Ratnadeep: Okay so is this what we call impactful development? If it is or if it is not, perhaps, what does a culture of impactful development mean to you? If you think about it as a concept,t what does this concept mean to you? Can give an example with it, it will be helpful.

Advait: So back when last year or so, I was reading this book called ‘Principles by Ray Dalio’. Incredible Book. Ray, I think he’s one of the top performing investors like the hedge fund and just the General market. And a large amount of his philosophy of how to actually grow a company and how to actually manage teams comes from the ACT fact that as long as you’re honest about everything that you’re doing, as long as you are very straightforward with how you speak, how you do, and there’s no correctness associated to you, it’s a lot more efficient. 

Impactful development is on the same idea which is just do things that create impact. The impact doesn’t have to be like, Oh I just grew the revenue of the company from hundred dollars to a million.’ That’s fine. That’s still impact, that’s a good thing. But sometimes impact can be, ‘Hey I just spent a year developing this tool that is probably not going to be used anymore’. A year is a long time, let’s say a quarter. ‘I just spent a quarter developing this tool that probably is not going to be useful but I know what we shouldn’t do ever.’ Or figuring out what the customer needs exactly and developing for that niche.

A large amount of how you do development has to be centric on the customer that you’re trying to serve. If you’re a startup that is just starting up, which has zero customers, just choose one or two of them and figure out okay if I have to basically create a product just for that customer, how do I do it? What impact can I make on that person’s life? How can it change the way the business is done to create a positive flow or a positive net flow down the line? 

That is all that means for me, at least. That’s what impact-free development means. Just do things that push the company forward, pushes your team forward, pushes you forward as long as it’s not illegal. Honestly, every sort of thing that you do is sort of impactful in certain ways, as long as you learn from it. As long as the team learns from it, and as long as it adds to the knowledge, base of the team or the company. 

Ratnadeep: You touched upon a very important point. I think it’s an interesting segue to my next topic, which is customer centric approach. So aligning engineering with customer needs is key to probably your philosophy as I can understand. Can you walk us through how you make sure this alignment happens? Perhaps with an example of a process that works well for you.

Advait: (Sorry I just sort of yanked my mic out of the way)

So a large amount of what the customer-centric approach is, it’s just talking to your customers quite literally. I think DoorDash first started this program back in the day, in which the cashers or the software developers in DoorDash would literally just deliver food. The CEO still does that. Because that’s when you understand what the issues are. Customer Centric approach doesn’t have to beat this massive framework of you have to do everything right or you have to do everything with the customer in mind. If you want to understand what the customer needs, just speak to them. Just literally email them. Get them on a call. 

Just text and whatever. If you’re small enough, you can text your customers. If you’re a SAAS company, you can do that. Just do it. Just talk to them, ask, ‘Hey what makes this platform better for you? What makes this product something that you cannot live without? What is that one feature that it misses? What is that one feature that you can’t live without? Develop that product, add that product. See if that customer is still using it. See if that customer still feels the same. Do it for one customer, then do it for 10, then do it for 100. Then figure out what customer profiles you have. If you have all students, figure out if they are engineering students or art students. Find out the materials. 

Semiconductors are a big thing today. Just figure them out. Figure out what every single one of these niches needs. And then once you understand exactly what that person wants, just develop for them. Developing niches is not a bad thing. Developing for one customer is never a bad thing as long as that one customer can now be converted into two. That two can be converted into ten, then two hundred. Because if you find, there are seven billion people in this world. No one is alike. As much as we say everyone is unique, no one is that unique. As long as you don’t micro-focus on just one customer, and as long as you understand what is it that makes your product stick for that customer and just iterate on that, and make sure that you have a significant customer base, you’re fine.

The company that I’m working on right now, it’s all CV. It’s sort of like a social networking platform for creative fields. So be that your video production, your film production, architecture, anything that is not programming basically. Because a large amount of these people rely on visuals instead of a resume, which is paper-based. 

Networking them, using LinkedIn or quality work for that matter, which is like a new platform now, it’s not really that useful. Because visuals are not really the mainframe of these platforms. And while we are developing this, and we recently just launched, we literally did cohort analysis. We had 10 friends that we bought here. 

(I don’t know if I’m allowed to say on this podcast, but they’re adult, thirty-plus. So everyone was of legal age)

We bought beers for them, we bought them dinner, bought them to sit in one room, all of them. We were like, ‘Hey this is the platform. You all are one of a few of the top people in LA who do illustrations. Some of them work at Netflix and all of these places. Just tell us what you feel, tell us what this product makes you feel like, tell us if you would use something like this, tell us what this is missing.

It was a two-hour dinner drink slash cohort analysis. Let’s just call it that for business purposes. But we understood that we lacked six or seven different features that are crucial because that is what people use Squarespace for. There’s a reason why SquareSpace and Linkedin cannot be merged together. Because they just are like so separate right and for a creative person. If you want to build social networking enter SquareSpace or SquareSpace into social networking. You need those certain features, you need that accessibility and ease of moving things around and making the pages seem like they’re just created just by you.

You want things that are up in the forefront because visuals are important so you want to make sure that those things pop. We’ve built a very simple platform that didn’t do any of this. It just took us two hours to understand everything that was missing, two and a half months to develop every single feature, literally the date we developed and released that feature. We called those same 10 people again and we called 20 more people because in the two and a half months we figured out okay why not. 

Let’s just get 20 more people, a cohort of 30. We told everyone we will buy your bear the next time maybe after you fundraised or something. Just help us out again, just create another profile, and see what feels right. Everything that you’ve done until now has been deleted. If your friends are fine with it, just do it for us. The one caveat of doing this sort of approach is that in this way you learn a lot as long as you have friends, especially those who are not always nice to you. As long as you have that one friend who will say the truth to your face exactly the way it is, you’re fine. Because you have to start from somewhere and even as you grow today, we are sort of growing. And that’s a good thing. But that also means that you lose that personal touch with the customer because know away from the customer. Because they are beyond your reach. 

In that case, just find the customers who either come back a lot for using the platform differently than you thought using the platform. Figure those people out, find someone who lives in New York, because again New York and India are the big creative field industries. Find someone who lives in New York, find someone who lives in La, find someone who lives in Chicago. From all the people that are using your platform, ask them why they use the platform, and what is so useful for them, because they’re living and working in a different industry. 

I asked my friend who does game development and I was like, ‘Hey, you literally cannot use this because you’re under NDA and stuff. It’s the companies that you work with. Just use this platform anyway, just go and do your stuff from your undergrad. When he was learning game development, he did that. He was like, ‘This is what works for me, this is what doesn’t work for me from a technical person’s perspective who is in design.’ which is again a creative field but is not really a creative field. Because you’re working with Unity and what Ubisoft thing was called anyway Unreal Engine. Figuring out these customer issues and then aligning yourself to allow them to succeed by using your product is the only thing that matters in a customer-centric approach. Honestly, I think products that are built that way fail a lot and iterate a lot because again you’re sort of navigating through what is right about the customer and what is wrong about the customer. Because customers have wild needs sometimes, which don’t really translate well outside that one person or two. But it’s okay.

There’s a reason why you get stories that are iterating out of the same topic. LinkedIn has stories. No one really used it. They moved it but they thought it’d be useful because some niche would follow or they would wait. And that’s okay for them to experiment. Because what’s the worst that’s going to happen? You’re going to have to roll back some changes. Who cares? You’re changing your code after every six months anyway. 

Ratnadeep: So customer feedback experimentation, those are the key. Now let’s talk about some challenges. So leading engineering teams come with their own set of challenges, especially in today’s dynamic markets. Can you share a story about overcoming a significant obstacle? And can you also describe probably a project where the technological hurdles were complex yet you managed to create something that truly resonated with the customers?

Advait: So when you’re building any sort of team, the first thing even before technology comes into the picture is their people. They’re human beings. And I know this sounds funny or rude or whatever but being a human sometimes is just tiring. Because every single person has their needs, every single person is different. And a large part of just building a team is figuring that out. A large part is even before you actually go into the technology of it. Even if you go into the market, just figuring out and building a team that sinks with each other, that is always on the same page, that understands exactly- if I’m saying something they get it before I complete a sentence, that is a great thing. If you can do that you’ve achieved half of your issues. Because then you know they’re on the same page. They’re in sync with the knowledge base that the company has.

The other thing that the team has to be is a child. They have to be able to move quickly. They cannot be the people who are too slow. Especially in the startup world in which the market changes every six months, in a world in which you don’t even know if your product is fit for that market. You have to have a team that is aligned for it. You cannot have a team that is like, ‘No. I need to perfect this one front-end thing.’ 

If you have that team you’re probably losing out on days, if not weeks of development. Because you are trying to perfect that one small thing. Just release it. What’s the worst that is going to happen? People are going to say, ‘Hey, it sucks!’ Great! Why does it suck? You have an answer to why. Just keep on iterating until that why just end, until the whole idea of it becomes ‘This is not bad’ start popping up.

In a cohort of 100 people, you just need 20 people to say this is not a bad thing. From those 20, you need to find 20 friends or evangelists. Get those 20 people to bring another 20 people right so you get a very multiple-level marketing scheme. 

But that is how products grow. The only reason why you use Instagram is that one of your friends uses Instagram. The only reason why he used Instagram was that she used Instagram, as opposed to their friend who used Instagram. That is how you grow a product. You just make sure that your product is in the vicinity of people who would use it and you have one person who is shouting that product out for you. That’s just the Evangelist that you want. The only way in which you can do that is by making sure that your team is ready for it. If tomorrow the market changes and suddenly it is illegal for you to send out emails without whatever reason, it’s not ever going to be that way which is assuming, your product and your team are ready with different things that they can do instead.

A major example of this could be something at the company at which I was the director of engineering, which is Buena. We were quite working part-time for a thing that we saw from the data that we were receiving after six or four months of development. It was that a large number of people would open their emails but they wouldn’t really click on them, which is worse than them not opening the mail. Because why are they not clicking on it? And we tried a ton of different things. We tried A/B testing, seeing if a certain design appeals to them. 

Or just to give you context, we’re now working in a creator economy space. So for a creator’s email to not be opened is surprising. For example, let’s just say you have like a favorite YouTuber or an Instagram artist and you received an email from them and you are not opening it. We sort of wonder, why? Why is that happening? Or if they’re opening a lot of it then you also wonder why is that a random thing? Is that a thing that where that data is corrupted? Because if the industry average is at two percent and you’re suddenly getting fifteen percent open rates, you’re sort of wondering why. That shouldn’t be happening. That’s stupid.

And that’s okay. That’s not a bad problem to have in any case but we used to see the industry average people opening their thing, but not industry outages things. No one was clicking on the links, even if we made links frontend center, even if they were highlighted properly, even if the text was proper. That sort of got us thinking that maybe the email approach is not the right thing. Because when you think about it, for a large number of creators, most people assume they’re trying to sell you something if they’re sending you an email and if they’re seeing a story about something. That’s great because they’re creating content and you love their content which is why you follow them. 

But unless it’s a niche, which is only for fans, or when you have some subscription for a service, you don’t really buy anything from an influencer. Unless it’s their makeup line or an Amazon sale that they’re getting into. And we realized that maybe that is an issue. And again this comes with a lot of maybes. Because by this point, we had about 200k monthly active users, approximately. So figuring out the exact answers becomes slightly difficult if you’re a small team. So you have to prepare your team and be like, ‘Okay, the next one and a half months or six weeks, we are just going to keep on iterating.’ And they’re figuring things out. 

We’re going to keep on testing small changes, we’re going to keep on testing alternate solutions. One of my team members went, ‘Hey, why don’t we just send them an SMS? Everyone in this world sends you an SMS, from your bank to like a random real estate developer. Why don’t we just create an SMS thing in which it’s like- hey it’s your favorite creator, I’ll now text you. Who wouldn’t like that?’ And we’re like, why not? 

(Sorry, I don’t know if I’m allowed to curse YouTube) 

But why not? Who cares? What’s the worst that’s going to happen? You’re going to lose five dollars in testing, will you? It doesn’t matter. Just create an account and test it out. We chose a small group of people and asked them to provide us with their phone numbers. Half of them did, and some of them did not. The ones that did, we just tested out with them. So did they actually like it? Turns out they do. Because it comes directly to your phone and you generally check your messages at the end of the day or at the beginning of the day. No one really checks their messages when they’re working. Most people don’t.

So they were just clicking on the links that we were sending them. What this also allowed us to do is create a situation in which not everything had to be a sales pitch. Some things can just be something like, ‘Hey! I’m really excited to let you know that today is my dog’s birthday,’ which is an SMS that actually went out. ‘It’s my dog’s birthday, just send me all your wishes on my Instagram stories.’ It’s a personalization on an SMS that feels different than a sales pitch because now you have a line of connection which is one way.

Don’t get me wrong. No one else could reply to them and the influencers couldn’t read those messages. But suddenly you had messages that were going in that were slightly more personal but they also could send in stuff like, ‘Hey! My Amazon wishlist is on sale,’ or ‘Hey! Here’s my Amazon shopping page. Why don’t you buy something from it for yourself? or ‘Hey! It’s Father’s Day. Why don’t you buy Father’s Day gift for your dad?’ 

That changes the dynamic drastically between an SMS and an email. Because you send your emails to your professional colleagues, you send an SMS to your best friends. That sort of dynamic we wouldn’t have really had if 

1. we did not have a customer-centric approach, in which you were speaking to our customers and figuring out, ‘Okay, is it something that they like? What they don’t like?’ Why are SMS or emails not working? Or why SMS is not working? 
2. If we don’t have a data-centric approach, which is, ‘Hey, why is something not working? Let’s just experiment and see if something changes with it. 
3. If we didn’t have a team that was really ready on their heels all the time and who thinks ‘It doesn’t matter, let’s just fail but let’s just try something.’ That’s all that matters when you’re trying to overcome market changes.

Ten years ago, emails were the king. Then Slack came and email sort of lost their thing. SMS is becoming a thing now because everyone’s sending you emails. I have 500 unread promotional emails all the time. I don’t read them because I don’t care about emails. I just get them from everyone because I’m using the same email ID every time. But no one really knows my phone. Or only a special few people know my phone number. So I read all my texts. My text boxes are zero and I think that is the way you overcome changing markets. Because the only way you can do that is just to keep swimming. The minute you stop swimming, you travel.