Ashutosh Garg, Co-Founder & CEO Eightfold.AI chats with Amit Somani, Managing Partner Prime Venture Partners.

Listen to the podcast to learn about

04:30 - How to Find if there is Enterprise Demand

08:30 - Find Pitch Market Fit before Product Market Fit

14:30 - “I didn’t build anything before I raised the first dollar. “

16:00 - Why Job Descriptions are Useless

23:30 - Moneyball for Enterprises

30:00 - From an Engineer to a Business Person

Read the complete transcript below

Amit Somani 00:40

Welcome to the Prime Venture Partners podcast. I’m your host, Amit Somani, and I’m delighted to have with us Ashutosh Garg, a former colleague, friend, and now serial and twice unicorn founder. So welcome to the show, Ashu.

Ashutosh Garg 00:55

My pleasure. Really excited to have this conversation with you, Amit.

Amit Somani 01:00

Thanks, Ashutosh. So love to hear a little bit about your journey since you grew up in India, went to IIT and so forth. So maybe just tell us a little bit about your background before you got into entrepreneurship.

Ashutosh Garg 01:15

So grew up in Himalayas in a small town near Delhi. After finishing my undergrad from Delhi and working in Bangalore for one year. Went to UIC whereby chance having to dabble into machine learning here in the early days. Got my PhD from there and after spending a year at IBM research and Google research where I got all the personalization efforts and there learned things like search, ad targeting and so on, and left Google in 2008 and that was the beginning of my entrepreneurial journey.

Amit Somani 01:45

Wonderful. So when you were back in grad school as well as some of the time you spent at IBM research and Google, was entrepreneurship ever on your mind? Or was it mostly around inventing new things around research, around patterns, around papers? Just how does that happen as a researcher to jump into entrepreneurship?

Ashutosh Garg 02:10

I think as a researcher we’re always solving new, interesting, hard problems. We are always trying to see where is the gap in the current research work is and trying to fill that. And at some level entrepreneurship is very, very similar to that. So I would say I was always excited to do something of my own but didn’t know what to do and how to go about doing it. So after spending a few years at Google, it became obvious that this is now what I can do.

Amit Somani 02:40

Got it. Got it. So maybe if you want to just share a little bit about the founding of your first company, BloomReach as to how that came about, what was the inspiration and then of course, we’ll move to Eightfold shortly after that.

Ashutosh Garg 02:50

I think each of these companies that I founded, people talk a lot about what we have done well. They don’t talk about where we have failed or didn’t go anywhere. So I left Google in May of 2008. And for the audience who know, 2008 was a really bad time with respect to markets. The world was falling apart. So I tried to do a few things which didn’t go anywhere. The key thesis in each of those things was I knew search, recommendations, personalization, ranking world. So what can I do in that space? It is started by building out a better search engine, but being a very capital intensive thing. In the year 2008, it didn’t make sense to pursue that. And later that year I got introduced to my cofounder of Bloombridge and together we decided to take that search technology and help enterprises succeed. So, that is what led us to start BloomReach. Initially, after some dabbling, we became an e-commerce personalization platform.

Amit Somani 04:00

And Ashu, maybe we should talk a little bit about Eightfold and come back to both Eightfold and BloomReach because one thing that I find extremely impressive is that from a deep tech and research background and identifying this need around search personalization and now AI, machine learning and hiring and recruiting, how do you find that there is enterprise demand for this? Because it’s quite audacious. A lot of people will start on the SMB end of the spectrum or the typical SaaS and so forth. But in both cases you started with the enterprise side, right? So how did you even identify that this was a big enough demand pool there? On both this and Eightfold. And maybe you can describe Eightfold for listeners before you answer that question.

Ashutosh Garg 04:45

So I left BloomReach in the year 2016 and I started Eightfold later that year with the primary focus of helping enterprises better manage talent. Our mission is enable the right career for everyone in the world. So we grew up to solving for employment. And we started by working with some of the largest enterprises across the globe which are today using us to hire, manage, grow, retain a diverse workforce. And we are using AI to better understand people’s potential, capabilities, what they have done, what are the skills they can quickly acquire next and using that to better match them to the right opportunities.

I think depending on whether they are selling to large enterprises or small enterprises, what you have to do is, what is your core advantage as a company and where the pain point is more serious? What’s interesting is that with large enterprises, typically you have to cross a number of hurdles, whether it is the legal reviews of your documents, the security review, the contracting data, the other systems that you have to integrate with, all those hurdles are there. Now, quite a few times we get scared of those things and we say that let’s start with SMBs, and then slowly, slowly, over time we go up market. On the flip side, sometimes I say that if the problem is hard but not too hard, maybe go after that because that becomes a competitive moat as well. What everyone else is scared of, I think of that as a key advantage.

So when someone complains that okay, large enterprise integrations will be painful. I’m like, no. Actually, why can’t that be my competitive advantage? So I think that mindset, combined with where the pain point is deeper led us to selling to large enterprises. But happy to dive more in on this one.

Amit Somani 06:45

Yeah. So I’d love to know, let’s say I’m an entrepreneur listening to this podcast and I have some expertise whether it’s in an area of technology or business or something, how will you do one customer discovery to validate that this is a real problem? Just in terms of access to these enterprises and figuring it out, number one. Number two, for you to iterate your way into getting your early PMF, getting your first two, three, five customers, maybe broaden this a little bit. How do you do that? So I love the fact that you say, “Hey, look. If you’re solving the hard problem, that itself will be the moat because many other people won’t try it.” And it’ll take you a while, but when you get there you are somewhat irreplaceable.

Ashutosh Garg 07:30

I don’t think there’s any silver bullet or any cookie cutter answer over here. But a few things, right? One is as you’re building a company, you’re building it for the next 10, 20 years, not the next two or three years. So think about where the world is headed. That’s very important. Second is don’t fall in love with your product. Maybe fall in love with the market. So in our case, in the first one it was the e-commerce market. In the second case it is the HR market. Fall in love with that market. What specifically you are building will keep on evolving over time, versus I’m going to force fit my product, my idea, my innovation and technology into this market. And third is what I call in the early days, try to get the pitch market fit instead of product market fit.

And quite a few times people will be like, “Okay, let me pitch my idea. Oh you’re not getting it. I’ll show you my product. And you need to use my product before you understand.” No. Building a product takes time. So sometimes in fact, I go to my product managers and say, “Just listen when I’m pitching to customers. And then let’s go build what I’m talking about.” If customers are resonating with that pitch, you can build that product. If customers are not resonating with that pitch, there’s no point building that product. So take a very low cost approach, call fifty customers, see how they’re reacting to your pitch. And if you see that there’s a lot of reception, go deep over there.

Amit Somani 09:00

I love that. Pitch market fit. I don’t think I’ve ever heard that before. I’ll come to that. But I loved another thing you said earlier which is fall in love with the market. Can you double click on that a little bit and expand? What does it mean to fall in love with the market? Let’s say you said HR tech is a big market or recruiting is a big market. Now what does it mean to fall in love with it?

Ashutosh Garg 09:20

A very common thing is that people are like, “Okay, here’s my product. I really love it. Who’s going to use it?” That’s one way to approach the market. But the market may change, people may or may not like your product and it may not go anywhere. On the other hand, if you ask me what is Eightfold about, Eightfold is about solving for employment, really enabling hiring retention of a diverse workforce. That is a problem I’m solving. Is there a big market for that problem? The answer is absolutely yes. Almost a trillion dollars gets spent on hiring, retaining, and managing talent. The problem is, as a CEO, do I have the right team? Can I hire quickly? Can I hire fast enough? Am I hiring the right talent? Fall in love with solving that problem. Now, how do you solve that problem? It’s just a means to an end to that problem.

So customers are interested in buying your offering to solve that problem, not buying your offering for the sake of buying your offering. So stay focused on that market, that problem that you’re solving and be comfortable adopting, adjusting a product to the needs of the customer or the market.

Amit Somani 10:30

Got it. No, fair enough. I think you also said earlier: find a macro trend and make sure that you’re doing something that will matter 10, 20 years from now, not just for the next quarter or two. So, that also helps. Back to the pitch… Go ahead.

Ashutosh Garg 10:50

So just to add, sometimes you find a macro trend, sometimes you create the trend, sometimes you have to dig in. As an example, in 2016 HR was not a trend in the market. No one wanted to do anything in the HR space. The trends were around crypto, trends were around autonomous driving and those things. But on the other hand, the trend that was obvious to us was that talent is going to stay important, and in fact will become more critical over time. So can we go after that problem? So sometimes you have to find those strengths then and then harp on it for a while before it becomes mainstream.

Amit Somani 11:35

Right. I think the trick in that one is to stay alive and solvent till it becomes mainstream. If you are doing interactive TV, you might be doing it for 30 years. Even AI got a second and a third coming. But I agree with you that finding something and sticking with it will yield to something assuming you’ve picked something at a broad level that’s going to pan out.

Ashutosh Garg 12:00

Yes. And just to add, put more color over there, see, the first three, four years, let’s say if my revenue per year is a million, two million, three million dollars, four million dollars, which is fine. For that, I don’t need market to be a 10 billion market. Even a 50, 100 million market is good enough for that. But if there’s not even a 100 million dollar market today, don’t even bother with this thing. There has to be some market today and a trend has to be there that this market can grow from a hundred million to a 10 billion dollar market. But by the time you’re a 10 year old company, you are targeting a revenue of almost a billion dollars to be successful. And if a market is only a billion dollars on that day, you’re not going to go anywhere. So on day when you’re starting, you have to have some market, meaningful market that you can go after and you should have a gut feel for all the data that says that this market is going to go from 100, 200 million to tens of billions of dollars, if not more.

Amit Somani 13:00

Understood. I want to go back to the pitch market fit as opposed to product market fit. That was very intriguing. So give us a real example maybe from the Eightfold journey early on where obviously you would’ve had some imagination of what product you want to build. I know you’re falling in love with the market, HR market, recruiting market. Just give us a couple of examples from the early days. In what way, based on what you were pitching and what you were hearing from the customers, did you change what you were building?

Ashutosh Garg 13:30

Two-three examples. What we did in the early days, a number of people asked me, “What did you build before you raised the first dollar?” And the answer was I didn’t build anything before I raised the first dollar. It was on a PowerPoint. Now, the second thing is that, “Oh, you raised some money just from PowerPoint?” No, actually that PowerPoint did have at least 25 customer interviews with a very focused pitch. So we would call HR leaders of different companies and ask them, is talent a big problem for you? Are you spending a lot of time just screening the candidates? Are you using candidates in that process? Are there people sitting in your applicant tracking systems that you’re not talking to who might be a good fit? And 90% of the people were like, “Yes, absolutely. I wish I can solve that problem quickly.” “I would like to build a product that will help you quickly discover all those silver medallists that are sitting in your database that you are not reaching out to because you don’t know how to access them, right? Will that be of value to you?” Answer was, “Yes. A hundred percent.” “Is that problem being solved today for you? The answer is, “No.”

“Will you be ready to write a check for that? Or pilot that offering with me?” The answer is, “Yes.” So now as a technologist, I have a challenge that can I build a solution for that? But I know people are ready to buy that. So, that is what I mean by hone down your pitch. Now, I was talking to a customer, they’re like, “Yeah, actually, that is interesting but can you also solve the same thing for my current employees, internal communities?” Well, that led us to talent management. And I had this gut feel on day one was that job descriptions are useless. They’re poorly written. So I went to customers, “And by the way, you know as a recruiter, I’m sure you hate these job descriptions, hiring managers spend so much time on those things.” And they were like, “Oh. You understand us. You know what pain we go through.” Now, instead of leaving it on that note, my main thing was to come back and say that, “Okay, job prescriptions are really not useful.” recruiters don’t like it.

Can we do something about that stuff at our end? How do we go about taking ownership of that? And that led to us what we call jobs intelligence, structured definition of each and every job. So even when you’re pitching and people aren’t, you keep paying attention to what is resonating, what is not resonating, what is in the back of their own mind and bring that idea back to how you should innovate.

Amit Somani 16:10

Got it. One thing I know for a fact because we’ve spoken about it in the past, is the name of the company, Eightfold. Obviously, does it have something to do with Buddhism and Buddha and can you talk a little bit about that as well before we jump into some of the role of AI in what you guys are doing?

Ashutosh Garg 16:30

Eightfold comes from this Buddhist philosophy of eightfold path . I believe all of us need to get to that professional nirvana. And if we can get a better job, better career, it’ll lead us to that. So, that is what led us to picking the name Eightfold, that can Eightfold become the path to the professional nirvana of getting the right job, right career.

Amit Somani 17:00

Understand. So let’s jump and talk a little bit about what the product actually does. And in particular, very curious to hear about the role of AI as you look at both companies looking to recruit talent, and also manage their internal talent, right? In what way has it been… Now you’ve been at it for a while, you obviously scaled up nicely. So maybe some of the learnings on using AI as a part of this journey and what has worked, what hasn’t?

Ashutosh Garg 17:25

So what is AI? To begin with, AI is, the way I look at it. It is nothing else for something which is ultimate human scaler. Brings consistency, uniformity and data driven decision making. In the past, I’m sure when someone came to you, your niece, nephew, friend asking for career advice. Or if you go back to our IIT days, right? What did we do? We looked at our seniors, where they went and just followed that path, right? Oh, after IIT everyone is going to US for masters. Let’s just do the same thing, right? After masters, these people are doing PhD. Let’s do the same, right? See, our decisions were made based on a very small amount of data. Looking at the career of 10, 50, 100 people that we knew. What AI is able to do the same thing, but looking at the career of hundreds of millions of people, saying what path they have taken and what can I learn about the path they have taken that can map to my own career. Now, that’s at a high level. Now practically what is one of the biggest problems today in the market? Everyone is talking about skills. But the challenge is the skillset, the rate at which they come into the market and disappear is faster than ever.

Whatever both of us have studied in college is outdated completely. It doesn’t matter what we studied at some level today, right? In fact, the companies in which you are probably investing in five years back, many of those concepts are already outdated as well. The metrics that you look at have changed. Everything is constantly evolving. In this new world, if we keep looking for people who have done the work, you’ll never be able to find enough people for that. Instead, we have to identify the people who can do the work, who can learn that quickly. And that is where AI comes in. Can I look at the people who know that stuff today? Let’s say blockchain. What these people were doing two years back and look at other people like these people two years back versus these people today.

And that helps us expand the talent pool. Saying that, okay, you’re looking for a Python developer, but maybe you should also look for Java developer and C++ developer and C developer and any of these people can be a good fit for your Python developer job. Or you’re looking for a marketing operations person, but look for a programmatic analyst because they are likely to fill that role as well. Or you’re looking for a data analyst, but maybe quant administer is good enough as well. So being able to help you understand who else you can look for the same job, who can learn the system quickly.

Amit Somani 20:05

Understood. I do have a question on that. But let’s say I’m the candidate, looking at the candidate, not the head of HR or talent acquisition or the CEO that is hiring and I want to figure out what is my path. Back to your IIT example, How would you help the candidate figure it out? Because I may not know if I want to go into blockchain or healthcare or become a marketing operations guru.

Ashutosh Garg 20:30

We do the same thing for candidates. Idea is that for each individual, if you’re employee in a company, you look at every other employee, use their career paths in the company to say that based on who you are, what you have done, these are the 50 other people like you and here’s the career paths they have taken and you decide which ones you want to explore. And based on what you pick, then guide you through what path, what courses you should do, what projects you should work on, who can be a mentor, who can help you get there.

Amit Somani 21:00

Got it. I can’t help but just ask you, I don’t know if you’ve seen this movie Moneyball. Okay. Because it reminds me a little bit of that in some sort of abstract way to say figuring out talent. By the way, in some sense the business that I’m in of investing in early share founders, we’re always looking to find people who are going to be amazing founders. It doesn’t matter whatever pedigree, most of the founders we back are first time founders. So you don’t really know are they going to be successful or not or how, and you’re trying to do this pattern matching. And I think AI could play a great role there. But just love to hear your thoughts on Moneyball and what is the cost effective part of the talent there. But also figuring out that, hey look, I’m not trying to get somebody who’s a good first baseman or to use a cricket analogy, a good one down batsman. I’m just trying to find somebody who can rotate the strike in cricket. And if they can rotate the strike, maybe they can do well.

Ashutosh Garg 22:05

So firstly I’m not a sports person so I can’t use that terminology. But I do say that we are Moneyball for enterprises crucially identifying the right talent. And more importantly, identifying the right teams. And we are still a little bit far from that. We are still working on that problem. But as an enterprise, you are not hiring any individual in isolation. If I take your example, Amit, is when you’re looking for these first time founders, more often than not you’re really investing in one founder, you’re investing in two founders or three founders working towards the same goal.

And then you’re saying that together, these two people, do they fit the profile? Together, do they have all the skills that I think this founding team should have? Versus saying that each person needs to be best in everything. And that is the fundamental thing that we are trying to bring to the talent space. For any project, any work that you are doing, you need a combination of the skill set. Instead of looking for all those skill sets in each and every person on your team, if you can say that, what is the ideal composition of my team? What are the different skills I need in my team? And then what talent I can go after that will help me fulfill all those skill sets in the most effective fashion. So yes, it is the same thing.

Amit Somani 23:30

Understood. Wonderful. What are some of the trends you’ve seen in the talent market in terms of, and I mean more on the soft skills or the attributes of… Like you said, look, you may need to hire people to do work on some skills that have not even been invented yet. So there was no blockchain 10 years ago or there was no social media 15 years ago or what have you. So what are some of the traits that you have found looking through your AI and through all these talent companies that you’re helping that say these are interesting common traits that are required for success?

Ashutosh Garg 24:10

Well, I think I can say first from my human perception part and then we can talk about the AI part, right? The most important thing is that more often than not when we talk about those soft skills, we are wrong.

So here’s an example, right? When I was starting Eightfold, everyone around me said that you can’t be a CEO. Why? You are an introvert, you’re having an accent, you don’t like to go out, you don’t go to parties, you don’t mingle, you cannot be a CEO. Right? When you look into the research and data, what you find is actually introverts do make good CEOs. Some of the best CEOs in the world have been introverts as well. So quite a few times when we try to look at these simple patterns, we end up making mistakes.

So I did my psychological survey and someone came back saying that you are a very risk averse person so you can never be an entrepreneur. And the answer was, I am who I am now. This is where I am. But the bottom line is yes, I’m a risk averse person at heart. So when I take decisions, they’re given by that mindset, which doesn’t mean that I’m not going to do the startups, but what I try to do is keep minimizing the risk at every step in the process. So the only trait I would say that matters is you don’t give up, perseverance and you listen well. But third is, which may be an interesting controversial one is I don’t believe much in interviews.

I would rather look at people’s past, what career path they have taken, what choices they have made, how they have succeeded in those. Versus saying that I’m going to be able to judge people in one hour. There are few things I can judge, but more often than not, what I’m judging is how well this person speaks and tells a story, not how well this person actually does the work. So I try to focus a lot on what people have done to get here, and that is where AI plays a very, very big role. You joined Google in this year, you grew through the ranks. What does that say about your perseverance, your hard work, your ability to deliver? You have been constantly job hopping, but you are able to get interviews in every big company. So what does that say about your delivery versus your interviewing skills? So we try to look for those kinds of patterns a lot through AI.

Amit Somani 26:50

Understood. And so when you say interviews don’t work, and yet obviously you’re interviewing people or do you have-

Ashutosh Garg 26:55

Interviews work more in terms of validation, finding red flags.

Amit Somani 27:00

So do you guys not interview people at Eightfold or are you just largely focusing on what people have done in the past and so forth?

Ashutosh Garg 27:10

We try to focus more on what they have done and use the interview as a way to validate that stuff versus using the interview as a primary way to discover what you know and what you don’t know.

Amit Somani 27:25

Of course. Yeah. I’ll do the controversial part another day. Today we’re hearing stuff from you. I’ll take that off the record. Wonderful, Ashu this is very, very exciting. Maybe just a little bit, switching gears and talking about your own evolution as an entrepreneur, as a professional, what have you learned about yourself? What are skills you had to learn, unlearn perhaps, relearn? So maybe you can talk a little bit about that and that might help some of our listeners who are earlier in their entrepreneurial journey in terms of what they could learn from you.

Ashutosh Garg 28:05

The biggest thing that I learned is that how little I know. And I have to keep learning, keep asking people, keep questioning my own judgment all the time and seeking out to people to understand what they would do, how they’ll approach things. So my journey, if you can look at it, at the very least, they’re having three careers that I’ve had. One was a pure research career, second was a technologist at BloomReach, and the third is a business person at Eightfold. In the first part of my career, it was all about going deep into a problem, solving it. And really not too worried about everything else going outside. As I became a technologist, what became important was that, what is the value of that problem? Is it worth solving or not? What is the right answer? What is the quick answer? Time to value became extremely important.

So that perspective changed. But the biggest shift happened as I moved from being a technologist to a business person. And the hardest one that I had to learn or unlearn was as an engineer you are constantly looking for any bug. If nine things are going well and one is not going well, you focus all your energy on that one thing and make sure your code is bug free. As a business person, nine out of 10 things won’t go well. So you have to accept that not every deal will happen, not every sale will close, not every candidate you can close. So don’t get heartbroken when that happens. Every day, certain things won’t go well, certain things will go well. Enjoy the ones that are going well.

Try to learn from the mistakes and keep going. Other big part was people skills. Getting out of that comfort zone, getting on stage, talking to hundreds of people was a skill I had to develop. So I worked a lot on that stuff. How do I deliver the speech? So as a simple example, any time I used to give a speech, I would practice numerous times. Even now at times when I’m driving before the event, I’ll keep speaking to myself in the car by the time I get to the venue. And that’s one way for me to practice to perfect my delivery. So there are a number of things that I’ve developed, but I think it’s still the beginning of the journey.

Amit Somani 30:40

Absolutely. And you are on your Eightfold path to Nirvana. So learning and discovery and perseverance. Those are a couple of things I’m taking away from what you’ve shared. And obviously you’ve reinvented yourself a couple of times and I’m sure you will do that many more times before we are done here. So thank you so much Ashu for being on the podcast. And it was a pleasure to have you.

Ashu Garg 31:10

Absolutely my pleasure. Thanks Amit.
 

Enjoyed the podcast? Please consider leaving a review on Apple Podcasts and subscribe wherever you are listening to this.

Follow Prime Venture Partners:

Twitter: https://twitter.com/Primevp_in

LinkedIn: https://www.linkedin.com/company/primevp/