In San Francisco, we meet co-founder and CEO of Gild, Sheeroy Desai. He shares his story how Gild was founded and how the current business model works, as well as what the current plans are for the near future, and some advice for young entrepreneurs.

The transcript of the interview is provided below.


Martin: Hi, today we are in San Francisco, in the Gild office. Sheeroy, who are you and what do you do?

Sheeroy: So, I’m Sheeroy Desai. I’m the co-founder and CEO of Gild. At Gild what we do is we help our customers, which are companies from startups all the way to very large enterprises. We have a software solution that helps them find, engage and end up hiring software engineers.

Martin: Great!

Martin: What is your background?

Sheeroy: My background…. I actually grew up in Pakistan and I came to the US to go to college. I went to MIT and got electrical engineering which I never used. I was fortunate enough that I graduated from the university right at the time when IT/software boom was just starting, which was the late 1980’s. My first job was actually  at a startup, even though back then I didn’t know what a startup was.

Martin: Was it called a startup?

Sheeroy: No, it wasn’t. That wasn’t even called a startup. It was just… I thought it was just a small company. It was like 40 person company and I worked for it and I wasn’t even quite sure what they did. But it turned out to be one of the very early companies that was pioneering in IT services. And it eventually became a company called Cambridge Technology Partners, which went public and eventually got bought up by Novell.

But anyway, I was there for about 3 years when we were very small and it was great training ground to just learn everything about, you know, I did some amount of engineering but really rapidly moved on to product development/product management. And then actually even I got a chance to engage a little bit of marketing and sales.

So from there, a number of us who are actually mapped up at Cambridge Technology in the early days and worked together, about 5 of us, ended up leaving and regrouped and started a company called Sapient. That was in, like 91. And then I spent the next 17 years of my life at Sapient in various capacities, from again, doing product, management, client management, to opening new offices, to eventually doing international expansion, and then I spent the last 7 years as a Chief Operating Officer there.

We took the company public, so by the time I left, it was over a 5,000 person company. It’s still listed on the NASDAQ today, I think it’s about 13,000 people. It’s enormous.

So that was, you know, a really great experience in many, many different ways. I mean, taking a company from startup all the way to IPO and beyond. You know, obviously you’ll be able to side our ups and our downs. Because we also lived through the dot com bust and then survived that, so it was a lot of learning in those 17 years.

Then, ultimately I left. I think I was really burnt out, quite frankly, after all that time in the company. And especially at least in the last at least 11 years, I was really being involved in running a public company that takes a toll after a while. And I was really missing being an entrepreneur again, doing something small again.

My problem was that I didn’t know what I wanted to do and I looked at a lot of smaller companies, or startups, and I just couldn’t get passionate, excited about anything.

Ultimately, I really came to comfort of doing a lot of soul searching. I came to the conclusion of what I was truly excited and passionate about is what we do here at Gild, which is really helping companies in…. Let me back up and say it this way, which is the way, companies find talent and the way professionals find opportunities is one area that actually still the underlined way it works, the underlined processes, the structure, hasn’t really changed in last 100 years. We rely in the same artifacts, whether that’s a resume or LinkedIn profile now, whether that’s an interview, we still do those things. There is no better way to get insights into people and to make better decisions. We use machine learning and analytics help our shopping experience for everything else today. But when it comes to shopping, let’s call it for technology professionals or profession in general, we are still in the dark ages. So it was really that insight that prompted me to start Gild, and this was about a little over 3 years ago, at this point. Well, Luca and I got together and we kind of, he had his own startup and I was kind of thinking about doing this. And we got together and decide to initially focus on software engineers and here we are today.

Martin: Great! How did you come up with this name and what are other alternative names did you think about?

Sheeroy: We thought about like a million of them. We probably made a list of about 100 different names. You know the challenge today in coming up with the name is probably a couple of full. Number one, it’s very hard to find a real word that can be used as a name because they’re all taken. So a lot of companies make up names, right. The second challenge, even if you make up a name, is getting the URL. So quite frankly, these challenge for us wasn’t being able to come up with names that we liked, that were creative, but actually being able to secure a URL and we wanted a dot-com URL. So ultimately, things converged towards Gild. The reason we like the name Gild because it’s a play on the word Guild, with a U, GUILD. The whole idea that there could be guild of technologist or the guild of other types of professions that we are trying to help. So that was the idea behind, but was taken and we couldn’t get it but, was taken but we were actually able to secure at a very reasonable price. So, that’s how we ended up with the name.

Martin: Great!


Martin: Can you briefly describe how the business model is working?

Sheeroy: The business model is simple. The business model is really a SaaS, enterprise SaaS model. Our customers pay us for basically getting access to our database. So what we have, at the end of the day, is we have a database of over 11 million software engineers from around the world, for whom we have generated all kinds of public information, ranging from just a profile, a biographical information, all the way to downloading and analyzing a new open source, code contribution that they may have. To analyzing their activity on Q&A forums, to analyzing the twitter feed. So we analyze anything possible about a software engineer and then trying to arrive to insights from that.

Over time obviously, insights are getting better because the more data you have, the more correlations you have, the better are your predictions. Since ultimately we are really trying to make predictions around how good is a software engineer, how in demand might they be, and how likely is it that they move to a different job or opportunity.

So those are the analytics that we provide to our customers, but then we also give them access to all the detailed data that we’ve collected on someone. This is a highly rich profile. So with our software called Gild Source, you get access to all of these. Then there are some work flow built in there as well, so you can do a lot of relationship building with the candidates, you can do all kinds, you know… think of it as almost like a CRM system for hiring. So you have a database and a CRM system, all combined together, and our customers get access to that and they just pay us on subscription base.

Martin: Is it only for full time employees that you are trying to match with the employers, or is it also for project related kind of matching?

Sheeroy: So, we don’t distinguish. We are simply out there, looking for information on software engineers and trying to tell a customer how good they might be. I’m sure many of them are looking for freelance opportunities and probably majority of them are looking for full time opportunities. But we don’t differentiate between the two.

Martin: Because I would imagine, if you distinguish between the two, then you can say that there’s some kind of trigger or analytics behind that. Okay, that there’s an 80 % chance that person would be willing to switch his job.

Sheeroy: Right. I think we probably can if we have enough data that we probably could analyze it and be able to say, we think this person is more interested in a full time opportunity or this person is more interested in, you know some kind of part time or consulting. We don’t do that. That’s an interesting idea. Something that our customers really haven’t brought up to us. And that’s one of the things we really rely on quite a bit is, our customers.

Very early on in the company, we made a decision to invest very heavily on what is today called Customer Success. Where 4 or 3 years ago, people refer it as customer service, we call it customer success from the start. The whole idea was not just so that we are making sure that our customers are successful in using our product but also, it builds a deep relationship so we truly understand, get feedback on the product. And that’s really important to us.

Martin: There’s one very interesting algorithm that you are using. Can you tell us a little bit more how this matching into supply and demand works.

Sheeroy: I mean, it’s number one out of many algorithms over here. We started off very simple. In the first iteration of the algorithm really was all about just trying to tell our customers of how good some of them might be.

The first iteration, that Luca built, was purely based on analyzing a software engineer’s open source code. So we would actually download the entire code repository that we would find, we parse it so we knew what programming languages they were using, how many lines of codes that were written, and then we would run analytics, basic analytics from the quality of the code, how extensive was the code, how well documented it is, etc. And based on those analytics, build issues of algorithms which would say, here’s how good we think this software engineer is. It is heck of a lot more sophisticated today, 3 years later.

So the software contribution is on the component of what we do, as I said, we also analyze activity and the level of activity on things like Stack Exchange, which is Q&A site. Then we start looking also at things, where the other developers that this individual is working with. So sometimes they collaborate on open source projects. Or maybe they are answering questions, or going back and forth with another developer, who is also very highly rated. Now, it’s like birds of feather flock together, so people of similar types tend to work together. So, if you are collaborating or working with really high engineers, chances are you’re pretty too. That’s an additional insight we can grasp.

And then it has gotten even more sophisticated because as we develop a very large database of what I called fact based software engineers. So now we can based on looking at quality of code or based on looking at their Q&A activity, etc., we can say that, in fact this software engineer is really very capable. But we can also start looking at matching people biographical backgrounds with the quality that we think they are. So over time, we’ve also been able to get insights into… Okay if we see a correlation on… I’m giving an example, I am not saying this is true, but an example maybe that anyone who worked at a certain company during the late 1990’s as a software engineer, the chances are… We see correlation that a lot of engineers there at that time were very high quality. So chances are, if we know nothing about you, all that we know is that you worked at that company in the late 1990’s, now we can start making prediction that likelihood is reasonable that you’re probably a pretty good software engineer. So analytics and as I said, there are many, many algorithms out here and the overall quality of the analytics and the complexity of the analytics, has gone up quite a bit over the course of the last 3 years.

Martin: When you were a very young startup understanding your customer needs etc, it’s very important and you also said that the customer success is very important for you. How did you try to understand what the customers really want in the early days?

Sheeroy: You know, it is still a challenge, not just early days, because, and I think this is the conundrum, because customers are great at telling you what is wrong with your product, what they don’t like. They are very great at telling you how to tweak a certain feature, about what they like to see. That is good.

Martin: Evolutionary step…

Sheeroy: Evolution stuff, they are very good at that. What customers are not very good at is revolutionary stuff. Because a lot of times, they don’t know what they don’t really need. So it’s really interesting because of what we focus on a lot. This is a trap, I think a lot of companies and entrepreneurs get into it, which is you start focusing on the recommendations that the customers are making.

We tend to focus more on the pain that the customer has: what is not working for them, what is the problem. And then, Luca, the product team is doing an awesome job, of analyzing all the things that are not going well for the customers to come up with a unique solution. We always ask ourselves: “Can we address this need in a way that no one else has. Can we address them in a unique way that is going to be significantly better than what anyone else is doing?” That’s the first place we go. If we can’t do that, then the question…  can we just get rid of the pain? But what we have found is, if you pay a lot of attention… So what really comes down to all of this is being close to your customer and enough of your customer so you can see the patterns, where is the most pain.  Where are they most unhappy? And then being able to really look at ourselves… Okay, is there a unique way in which we address that. So that’s how we come up with the revolutionary solutions, and that’s hard. That takes time.

Martin: Would you describe the Gild business model more of a local market place or would you say it’s a more global market place where I can find and hire developers from another country?

Sheeroy: It’s global. Our database is 100% global. So that’s number one, the database is totally global. You can search developers anywhere. In terms of how customers implement that obviously smaller companies are more regional. They are looking to hire potentially in 1 or 2 locations. And our larger customers are more global. They’re looking to hire people anywhere. Now having said all that, I think there are such a shortage of software engineers, they are such in high demand, we are even seeing startups now looking for software engineers anywhere in the world. You know, we are such a great example, actually more than half of our engineers team are actually in Europe, in Milan. So, we’re seeing even smaller startups, 50 people or so, having operations in multiple continents.

Martin: When you started, what were the minimum features that you that you tried to develop in the product and ship to the market in order to understand what the needs are, so you can decide for a revenue model etc.

Sheeroy: The minimum features when we first started was basically was search and results. That really was the minimal things we were trying to do, which is a reasonably easy interface by which you can say: “Okay I am looking for an iOS engineer in San Francisco or the Bay area”; and then being able to give them results that are relevant to that search and having analytics be good enough that they would make sense.

I remember reviewing the first version of the product, it was really rough. The UI (user interface) was extremely rough, we made a lot of assumptions. But I do think among things we did well was, we immediately start taking it to market. We just started talking to tons and tons of prospects, showing it to them, getting their feedback and then about 3 months later, we actually started charging customers. Not because honestly that I thought the product was so great that we should be charging. But again, a startup is all about testing a hypothesis, right. First half of the hypothesis was, if we give this level of intelligence is anybody going to care? That was getting anybody to say, this is great, I really like it. So second half of the hypothesis was, has anyone willing to pay for it. That’s why we start charging for it. And within a couple of months we had actually…. I think within 2 months we’re like 20+ paying customers. So like… Okay, this is working, people are paying for it.

Martin: How did you solve this hen egg problem? Because when you started, this is basically a type of market place which is connected by some kind of scoring algorithm. How did you create the first 1,000 or 2,000 of developers and their first 50 – 100 companies.

Sheeroy: It’s a great question, which comes first, the hen or the egg. Actually, right before we started doing what we’re doing, the Gild Source business, the prior incarnation of our business actually was, we were building a website where software engineers could come and take coding challenges and puzzles and write code and we would analyze that code. So that’s truly trying to build a market place. On the one hand, trying to get the software engineers, on the other hand trying to get customers and kind of have supply and demand. That is very challenging. It was actually an outflow from them, which is as we kind of work on there for months and months and we finally realized… Gosh, we were trying to build a two sided market is very hard because just the value proposition for the software engineers wasn’t compelling if there weren’t any employers. And for employers it’s not very compelling if you don’t have at least tens of thousands of software engineers that they can search and choose from. That was a big problem.

That’s how initially, Luca came up with the idea. He said…. Well, let’s address the supplier issue just by creating profiles for this software engineers by looking at their public activity. So in a way, we started creating an inventory of software engineers, so then we can then just focused on building out the customer’s side of it. The initial thought was we would go back and build a community of software engineers but as we continue to move down, we realized, we don’t really need to do that. So that’s how we got going.


Martin: Sheeroy, let’s talk about the corporate strategy. You’re currently have this kind of market for software developers. What are your thought on adding other verticals and whether you can also apply this kind of data driven decision process.

Sheeroy: We started a company so that eventually we solve the problem of hiring any professions, not just software engineers. So that remains our mission over time. We want to provide this level of analytics and insight into any professional. Lots of challenges to it.

  • We’ve been thinking about this for a while, and we have a viewpoint on how we can eventually get there. Which is a secret, I’m not going to share that. What I’m thinking is on that. But no, we obviously have been thinking about that and we’ll continue in the long term to execute towards that. So that’s one answer.
  • Second answer is, we’re building a company here for the long term. In the way we think about ourselves and talk about ourselves, we want to build a company, this type of problem we’re trying to solve it’s not going to get solved overnight. It’s going to take some time. So we’re Okay with that. We are in no rush to just get there tomorrow. With that realization, we also have realized that, if you want to build a very large company, there’s a saying if you want to build a monopoly, a broad monopoly, first you got to build a monopoly in a small market. Get that going, get that right, and then you have the opportunity to build a monopoly in the greater market.

So that’s what we’re trying to do, we’re really trying to get it right in software engineering. And really understand what it takes to win and be successful there. Because once we have all the answers there then duplicating this to other professions would be a lot easier for us.

Martin: In terms of competitive advantage, would you rather say that Gild has the advantage of having a large network effects, or is it more that Gild scoring system is the key source?

Sheeroy: I think our competitive advantage is our analytics. I mean no one else that I’m aware of, no other competitor… I mean competitors are doing very resourceful things, they are competitors who are aggregating profiles. So to your point earlier, I’ll go on LinkedIn and scraping it, I’ll go scrape Twitter, and I’ll go scrape Facebook, and I’ll try to give you an aggregated profile, and we do all that. But no one is doing is really the additional insight of really telling you how good some of us and why we really think this person is good and what they are good at. Why they are in high demand or not, or when they might be looking for a job. Those are analytics that no other competitors are providing. That remains our competitive advantage.

Martin: Great!


Martin: Tell me about market development. Can you give us a brief overview of how you perceived the kind of market and the players in the market and what is currently happening over there?

Sheeroy: So we play in kind of Recruiting HR market. Like most markets, it’s  fairly saturated, there’s a lot of noise. It’s a noise in market, which is a few are the head of HR, a few head of recruiting acquisition in a company. Chances are high, you probably get a 5-10 emails, voicemails a day telling you, “Oh, I have this new amazing thing.” A challenge for them, the challenge for us how do you actually get yourself above that noise level. There’s no easy answer to this. I mean I wish I could say, “Here’s the magic that we came up.” But there’s no easy magical answer to this. It is a number of things.

For us, what we did is number one, that I think is a little bit unique and different than I see some other startups doing. We do all the normal stuff you would expect us to do: we do contact marketing, we do lead generation, and we do lead nurturing and all that stuff.

But I think some of the things that we did differently is, we focus on brand building very early on in the life cycle of the company. So, we were only in the market less than a year, when we were able to get a business cover page story on the New York Times on Gild. That was huge. Probably when I think back in my career like years and years and years from now, I’m most likely will remember that Sunday morning, getting the New York Times and right there on the cover. I mean, that doesn’t happen very often. That was a good nine months of effort on our part. We had to cultivate a relationship with the New York Times, we had to stay on them, we had to give them really meaningful story that they could write about. It took a lot of effort, a lot of time, and a lot of money, I guess in some ways. I got to get to invest in PR, as an example. A lot of startups don’t believe in that stuff and I understand why they say that. But I think in our case, that’s one example of many different things that we’ve done that have allowed us to stand up above everyone else. And as you know, I’d say, I think, one of the things that has distinguish us and hopefully will continue to distinguish us, is our efforts and our emphasis on brand building, which is a long term payoffs. It doesn’t payoff short term.

Martin: Okay.


Martin: For our readers, we always try to share some knowledge, so they make less errors when they start their companies. What would be your advice when starting a company? Maybe you can also talk about when starting a market place.

Sheeroy: Starting a company, a couple of things… I could probably go on for a while. Just imagine a couple of things.

  • One of the things that I said earlier, which is don’t be afraid to put your product into market early. I think one of the mistakes, entrepreneurs make is that they want to put out a highly polished product and get it right. The fact that matters, you won’t get it right, you just won’t. Because you don’t know, you don’t know what the market really wants. So don’t worry too much about that, get a product on the market and then get real feedback. So that’s the first thing I would say. That’s a process of constantly iterating until you achieved product market fit. Stay really focused on that. So, that’s not a unique advice that I’m giving, lots of people do that. But it’s harder than what people think it is.
  • The other thing is, and maybe this is a little more what I see here in the Valley. So I’m trying to give some slightly different advice. There’s a lot of  focus on fundraising. Of course, especially if you are going to be a SaaS startup or something like that, there’s a lot of investment you need to make, and the revenue is kind of trickle in slowly. So, believe me from first hand, I totally understand that financing has a key role. But I do think there’s a tendency to get a little bit consumed by financing and worry a lot about evaluation and this and that. From all my experience, I can tell you, worry about the quality of the investor, someone who’s going to be a long term partner. Someone you trust and you’re going enjoy working with for 10 years because that’s what you’re signing up for. I think a lot of people get focused on evaluation and the amount they raise, and things like that. I kind of emphasized enough spend really time vetting the investor. If it is a VC firm, vet the partner. This is someone you really want to be talking to on a Sunday evening because that’s going to happen. This is the person you want to be seeing at least once a month and this is going to happen because you got to spend a lot of time with this person. So that’s the other piece of advice I’d give.

Market place is tough. I mean, what I’d say is learn from us and learn from others who’ve done the same thing, which is you’ve got to create a supply somewhere. I think the best way to get a market place going is to create that supply artificially.

In our case, it was really aggregating and creating these profiles, which end up really being our business. That’s the funny thing, because we haven’t gone back to building a market place. But, I would say, try to figure out how to artificially create one side of the market and then go aggressively create the other side. The other side hopefully should be the side that pays, so you can start monetizing.

Martin: From my understanding, a lot of people when trying to start a market place, they can start by artificially creating some kind of supply. Like some people, what I would call a good artificial supplier, which is real. And some others are trying putting fake profiles, which from my understanding, the majority of people are doing that. What would you recommend, I mean obviously, the artificial supplier would be supported but in most cases it won’t work. Would it?

Sheeroy: I think it depends on the market place. This is a question of what happens when the rubber actually hits the road, right? What do you actually do? There is no easy answer. I think each market place is a little bit different, but what I would say is, number one is be clear where you are going to generate your revenues from. That part you can’t fake. What you got to make sure number one is what is going to be valuable value proposition. Get that value proposition right, then on the flip side, can you generate that artificially. There are many different ways to do it artificially, you don’t have to have  a huge market. Say, I wanted to build a food delivery startup in San Francisco. Think what I would worry about most is who is going to want the service and are they willing to pay for it. I could artificially put up hundred restaurants, you can order from any of these restaurants, and the people start ordering, then I got to make sure maybe it’s me once the order comes in, is running to the restaurant and getting the food or I have some freelancers that’s doing that for me. The other side doesn’t need to know how you are satisfying the demand. So those are many ways in which you can create artificial supply without really, you don’t have 1,000 restaurants. You don’t have any agreement with 1,000 restaurants but the buying side doesn’t need to know that.

So those are the creative things about…. So that’s why I think market place is tough and I have a lot of admiration for people who pull off market places. Uber, for example has done great. Until you have a startup where you dealt with a market place problem, I don’t think you understand, how logistically hard it is to pull something like that off.

Martin: Sheeroy, thank you very much for your time. When you hire the next time one of your great employees, whether it’s a techie or a finance guy, maybe you should be thinking more about data driven recruiting. Thank you very much!

Comments are closed.