Uday: Hey hi Rohit!
Rohit: Hey Uday. How are you?
Thanks, guys, for joining. I think a lot of people are here. Just to recap, this whole conversation is about, to build or buy? This I think is an eternal question in the product-led growth world and to talk about this, I’ve invited Sushma Nallapeta. I am waiting for her to join in. Give me one second.
So, here’s what we’re trying to talk about, just for the recap of the session, right? The conversation is all about, hey, product-led growth is the new momentum that’s been building for a long time now. What should we do? Should we build the tools or buy them or what is supposed to happen?
That’s what we’re trying to do over here, so my goal through this conversation with Sushma Nallapeta who’s head of engineering and VP at Apartment List, a start-up here in the Bay Area, is to talk about what is her process in deciding when to build when to buy toolsets that are required.
I’m just going to check quickly on Sushma to see where she is. And if she’s having trouble. In the meanwhile, Anand, Uday if you have any thoughts on what PLG is, what are some of the challenges?
Anand: Rohit, we will wait to hear from you and then as we go along, if we have questions will raise them.
Rohit: Ok, I will invite Sushma here with me. I think there is a little bit of a challenge that she’s facing. Sushma is also, for the sake of disclaimer is also my better half. She’s also head at engineering. I will let her introduce herself, Sushma is with me here guys, she’s unable to join through Mentza.
Uday: That’s an awesome thing. Hi Sushma.
Sushma: Excited to be here. Thanks for having me.
Rohit: True, what I was hoping to do. If you can, do a quick introduction of yourself. That would be great.
Shushma: Sure, I’m Sushma, I currently lead head of engineering at a company called Apartment List. We are a two-sided rental marketplace. I’ve been here for three and a half years and I’ve been tech now for almost 15 plus years and I’ve seen and been part of companies and seeing high growth being part of startups, being part of bigger companies and excited to talk about some of those experiences today.
Rohit: Thanks, Sushma, welcome. The question Sushma, you know how product-led growth is such a big deal for everybody now that the new set of tools addressing the new momentum right?
So, the question is, again, every company wants to build its own set of tools because the set of skills that is needed in product-led growth is not available in a single tool set so people think of building everything, because that’s what engineers are. What is your view about the whole of this build versus buy?
Sushma: Yep, in my experience being an engineering leader, I’ve seen companies fully building everything you could possibly think of and also the other side where companies have built a huge business of the software, they just bought on a credit card, and I’ve often wondered what’s the right approach. So, I kind of take my own framework. That has helped me kind of internalize what we should be doing in those situations. So, I call these the seven C’s and these are capability, cost, complexity, competence, cohesion, competitive advantage and continuity.
I can go into each of these in a moment, but basically these form a framework for us to really understand whether we should be building or buying software. Now, if you think about capability, for example, this is what do we want to solve for what is core to our business? What capabilities do we want to achieve, do those exist in the market? And this itself like becomes a huge input to our decision making.
As an example, one of the companies that I was part of, we built our own AB testing framework and the reason for that.
Rohit: You really built AB testing framework.
Sushma: Yes, we pretty much built AB testing framework. It was almost 10 years ago, the ‘optimize lease’ of the world did not exist back then, and there were also any out of the box solutions that were available at that time were insufficient for the business need that we had.
We had to create a ton of experiments; we needed that speed. We needed to get features in front of the customers. Heavy customization and personalization and we needed to win against our competition, so we needed to build these capabilities ourselves.
Rohit: what kind of industry was this that you needed such speed?
Sushma: This was a dating industry. I’ve done this again at a previous e-commerce industry as well, so it is a very common thing that we see. So, capability is such a huge factor in really determining whether we should be building or buying. The second C that I would like to talk about is cost.
And cost is such a big factor in the decision-making process. People often think about cost in two dimensions. One is resources, the other one is time and then they compare it against buying and go with the cheaper option. But usually, cheap does not mean it’s the right solution for us. And there are so many other factors.
Rohit: Got you, so Sushma just to talk about right, you talked about capabilities so right now in the PLG world there is not a lack of capabilities. But what about the fact that you did mention one of the C’s as competitive advantage so people may push for PLG as a competitive advantage because they build their own growth tools, right? What do you think about that?
Sushma: Yeah, it is, as long as it gives you an edge, it’s good for your business. I think in PLG often the focus is on driving growth, maintaining retention, conversion, those are the factors, but as a company we are also trying to build for the core competencies of the product itself. Growth is one part of it, but once the user actually ends up signing up for your product, you need to kind of provide the actual solution. Actual core, a feature that you’re responsible for.
So, with that, my focus kind of shifts to what is core to my business and I need to focus on competitive advantage by building the core key features.
Rohit: Got you, so we talked about capabilities. I think we talked about cost and I think you also talked about competitive advantage and you’re saying that look unless your business is super unique, growing is just a part of who you are as a startup and that is not a competitive advantage unless you have a unique insight into it otherwise, don’t build it, is what you’re saying?
Sushma: That’s correct.
Rohit: What is the thought on the business complexity? Because businesses are all complex, right? So, where and what have you done when complexities have come your way?
Sushma: Yeah, when I think about complexity, basically it is what is the opportunity cost of not doing something else and doing this. Finding the easiest way to perform a task and have a great impact on productivity and business results. If it is simple, go ahead and build it.
Like you know if it is complex, you have to figure out if this is really the right thing or should I just buy something off the shelf.
Rohit: Got you. Thank you Sushma, on the complexity of it. Right. Now what about the other aspects like Oh well, we want to build things because we want to build it. So, in your experience what are some good examples of complexities when somebody should decide to build it?
Sushma: Yeah, I’ll give you an example. Again, one of the companies that I worked at, we built our own billing system in the world of having Stripe and other systems that is readily available we actually built this.
Rohit: Why would you do that?
Sushma: I think basically two facets to this problem. One we were in an industry. There are some industries like dating or you know furniture or e-commerce that has a very high charge back rate. So, because of that it’s impossible to actually maintain your business margin.
Rohit: When you say charge back, you mean the credit card charge back, right?
Sushma: That’s correct. So, we were in a subscription business. And then if we ask customers to pay for our service. We ended up getting very high charge back rates on these credit cards so because of that we needed to build our own billing system to handle this and connect to the processors.
We also maintained, you know, for fail overs like meaning if one processor wasn’t available, we had to fail over to another one, also to mitigate fraud. To maintain multi country presents. To continue to evolve and grow our user base. This was a very complex problem to solve and we needed to build it on our own.
Rohit: So, you’re saying we need to build when there’s a real problem of the business complexity that the current tools don’t actually address? But you need to kind of, you know, wait, what’s right or what’s wrong, right? That’s what you’re saying.
Sushma: That’s correct.
Rohit: Ok, what about the other C’s that you have, you didn’t talk about competence and cohesion? Why don’t we talk about that?
Sushma: Yeah, I think when it comes to competency, the biggest question that I always ask myself is let’s not chase shiny objects, especially it’s very common for folks to say I want to build my own. You know, data science team and ML and some of these core capabilities, but essentially what I want to be asking for is how good is my team? How long have they worked together? How much do they really understand what we’re building? Are there new technologies involved and how long does it take to actually ramp up and then also, do we hire specialists to do this and hiring these days takes anytime, anywhere from three months to six months, and that is a huge time. The time that is essential for me to be launching something.
Rohit: So, what you’re saying is If it’s not a core competence of your business and growth is something you need, don’t be building it unless its core competency for you because you’re like if your product, if you’re selling accounting, sell accounting, don’t build growth tools because that’s just a part of one of the stories, unless there’s a hidden competitive advantage, is what you’re saying.
Sushma: Correct, right now I work at Apartment List. We are a marketplace. What is core to our business is how do we actually match the right renters with the right property? So, I want to actually focus on building that. And today we actually invest in a lot of growth like we invest in product-led growth.
We invest in getting a return user back some of the users who came to our site about a year ago. Three months ago. So, we actually initially had built our own email software and we moved away from it. We decided to go the buyer out. Because we could not maintain it, we did not have it. It was not our core competency. We actually needed to build more product features. We needed to go to the market quickly, so we decided to actually change our decision.
Rohit: Ok, this sounds contrary, so Shaveta has a question over here. What about buy in the short term for cost and speed and move over to build in the long term? I think you have an answer for that, but you just told the rivers of it right now because you actually build something now the cost of maintaining it is hard and you can continue to deliver with the same speed, because that’s not your core product. In this case you went and bought software, right?
Sushma: That’s correct, and I think to answer Shaveta’s question, a lot of times, especially in startups, all you’re trying to do is to get to the next level of funding the next round of funding. So, if that’s what gets you there, just build it. Whether it’s simple, whether it’s complex, if that’s what your goal is, I would totally recommend building it, but as the company continues to scale, you need to be thinking about how do I allocate resources to get the best return on investment and then it’s not a competency you should not be building it.
Anand: Can I ask you a question here Sushma.
Rohit: Yeah, go ahead Anand.
Anand: So, you know. Especially for companies in this age, data is almost considered core to your business right and so for example, analytics tools, visualization tools. There are really good ones there. At the same time, you will need to have a set of data that you are tracking yourself. Now again things that are so core to you, like data, does it make sense to look at building it yourself in the long term rather than relying on an external, you know suppliers the good thing with external supplies obviously they’re on top of the job all the time, so you know they would be making improvements all the time, which you might not be able to do that.
What’s your sense? Are these things that you would look at? You know outsourcing or building it.
Sushma: Yeah, actually I think I completely relate to that. That’s a great question Anand. so, we actually bought and built our analytics tool, so let me explain. So, we actually have a ton of data. We use all of that data to continue to build features to continue to grow more users and bring more users to the side.
So, what we realized is. Yes, we have all that data. There are certain parts that are extremely core to our business and we needed to build and there are other things that we can augment and buy. So, we actually built our own eventing framework because we needed to exactly understand all the user behaviors. That is extremely customed to our side.
So, we actually figured out that we needed to build the actual capturing of that type of events that we needed and then we, you know bought like Google Analytics 360 or like you know we bought, We signed up for Looker and Big query. To augment some of that capability and this way, like my, you know task became easier because on one side I’m focusing on building the right thing that for my business and then on the other side I’m actually like, you know, speeding up the process to get all of this in front of our employees, in front of our customers so we can actually capitalize on what is very meaningful for our business. So, in this situation I would recommend building, buying and then building on what you buy.
Anand: That explains.
Rohit: I mean did that answer. What I am starting to get a sense of right, in all of this thing is there is no one golden answer at the end of the day. I mean, you would say that is the golden answer, but that’s not the right thing. Like Anand said there are situations where the core of the business has to be evaluated, and again, I’m going to Anand’s discussion point right. Not all data is equal. Your customer records are customer records.
There is a dimension to it, but there’s nothing special about it unless the uniqueness is in your business, right? So then use readymade software but in PLG what happens Sushma is, what we see is the marrying of the data silos. In that case, the tools are not there yet, but there are new tools that are coming up. So, what do you think about that aspect of PLG?
Sushma: Right, I think one of the C’s I mentioned was cohesion. How connected are my systems and how well they work together with everything that I have like I might have Salesforce. I might have other software that I’m actually, you know, working with, and then I might have my own internal systems that I’ve built out.
How do they all work well together, and what is the continuity like? What is maintenance, support, upgrades like that, so I think those two factors play a huge role in what you said, especially in the PLG space.
Rohit: So, what you’re saying is if they don’t work well already today, building on it and maintaining it is still one thing I contend with as myself, a product-led growth tool supplier, right? Because Glance does the same part of product-led growth through email marketing. What we think is you don’t necessarily have to build those. You can build things above it, but essentially marrying the data silos and figuring out what the insights is. We have to figure out that integration itself becomes a long system because there are about 20 systems anybody uses today like they use segment like you said, Google Analytics, Looker Salesforce, email software, marketing automation, sales engagement.
So how many such integrations and every company has to go and build the same set of integrations. Because systems of records are spread out, including the billing system for that. So that is something that I think is a thing. What is the aspect of continuity that you were going to talk about in your 7th C?
Sushma: I think the most important thing about continuity is that there is always you have to assume there is always going to be evolving business needs. That is just the nature of the business. What you build today might not be what is needed like as the company goes through the next stage of growth. So how do you build continuity on your previous investments and what does that look like? There is no one solution to rule them all, so how can you actually make those decisions and figure out in what areas you can build it in a way we can easily replace solutions?
Rohit: Correct, correct. And I Also think the trouble of data continues to be there. That there’s a lot of data continuity challenges that come in, right? What I mean by that is, there’s a lot of data silos you need to build, and there are some existing investments in systems that you are doing.
What are some challenges as an engineering leader when you have to make decisions which are exactly the opposite of what you advised us right now?
Sushma: yeah, I think there have definitely been some contradictions. I think we have built software as a necessity just purely to attract talent or foster innovation. We have built only like you know as part of just trying to get the next level of funding. We’ve also built as a necessity to creating new lines of businesses, especially when you have a core business and we’re trying to really figure out the product-market fit for maybe a new revenue stream or a new idea we’ve also built different proof of concepts to get to market quickly, as Shaveta said and then try out, you know various features pretty quickly.
And then the other part of the contradiction on the other side. We’ve also bought as a way to get to the market faster. There have been companies that have actually launched completely off of using off-the-shelf software in one of the companies I was in, we used, we launched our entire e-commerce site based off software we bought on somebody credit card for $1000.
Rohit: Oh wow, so you’re telling me? you need to buy software when you think it gets to mass market, well it makes sense, right? You want to buy software. In the long term when you look at this overall thing, I do understand the contradictions like you want to build because it is what you want to build. But there are times then when you have to make these things right, but Anand has asked one question, so I’ll ask my question after I read yours ok.
How does one leverage these capabilities that one builds when pitching to investors? Well look, I let Sushma answer it also once and then I’ll tell you as an entrepreneur what I am doing.
Anand: Go ahead.
Rohit: See as an entrepreneur what I think is. There is more to the business that you build. There is an insight into business or technology which is mostly a competitive advantage. Or ease of use or anyone such offering that you are actually focusing on when you build these capabilities.
Sometimes it is redundant. Second thing cost if you cannot afford to use something you build it. But again, you have to always figure out whether your engineering dollars are spent better on just something that you can afford versus something that you have to build.
It’s all about time in mine line, so the capabilities when you build for an investor, I will only build that is super important and seems like a competitive advantage.
Sushma: Yeah, and I think especially when you are pitching to the investors, the most important thing is what is your idea? How will you grow your users, and how will you actually like, how will you find the product market fit? And I think when that’s the criteria, it’s really not about the technology underneath it’s about how quickly can you grow your user base and how quickly can you get to a monetization strategy? So, technology just enables it, but it is not the only defining factor for an investment.
Anand: Thanks so much. That helps a lot.
Rohit: Uday did you have a question, oh, how do you keep accessing capabilities, especially after. Uday please unmute and ask the question rather than me reading it because I can’t see the full question in the comment section.
Uday: Yeah, I was just trying to say that sometimes you could just be redundant to believe that you have the capability. Because you would have had a series of successes in the past. You know building something instead of buying something. So how do you keep developing that muscle to assess it rightly that this is the time we need to buy and this is the time you need to build and not fall for the temptation of just depending on your own competencies and capability.
Sushma: Right, I think there is no one good answer. Sometimes it’s making some mistakes and learning through it, but I think over a period of time I guess the most important factor when you think about capability is do you want better control and also what does the ease of use look like? Is it easy to use if you build it? Are you going to be able to have the time to maintain it? What does that investment look like? And also, is it portable?
Like you know when you build something and if you want to actually migrate to something else or like you know, when you get to the next stage of your growth, can you actually port that solution? Is that scalable?
And then also ultimately what it boils down to is shortest time to value, because I want to get things in front of my users quickly, so if that means I have to build something then I’ll build it. If I have again, going back to, do I have the right team? Do they have the right talent? Do they have the right skill set? If that is, if my answer is yes, let’s just build it. But if my answer is no, then I actually want, my most important value is to get it in front of the users. so, I will go there and buy. Right?
Uday: That’s super. Which actually puts a lot of focus on leadership and wisdom there, right? And having your ears and eyes to the ground completely.
Rohit: So that is where I think you also did the email system right? You used to build your own email system and you moved away to and actually bought software because you can continue to deliver the value and flexibility because your business is going and dependent on email right.
Sushma: Yeah, we built our own email system and it was like, it actually served us through several years and almost three to four rounds of funding in the life cycle, but after a point you know it was not enough for us to scale, so we decided to go use it. In fact, we decided to use two different email systems. We used Iterable for our marketing emails and then we used another software called Courier for our transactional email, so we decided to buy two and build some services to interconnect the two.
Rohit: Interesting! Shaveta has a question about large banks are built on tech that is bought after then rather than build. That’s absolutely true, because the banks need business continuity. If you look at Sushma’s point about continuity of evolving business needs, build on previous investments and continue to keep on pushing updates and time to value. I think that’s where it all boils down to Shaveta and the guarantees of software services and SLAs. So, Sushma? What would be your final conclusion?
Sushma: Build, buy and build on what you buy.
Rohit: There you go, folks.