Arlen: Welcome to the e-commerce marketing podcast. Everyone I am. I’m your host, Arlen Robinson. And today we have a very special guest Gorav Bhattacharya. Who is the CEO of Involve. Before starting Involve he worked in product for ASML and PwC. His background is in software & product development. At the age of 17, he started his first company Iti which created a platform for radiologists and helped them prevent female foeticide in India. Currently, Involve employs 40 people and is growing rapidly building products for distributed teams. 

Gaurav: Thank you for having me, Arlen, I’m really excited to be here and contribute and also connect with your listeners. 

Arlen: All right, great. It’s awesome to have you. And today’s topic of the day is going to be artificial intelligence and the use of machine learning and predicting customers.

You know, it’s really amazing how technology has advanced in such a way. And it’s awesome that companies such as yours are able to really leverage, you know, these advancements and helping. Businesses predict customers and effectively reach the people that they need to reach. And so that’s awesome. Things are a lot different than when I first started in the whole internet business, which was about 20 years ago, yet it’s a different animal these days.

But before we get into all of that, why don’t you tell us a little bit more about your background and specifically how you got into what you’re doing today? 

Gaurav: Thank you for asking that question Ireland. So a little bit about me, as you mentioned already, I come from a software background just growing up. I was really excited about computers.

I had lost my father to cancer and our Stu uh, had this single mom who was a blue collar worker. She worked super hard to give me and my brother a good education. And one of the things that she brought home, was it old windows? Computerizing. It was like in 92 windows PC, and I was really excited about it.

And I got passionate about computers, about coding. That kind of became an outlet for me and took on a lot of projects to bring in money to the family and just stayed excited. I had the opportunity to start a company with my co-founder. Samia who was one of the only, I think we met at a coding class when we were in high school and she was the only girl out of one 62 guys in that coding class.

So that was really funny. She has a great background. She loved coding. So we started coding together. After we did our first company, we had an opportunity to sell it. And then we moved to the U S of your boats from India. And the way that we stumbled upon this opportunity was we had another. Successful business that we were working on.

It was a software as a service company, similar to OSI, and we had a subscription model, but we found that it was really hard to predict your revenue. It was really hard to understand when customers are going to churn. What are the predictions for churn? And then what are the customers that are ready to buy more or expand?

That was a huge challenge at visa. And we built this like project in-house where it was a simple dashboard that took Excel spreadsheet, data, and it applied artificial intelligence on it to predict which customers are showing. Almost like an early warning system that are showing signs of churn that might leave you.

And then also bridge accounts are happy that we could expand into. So that’s kinda how the origination of this product and this platform became, and that’s what involvement involved AI started. That’s kind of our story of getting started. So we kind of saw a problem that we faced and we wanted to solve it and found that this is a problem.

That’s very common with others and decided to scale. 

Arlen: That’s awesome. And you’re probably not the first person that I’ve interviewed before that has told me the same story. As far as you birthing a business out of a need that you had directly you and your business partner head. And that’s really awesome because those are usually the types of businesses where the founders are really kind of all in, because you’re really.

Doing it because you’re so familiar with that particular problem and you’re able to understand it more and then effectively try to solve it, not only for your own company, but for others. So that’s awesome. And it’s definitely a subject that I’m very familiar with as far as being an owner of a SAS company and dealing with churn.

Yeah. It’s something that we battle every day and you were right. It is very difficult to do predictions with. Customers and finding customers that are at a certain point in predicting wind customers churn in it’s, uh, like you said, I’m definitely not alone in this most SAS companies deal with churn and doing accurate customer predictions as well.

So definitely a challenge that we all face. Now, you know, we really hear a lot about artificial intelligence these days. It’s kind of a buzz word as well as machine learning. So before we kind of dive deep, Why don’t you explain really what it is and how it can relate or how those relate to helping businesses grow and find customers.

Gaurav: Yeah, absolutely. That’s a great question then. And I’ll give a brief background about machine learning and artificial intelligence. Every time we think of artificial intelligence, just because if media it’s always a negative connotation that AI is gonna replace all jobs. And I don’t think it’s going to replace jobs.

It’s just going to create the workforce to become a little bit more smarter and the jobs will just relocate to something else. That’s what I feel. And I feel there’s also a time between now and for the next 50 years until AI becomes evil. That AI can really help us and really help us do our jobs better.

So that’s what I’m really passionate about. And the Genesis of machine learning and artificial intelligence has been going on for about a hundred years. So it’s not a new concept, but what’s happened in the, in the recent past. Is that technology. And computing has become so fast with everything in the cloud and us continue to create these ships that can process data really fast, that AI is now accessible for everybody.

So just to be back on, on what artificial intelligence and machine learning is, and deep learning is the concept is that we take data or we take historical data to be able to. Take certain and predict certain outcomes and why that’s happening based on the outcomes, the machine is automatically able to change the algorithms to improve the outcome.

So just as an example, it’s a, it’s almost like an Excel spreadsheet. The Excel spreadsheet says that round. Let’s say if I have 95.6 and I want the Excel to round the number for me. And it makes it 96 because it’s a little bit above 95.5. Right? So it makes it 90, 96 for example. But what machine learning can help you do is it can also treat the algorithm to say that if the output would be better, if I rounded to 95, then it would change the algorithm.

It would say that. Okay. Unless the number is. 95.7 and below it should become 95 instead of 96. So the machine becomes smart enough to understand what is the best outfit I’m trying to achieve and change the algorithm as well as whatever the algorithm is based on the historical data. And then also looking at the outcomes that we want to achieve.

So that’s the basic difference. So it’s just that the algorithms can change automatically based on the data you feed it. 

Arlen: Yeah, that’s a great example. As far as this spreadsheet is concerned, I think it kind of definitely paints a picture of what it is on a simplistic level. And now that we have an understanding of really kind of what it is in order of some of the things that you can do at a very basic level, how would this apply to assisting customers specifically, typically over businesses, rather e-commerce businesses, right?

Yeah. There find customers and grow their business. 

Gaurav: Absolutely. So Arlyn, the two ways that I feel that businesses can really leverage artificial intelligence is a, is for automation and B is for prediction. So let’s talk about automation. So automation could be all the manual tasks that e-commerce businesses have to do.

Everything from, let’s say, manual emails that they send to their customers, to customer advocacy programs that they might be doing from lead generation to replying messages, to answering support tickets, to basically make sure that the shipping goes on time, make sure that we have loyalty programs in place.

All of these things can be automated. If we have enough historical data that, that we were talking about. If we know. What are their expertise? How do we reply to customers? How do we find the right customers and the right personas? These things can be automated using AI, and it can also be done in-house but it also, if there’s no expertise is a lot of softwares now, similar to your platform, there’s the softwares that assist with any part of automation.

So you don’t need a person to manually do that. The second part of it is prediction. So prediction can happen in lead generation. You can predict which leads. Have a higher score attached to them, which leads have a higher conviction of, of closing. And then you can create a similar kind of audiences for outreach programs, or really understand the kind of audiences you’re attracting for your products or your services.

So that’s where prediction can really come into place, really predicting where the market is headed and which, what kind of products and what services should we be offering for which segment in the market for best possible results. So, yeah. I feel these are the two things that we can really leverage AI into so that we can improve our productivity and have better conversion and ultimately improve revenue for our businesses.

Arlen: That totally makes sense. And when you were describing that you were talking about, I guess the problem with a lot of businesses have, is finding the right customer because, you know, you can do some blanket marketing campaigns across a variety of different channels and places where customers. Are found, but if you really aren’t targeted enough for you, if you’re not really using your understanding of who the right customers are, then you’re definitely burning a lot of, of advertising dollars, for sure.

And I can see how artificial intelligence can help you with all the marketing efforts and analyze that data and then understand what types of channels you need to go after what. Ads should be shown and kind of the whole thing. It almost sounds a little bit like it does kind of fall even under the category of.

Conversion optimization, really conversion rate optimization because you’re utilizing these strategies to really just prove conversions where you’re helping a business find that right customer and then ultimately convert. Does that seem like that’s the kind of under the same case? 

Gaurav: Yeah, absolutely. I think you hit the nail on its head, basically.

It’s all about how do you improve lead generation? So how can you get more pipeline for your products and for the customers as well as how do you improve conversion? So your win rate improves and your customer expansion improves. Once you have a customer, how do you keep them longer? How do you keep them buying the products or services that you have?

So, absolutely. I think it’s a lot about filling the pipeline with more. Qualified opportunities and leads and customers that kind of would be the best fit for you as well as improving the conversions at different stages. I think that’s an opportunity really easily that companies can leverage AI for 

Arlen: now in your experience, you know, helping customers.

Do this, what are some specific strategies that you’ve seen or you guys have put in place to predict and really acquire the right customers? And, um, can you provide us some examples of maybe some companies that have used these strategies and kind of what they’re doing? 

Gaurav: Yeah, absolutely. So I can give an example that we do that can be extrapolated for a different business.

And obviously every business is different. So the KPIs and key indicators are going to be different for them. But for us, what we do is we have. A prediction model for pipeline and a prediction model for customers. And what we do is we look at a lot of key indicators, key indicators from what is the order size, how often does a customer order, what do they order?

What’s candida problem that they’re facing with our product. So we have a software product. How much usage do they have? We look at executives. Do we have direct relationships with executives over there? Is there any movement with executives happening? What have the customer been talking about on press releases recently?

Again, these are key indicators that apply to us, but a similar model can be built for e-commerce businesses as well. And then ultimately what ends up happening is you look at these things and when these things are changing, we build an early warning system to tell us that what are the top opportunities that can convert and what should be dropped?

Which ones should we not focus on as well as focusing on the top customers, which are the customers that have high risk right now, or they have a high probability for expansion or buying a new product or buying more products. So that’s what we have built through. Basically they key indicators fi feeds the algorithms, teaching it.

How the key indicators impact the likelihood of buying more or going to a different competitor. And then we just monitor that over time and the AI system can do that automatically and can do that. That’s a little bit more complicated, but in terms of AI products that e-commerce companies can use, there are some products that helped with better messaging.

So one of the products that we use. It’s called outreach.io, but that’s a product that’s used for messaging out to its users, making sure that we do proper messaging handle support tickets properly. There’s obviously a lot of chat systems that provide automation, but these are the two big spectrums over here.

So there’s automation and prediction. So our focus is on prediction, but automation can be equally valuable for a. Small business when we have their hands full and they really don’t have time to respond to hundreds of tickets that are coming in and also answer questions about their products all the time.

That’s something that could be really valuable as well. 

Arlen: Yeah, definitely. Yeah. I can definitely see what this small business owners e-commerce businesses. Um, as you mentioned, that are really kind of overwhelmed with their day to day responsibilities. And a lot of times communication is tough to be able to.

If I could, we communicate to all their costumers where you’ve got inquiry inquiries, you have problems, or, you know, you just have general, you know, messages that are coming in that have to be responded to. And I can understand where a lot of these responses are going to be similar. So that’s where I would see an AI system when she let me coming into play, because it can, it can see these.

Questions and inquiries and then respond accordingly and whole there’s a ton of, like you mentioned chat solutions, there’s these chat bots, which are really intelligent. And there’s so many of these days where it’s like, you, you go online. I was just, I’m logged in earlier to American express checking one of my credit cards and fall.

It seems like these days is they immediately use it. Some type of bot technology immediately with most of these systems and they don’t really engage. The actual live person until it kind of reaches a point where it doesn’t really know what to do or respond and then go from there. But usually I’m the type where I usually bypass it immediately and I easily type an agent.

And I think the systems can immediately understand. They all know, as soon as it sees that word agent or representative it immediately then pings one of their staff and then they jump in. 

Gaurav: Absolutely. No, absolutely Eileen and this, I think it’s, I also feel that these chat bots are overused or misused.

Because nobody can take away that authenticity of a live agent or someone helping you there’s there’s companies that, that we have seen where this automation process can be at different points of the process. Instead of at the front of the process, when an actual person triages a ticket or an issue that’s happening, they can then send it to an automated system for solving the query and then closing the loop on it.

Notifying internal resources. Let’s say if there’s an actual bug with AMX, a bot and Amex can create a ticket, can notify the engineers. They can use that information on when that issue will be solved to go back to the agent so that agent can get back to you. Instead of keeping all follow-ups and notifications, they can almost recommend the next best action.

So in two weeks, The bot could tell the agent that, Hey, you got to reach out to Arlin and make sure that you tell him that this issue is not resolved. And we were able to make some changes. So there’s a lot of stuff that can happen, not just at the front end, but also middle and the back end. But it all depends on how you want your customer experience to be.

Arlen: Yeah, it makes a lot of sense. And the ultimate thing to note here is that is I think the big advantage has been one of the biggest benefits is the time saving factor that these tools and solutions will provide where it doesn’t, these things don’t have to manually manually be done. Like you say, where you could have things escalated to an engineer at a certain point where, you know, there’s not somebody manually dictating at information and passing it on to where it’s automatically happens.

Gaurav: Yep. Absolutely. 

Arlen: Now you mentioned that you, uh, tools that can be used, there’s a lot of chat bot tools, you outreach, diet IO, um, that kind of fall into the category of this, uh, artificial intelligence machine learning and, and effectively automating a lot of your communications. Are there any other tools or resources that you could recommend to make all of this happen?

Gaurav: Yeah, absolutely. So there’s definitely a lot of tools that we use. Another one that we have heard of, which is really good, it’s called zoom info. But what zoom in for does is it helps you complete. Some of the information. Let’s say, if you have a form on your website and you have a certain lead information, you have their email, you have their first name, but let’s say you don’t have their last name.

So the system can automatically go search to it. And I can help you flag some of that. That’s one of the things, another one that I like to use or have heard about is easy ask, which is a voice search engine. So it basically helps like e-commerce stores use voice commands so that anybody who is. Coming on for your products, you can use a lot of voice commands.

They can use voice commands to find different products. Internal people can do a lot of things there. I know for e-commerce companies, there’s something called spark Toro. It’s S P a R K T O R O, which is for influencer marketing. It has influenced your intelligence. So it. Constantly updates, influential information looks at brand mentions, looks at which influencers are doing what so that you get updated information and you can reach out to them.

So I think the areas that I have seen the most is auto-complete search plugins. So like I talked about ZoomInfo over there. I don’t know if it’s only for e-commerce companies, but I know they serve a bunch of other customers. There’s a lot of marketing communication, automation, outreaches one, but.

There’s so many more they’re chat tools that are AI tools, his voice command APS that I was talking about, like easy ask. These are some of the ones that I’ve come across. I know I rambled on a lot, but if anyone has questions, I can obviously answer more and reach out and brainstorm on ideas too. 

Arlen: Okay.

Great. Well, yeah, that’s awesome. Yeah. Those are definitely some tools that, that I wasn’t familiar with. And I know aren’t going to be useful the listeners as well. So yeah, it’s good to know that there are a lot of kind of leading edge companies that are out there that are kind of leading the way with this type of a technology, because there’s so much that can be done.

You know, as we get ready to wrap things up, where do you see the future of artificial intelligence and machine learning, headed with regards to e-commerce and e-commerce marketing. 

Gaurav: Yeah. I really feel that it’s a lot about personalization for e-commerce. We call it like the fourth industrial revolution, AI too, you know, it’s coming, it’s going to be integrated in every facet, just how internet and software is now integrated in every single facet of every platform and every.

Company and every business, that’s how AI is going to be embedded. I feel for e-commerce a lot is to do with personalization so that everything that you buy and every, that everything that a consumer or a customer is getting is completely personalized for what they like, how they like it. That’s what I feel is the future.

Obviously, the. People who are in e-commerce businesses, running e-commerce businesses have know about the problems that can come up and that’s, what’s going to be solved first, but ultimately we feel, I feel that just from smart shopping, to really being everything personalized and while still having that brand feel and that having that brand, which is very personal, I think that’s, what’s going to be the future of 

Arlen: e-commerce.

Yeah. Yeah. I think you’re totally right on that for sure. One of the things that I was thinking about with regards to personalization, and I can definitely see this coming soon is you’re probably familiar with the e-commerce companies that sell these. What do you call it? Uh, kind of. Curated kits of different products.

There’s a lot of these curated clothing line types of companies or companies that they may not necessarily be creating the actual clothing themselves. But what they do is they curate a set of, let’s just say, men’s clothing. So, you know, a lot of these we’ll put together. Different styles and different fashions of let’s say pants, shirts and things that, you know, are fashionable these days that, you know, the average man would like.

And I, you know, I’ve seen a few of these and I’ve actually been interested in checking some of these out myself, but what I see as far as personalization and where these companies can kind of take it to the next level is as they get more data and as they start using AI and machine learning, what they’ll be able to do is analyze data.

For a bunch of different demographics have different type of men. And so let’s say for instance, I get an ad specific to me or our curated box with an ad specifically created to me based on where I live in my state, the climate of my state, my particular age and all of that. I mean, it’s all presented to me and I’m like, wow, that’s definitely something that we’re aware that really truly uses this data to personalize.

The ad and the messaging that they’re sending out to kind of a whole wide array of people. So yeah, I can definitely see that coming soon. 

Gaurav: Yeah, absolutely. 

Arlen: Yeah. Well, great. Gorav it has been awesome talking to you this, as we see AI machine learning is definitely where things are headed and it’s no slowing down for sure.

So I know our listeners are going to take all of this information and then see how they can apply it is their own businesses. For sure, because it’s, you’re at the point where, you know, if you don’t do these things, you’re definitely going to be the left behind. But what I always like to do is just kind of shift gears here with my kind of final fund effect question, just to close things out and switch gears.

So our audience can get to know you a little bit better. What’s one fun fact that you can let our audience know Bart. 

Gaurav: That’s a really question. So one of the things that I like to do for fun, I love playing chess. And what I like to do in my fun time is I always try to play chess against different computer systems.

And what I try to do is find. Any loopholes that the computer systems may have, which is really fun for me, it sounds pretty stupid, but I always try to like go back to like these old chess systems that were built. Let’s say like in early two thousands or even late nineties. And they’re these different games.

And I always try to find like these loopholes that they don’t, that the system wasn’t built to understand. What are the things I’m really excited about is like how AI is going to impact games and how AI influences everything. I know we were talking about NBA, our Lin, right? So everything is analyzed.

Golden state warriors always uses everything that, where should they shoot from? How should they have their place? They use a lot of AI and now a lot of NBA teams are using AI to make. Data driven decisions on which players, how many minutes who plays when obviously you cannot change that replace that system.

Every play cannot become like Jordan or LeBron James with that, but that’s really exciting to me. So I always look at like, Things that I can, things that impact the games and how AI and bags games, um, but for fun, I look at these old chess systems and see, where does the algorithm break? What, what did they do wrong?

And then try to beat that test system. 

Arlen: Gotcha. Yeah. That’s interesting, man. That’s cool stuff. Yeah. I recently started learning to play chess as well. I got to get back into it. There’s really so much to it and I can definitely see why. Looking at those old systems, there’s definitely some holes there, loopholes and 

Gaurav: yeah.

Arlen: Finding those ways to kind of get around those loopholes. I’m sure is exciting and kind of a challenge there for you. So, yeah. 

Gaurav: Yeah. It’s like bugs, right? Like it’s bugs everywhere because it’s software. So if, you know, sometimes if you’re ordering from Uber eats or any kind of online delivery system, you use like a old promo code, it doesn’t work.

You use it like a couple more times. It probably starts working like the loopholes like that. Everywhere. And that’s kind of what I, what I enjoy finding. 

Arlen: Yeah. It’s good stuff, man. Definitely. There’s always a way to break a system for sure. Well, great gruff. It’s definitely been awesome talking to you and thank you for sharing that.

Um, fun fact. Finally, if our listeners want to get a hold of you and pick your brain anymore about artificial intelligence or machine learning, what is the best way for them to get in contact? 

Gaurav: Yeah, LinkedIn is great. You could find me easily on LinkedIn, but also email. So my email is. G a U R a V. That’s my first name.

Add involve soft.com. That’s the company website and that’s the easiest way to get it. Get ahold of me. Okay. Hey, 

Arlen: great. Well, thank you for sharing that girl was, it’s been a pleasure talking to you and I hope our listeners take advantage of getting in touch with you so they can learn a lot more about how the artificial intelligence and machine learning can be applied to their business.

Gaurav: So thank you for joining us today. Thank you all. And thanks for your time. 

Arlen: All right. It has been a pleasure. Take care. Thank you for listening to the e-commerce marketing podcast.

Podcast Guest Info

Gaurav Bhattacharya
CEO of Involve