your phone is in your pocket. Traffic can use myself.
Niti Sharma: Awesome. And that’s so much needed, right? That sort of detox. I think we, we all, uh, really need it. Awesome. Awesome.
Closing Remarks
Niti Sharma: Uh, we wish you, uh, the very best, uh, with that and, uh, sincerely hope that this dream, uh, comes true. I’m sure it’ll actually Thank you. Vinayak, I thank you so much for taking out time to talk to us, uh, today. I had a lot of fun talking to you and, uh, I genuinely learned a lot today also. So big thank you. Big thank you for that.
Vinayak Purshan: Uh, experience working at CRO.
Niti Sharma: Thank you. Awesome. And, uh, thank you to all the listeners, uh, who will be tuning in for this episode. Thank you everyone. Thanks. Bye. Okay, we’ve started. Awesome. Hello everyone, and welcome to Behind the Wins by the VWO Podcast. I’m Niti and I’m the editor in the marketing team at VWO. I’m super excited today to be speaking with Vinayak Purshan. Vinayak is associate marketing director in digital marketing at Cashfree Payments. Cashfree Payments is India’s leading payments technology company, empowering over 800,000 businesses with fast and secure payment solutions.
At Cashfree, Vinayak spearheads website management, SEO leadership, digital marketing innovation and content strategy. He combines expertise in social media, paid acquisition, analytics, and various tech platforms to build integrated data-driven marketing systems that strengthen Cashfree’s presence and also accelerate business growth.
As part of his journey, Vinayak and his team have also been leveraging VWO to embed experimentation into their digital strategy and unlock meaningful wins. I’m very happy to explore Vinayak’s CRO experiences and his overall journey. Hey, Vinayak, welcome to the VWO Podcast. How are you doing?
Vinayak Purshan: I’m doing fine. I’m perfectly fine. So thanks for having me. It’s a real pleasure to be here and joined the conversation today.
Interview Content
What Makes CRO Exciting
Niti Sharma: Awesome. Awesome. Um. Uh, has there like been some, some sort of moment in your CRO journey, which has made you feel like, you know, yes, this is exactly why I love doing this?
Vinayak Purshan: Uh, so as a marketer, uh, and as a CRO person, right? So there is one thing that, uh, as far as me like experiments, like, so the moment you do things that you already know, it doesn’t excite you even in your career and also in your life, right? So, we go for a bike ride and we don’t know how the road ahead. So when you get to know that, okay, it’s a simple straight road, uh, no photos, nothing, it doesn’t excite you to drive, right? Same thing applies for marketing as well.
So when you do CRO, uh, we we’re actually playing with the user sentiment, right? So we are trying to understand what the user is actually looking for and each person from a different point of view. So that is what excites me because the strategy that works for me, uh, for an X audience would not work for an Y audience for me. So that is what will keep you to be on your toes and try something, uh, exciting day in, so that is the video.
Balancing Experimentation with Security in FinTech
Niti Sharma: Yeah. Yeah. Wow. Yeah, I mean, well said. Really. I think the whole, uh, study of, I think human behavior behind, uh, the behavior on the website or the app is, is it’s so important, uh, right to diagnose. I think, um, so I wanted to actually ask you, uh, because you are with Cashfree, uh, payments is such a sensitive space, right? Because users come in expecting a lot of speed. Uh, they expect trust, they expect security, right from the very first interaction that they have with this digital property. So while you also want to boldly experiment, uh, in CRO, how do you also ensure that these fundamentals are protected?
Vinayak Purshan: Yeah. Uh, so, you know, uh, when we talk about speed, trust, and security, right? It’s not only a concern about the FinTech space or a payment space that should be like our main risk before any business.
Niti Sharma: That’s right.
Vinayak Purshan: So, so, yes, absolutely. Uh, FinTech is a very sensitive space, but yes, we, we, we do take it very seriously, like. So our, our main, uh, foundation is always around speed and security. That’s non-negotiable for the Cashfree. So when we experiment with CRO at Cashfree, so our mantra, simple optimize without compromise. So just to make sure that we wanna optimize something, we don’t wanna compromise on our standards, so we don’t wanna on our values or the use piece that we’re trying to say.
So what this means is like, uh, say we’ll never trust something that is having the risk per. For our client, uh, it’s like at the right. So we can’t, uh, try out multiple things over here if there is any risk with, uh, state or the payment that they’re trying because they trust, they’re trusting us in terms of payment. They’re allowing us to manage their payments at the end of the day for their business. So, uh, some of the experiments that we would never go ahead with is like hiding the trust batch, changing the critical close, slow down the load times because payment needs faster and speed, right? So we’ll not do any experiments that would impact our speed and also their payment speed or something like that. So make sure that your website is trustworthy in terms of speed, in terms of security, and also make sure that you have all the details mentioned right in the festival. So that is what we do.
Combining Qualitative and Quantitative Research
Niti Sharma: Hmm. Yeah. Excellent. Yeah. Yeah, that makes so much sense. Um, the other thing I wanted to talk about, like we, we, you know, we started off this conversation by talking about user behavior and understanding user user behavior, and there is so much talk about, you know, combining qualitative data and with, uh, quantitative research also. So like, uh, data from analytics and funnels and your traffic numbers, and then combining that with qualitative research from heat maps or user feedback and all of that, uh, before you decide what to test. Right. Um, and while all the sounds very good in theory, uh, for someone who is doing it practically, like, could you, uh, share how easy or how difficult it is and like how does it work practically? Or how does it, uh, ultimately convert into a more impactful experiment for the website?
Vinayak Purshan: Yeah. Uh, so we generally take a layered approach where, uh, quantity to data tells us what is happening.
Niti Sharma: Yeah.
Vinayak Purshan: Analytics, we have a number of tools that use us quantitative data. So we generally see, uh, say for example, given or if I’m doing an optimization for any given page, so what is the rate of it? Uh, people, the page, uh, they satisfied the page because rate would give the engagement rate of that user, right. So we, we successfully landed a user by a different source. So we might get them via SEO, we might get them via pay, we might get them via social media ads, something like that. But at the end of the day, the bond rate is very high, that we know that our ads and our approach marketing approach was very good, but we are not selling the same thing on our website. So that will give us a quantitative data, right? So that is, that is what we, we know what, what is happening.
So when qualitative research, based on this quantitative data, we know why is this happening? So, for example, we noticed that a high exit rate was there on our driving section via analytics. So we saw that, uh, okay, there were a good number of traffic coming in, uh, but the bounce rate was very high. The section was going down. Uh, people were not engaging with that page. So, so number help us out. Say for example, we know what was happening, but we never know why is this happening. That is when we layered the approach with heat map and reporting. So, okay, when we, the heat map, we saw users. Percent of the are not able to find the place we actually want land or section.
Niti Sharma: Okay.
Vinayak Purshan: And we generally go through on a weekly right, get the, uh, actual use case of the. So why we saw this and then we made sure that, okay, this is the area where a user is not able to understand what is it trying to look for, the messaging that we conveyed on the ads part. So that is how we, the experiment, we made sure that we clarified the copy and so result this approach was, yes, there was a lifting, ation, uh, conversation would be like a final metric for us. But yes, the first win for us was we saw that the was going up. And people were engaging on those pages, people were interacting on those pages. So that is how we, uh, take the mixed approach of quantitative, qualitative. And generally we say that, uh, the magic happens when we marry, uh, what with white. So the both aren’t work individually. So that is how we take the.
Niti Sharma: Right. Right. Um, and you said like you are looking at, uh, session recordings on a weekly basis. Right. That’s so, uh, that is part of your overall process. Is that correct to say? Like, and you, you start with looking at quantitative numbers and then you look at, uh, how does that whole method if that’s okay to share?
Vinayak Purshan: So, so use this feature, uh, on a video basis just to understand what’s actually happening because FinTech, in FinTech, right, so the traffic and session also works on a seasonality basis. Say for example, now. So most of the C brands who wanna have a payment, uh, or a multiple payment gateway for their business, right? So that is where the traffic flow will be high for us from PC brands or other brands. So then, then we, we don’t wanna lose that value, right? Say for these are very valuable customers for us. So leveraging or seeing this data on a weekly level, use us, uh, uh, uh, uh, HD. What should we do? What should actually be done, uh, so that we take a proactive approach.
Learning from Failed Experiments
Niti Sharma: Right. Excellent. Yeah, that makes, uh, so much sense. Um, and, you know, while this happens, uh, while we are experimenting, um, not every test really goes the way, uh, we expect, right? And sometimes it also happens that the ones that don’t do, uh, so well end up teaching us something. So has there been an experiment that has maybe failed or, um, you know, but surprised you still, and it kind of changed the way you were probably thinking about user behavior or, you know, your, uh, approach to like finance websites. Anything that was like a little bit of a surprise for you?
Vinayak Purshan: Yeah, there are a lot of experience there. Plenty of experiments that.
Niti Sharma: Wow.
Vinayak Purshan: We a marketing head, right? We a marketing and we know that, okay, this thing would work and we go with confidence saying that, okay, we’ll create a campaign. We’ll run the experiment on that. Okay? 90% because most of the time it fails because generally no one is perfectly understanding the user sentiment. And that is where. Come into picture. So yes, there has been a lot of such incidents where, uh, we thought that this would be a successful campaign and it failed miserably. And there were certain campaigns which we are not sure about. Uh, we thought, okay, this might not work. Only 10% of the success rate is there in this campaign, but it worked out to be a best campaign that we ran so far.
So yes, one, one example is that, uh, generally, uh, when you are dealing with, uh, form filling or a optimization part, right? So generally people say that, uh, a shorter form can split up is, that is a hypothesis that every business have. The, the shorter the form you give, uh, the more, the lesser the friction point is the conversion that we have, but that is not the case with FinTech. Sometimes the, the moment you reduce the form size, uh, you, you play with your trust part because it’s not only about filling the name, phone number. We just want to understand, uh, what the user is actually wanting for. What is the need from, so those friction points, the all friction points. It’s not a hundred percent true, it’s not always true, but in, we have seen that the more friction points you pay, we get quality. It’s not just the junk that sitting on your database where sales team needs to go through all and fill it up.
Niti Sharma: Yeah.
Vinayak Purshan: So the more approach, like a specific approach we take here, it’s easy for our sales team and also good for the business that we are driving them quality rates and uh, the conversion is high.
Niti Sharma: Yeah. Yeah. Wow. Like, this is like gold. Uh, because, uh, I, I can imagine like, uh, you know, even the listeners going like, oh, how did this happen? Because, uh, yeah, I mean, I think the general consensus usually is that, uh, the shorter the form the better would be the form submission rate also, right? It, it would help people fill it quicker and there would be less hesitation time and, and all of that. But I think just goes on to say that actually, like there is nothing like best practices anymore, right? Like you cannot, uh, really put uh, anything like in a box, like a best practice because it’s going to differ from industry to industry. And then even within industries also, I think from company to company and maybe even like, which part of, uh, the website and what is the intention of the form, uh, and, and all of that. So yeah, this was a brilliant example, I think. Um, super valuable.
Documentation and Cross-Team Learning
Niti Sharma: Um, and do you have like a process of extracting learnings, uh, from your experiments so that they can then feed into like your long-term website, uh, strategy, uh, or even into your, uh, next campaign? Like is there a, a method or a process that you follow?
Vinayak Purshan: Yes. Uh, at Cashfree, we generally don’t create experiments as campaign. It’s like a.
Niti Sharma: Okay. Ah.
Vinayak Purshan: So for example, every test that we running is document. Hypothesis, results and insights, whether it’s a win or a loss. Because to be very honest, CRO is not a onetime activity. I might run an experiment today, uh, or a festive campaign six months down the line. I might also, uh, there might be another team member who wants to learn from the mistakes that we made in the past or from the success that we drive in the past. So we, we own this. We make sure that this is shared across the team, across the marketing team.
So that, let’s say, for example, we run a campaign, we know how this worked, how the campaign was successful, but this is used as a learning curve for the product marketing aspect because product marketing team will understand what is the USP that they, they want to build, right? And people who, who gives you the messaging? What needs to be done? What is the product all about? Who we wanna sell. Digital marketing is a channel that we, we use to get traffic, but I think.
Niti Sharma: Yeah.
Vinayak Purshan: And with these experiments, yes, they, they get some insights, they get a very valuable insights, and that is how we make, we maintain a record of all the campaigns, whether it’s we record everything.
Niti Sharma: Right, right. So I think I get two things, uh, from your answer. One is like how important a role documentation plays, uh, in CRO, how important it is to, you know, have a record of, of everything. And the second is also that, uh, digital marketing or, uh, you know, the team responsible for conversion rate optimization cannot work in silo. Uh, right. There has to be, uh, it has to be a communication flow between various teams and insights are not just for the digital marketing team. It is these insights are for various, uh, teams, right? Like, um, again, like is there some mantra or, or something, some philosophy that you guys follow to make sure that, you know, various teams are collaborating for experimentation?
Vinayak Purshan: Uh, generally we have a practice at Cashfree that every month we have a all hands meeting specifically for the marketing team where we celebrate small wins, where we celebrate awards for our peers, like if they have any exceptional job. Because that’s a way of encouraging the audience also, right? So in that way, we, we, we we don’t ate events. We also highlight the losses and what would having done, so say for example, like CRO as it’s an experimentation part, right? So we should have a mindset that not everything we passed in. So I failed from something. I failed a campaign. So we, we communicate that to the team saying that, Hey, this is what we tried. This is what we, and this what impact we could have done this in. So this way, uh, even all the teams will get a sense. So there is a saying that there mantra people say, learn from your past mistakes. Because people tend to do mistakes, because nothing works as you expect. And we want our team, our peers, to learn from the mistakes that we did because you can’t always learn from doing. So you also sometimes need to learn as aspect. So that is what we call and we communicate.
