Twine founders Joseph Quan and Nikhil Srivastava talk about their award-winning HR analytics startup.

The winner of Penn Wharton Entrepreneurship’s first ‘Startup Challenge’ is Twine, an HR startup that uses algorithms to find the best internal candidates to fill a company’s open positions to boost employee retention and reduce the cost of new hires.

Founders Joseph Quan and Nikhil Srivastava, Wharton MBA students from the class of 2017, won in a field of 29 semifinalists after making their pitch to alumni and others to compete for funding, services and support. They received a grand prize of $30,000 plus $15,000 worth of legal, accounting and strategy services. Twine also is a member of the university’s VIP-Xcelerate, a 4-month startup incubator program.

The founders spoke with Knowledge at Wharton, which airs on SiriusXM channel 111, to talk about their startup.

An edited transcript of the conversation follows.

Knowledge at Wharton: Where did the idea for Twine come from?

Joseph Quan: The genesis of the idea was a little bit circuitous. About a year ago, Nikhil and I came together with this idea of building a people-recommendation engine, essentially software that would help any person inside of a large organization algorithmically identify with other people they should be meeting. This was mainly for networking and mentorship purposes. We ended up piloting it, building it out at Wharton, signed up over 1,000 classmates — about 60% of the class — within a week.

… We had two fundamental hypotheses as we [sought] to build this out from a side project into a business or a company. First, we wanted to go B2B (business to business) rather than B2C (business to consumer) — sell it to institutions because a lot of value accrues to institutions over individuals when you deploy a service like this. And secondly, to go [to] corporate [clients].

We ended up taking this to the University of Pennsylvania and Wharton. They actually became our very first paying client. From there, we said if we validated the fact that an institution will pay and get value out of this, let’s take it to the corporate market. We ended up taking it to Fortune 1000 companies. … We discovered that while this was an idea that was really interesting to HR executives, there wasn’t enough pull in the short term to push it forward. But we did end up surfacing this overarching problem around talent acquisition, talent retention and recruiting that was immediately actionable for a lot of large companies. They care a lot about keeping and retaining their top talent.

Knowledge at Wharton: I can’t tell you how many times we’ve talked about this issue and the losses that companies see each year from either not being able to match up a job with the right person, not retraining the people they have to fit those jobs, or losing people who don’t feel like they are doing a job that they want to do.

“Internal hires generally tended to perform about 35% better … and they tended to cost about 20% to 30% less.” –Joseph Quan

Quan: You’re witnessing a sea change in the labor market where these millennial employees are jumping ship every two, three years because they’re feeling like they’re outgrowing their roles. We find that there’s a subset of 10%, 20%, maybe even 30% of the market that really has forward-thinking HR departments and people analytics departments that focus heavily on those issues, but it just didn’t get the reception initially that we had wanted. We are starting to work with this pilot contingent of companies that are really passionate about these issues, and they’ve been great strategic partners in helping us advance this mission.

Knowledge at Wharton: What other companies are interested in this type of project?

Nikhil Srivastava: We estimate that the average Fortune 500 company loses $100 million every year to preventable employee turnover. It’s a huge cost, and in many ways it’s a hidden cost. To answer your question, we signed up Wharton last year. We also work with a handful of other small and medium-sized businesses. About four months ago, we signed up our first Fortune 1000 client. That’s Nielsen based out of New York. We’re doing a project now with 10,000 of their North American employees, 20 of their internal recruiters. And we are currently building up our sales pipeline of other large institutions. We are mid-pipeline with a couple of those and in the pilot stage with a few more.

Knowledge at Wharton: When Nielsen comes to you, what are they specifically looking for?

Srivastava: [They wanted to solve] two big problems. One is that their top talent often leaves the company. You look at the average tenure, especially of millennials and junior employees, it’s now two or three years. The average student coming out of college is expected to have 12 jobs in their lifetime. Thirty years ago, that was five jobs. Every time they lose someone, it costs anywhere from 50% to 100% of their annual salary to replace and retrain someone. The first problem is just losing skilled people. The no. 1 reason these people leave is because their jobs no longer fit their skill sets and their aspirations.

Second is this issue: We have these skill sets and this knowledge base, but is it being effectively deployed in our organization? [Should we] move this person from department A over to department B? Put them in a better role where they’ll be more engaged, more productive? And it’s better for us because instead of [their] going back to school, more employees are receiving on-the-job or experiential training in this new labor market.

They brought us in to solve both of these problems. We do it through deploying these algorithms to match people into new roles, and they’re using a web application to surface these recommendations.

Knowledge at Wharton: Not only is there cost savings from incorporating people that are already in the company, but there’s a bottom-line benefit to the company to have somebody that is successful, motivated and really engaged in their job.

Quan: That’s exactly right. We’ll give a shout-out to Wharton professor Matthew Bidwell, who did research on this and published something through Knowledge at Wharton a couple of years ago where research from his department showed that internal candidates tended to disproportionately outperform external candidates. Obviously, you don’t want an organization that’s 100% internally filled because that starves it of external innovation, but a lot of companies have a great potential to up their rates of internal hiring. They found that internal hires generally tended to perform about 35% better when you looked historically and empirically at all their performance ratings, and they tended to cost about 20% to 30% less. There’s a huge benefit and a huge boon just by leveraging the implicit knowledge that they’ve developed inside the organization and all the existing networks that your employees already have inside the company.

Knowledge at Wharton: What was this experience like? When you’re doing the finals of the competition, you are presenting this idea to not only people of Wharton and the University of Pennsylvania but to outside business people as well. You are reaching potential clients in the process.

Quan: That’s why we were there. To some extent, we’d been training for that for the last six to 12 months, not by explicitly getting involved in these entrepreneurship or pitch competitions, which we’ve always actively avoided in service of really building up the business. But we had been speaking with sales prospects, with advisers, with investors for the last six to 12 months nonstop. It was very much a return to form to do that through the business plan competition as well.

“The average Fortune 500 company loses $100 million every year to preventable employee turnover.” –Nikhil Srivastava

About a week before that competition, Nikhil and I sat down and really thought about selling this. What isn’t traditionally considered the sexiest company? It’s a B2B enterprise. You’re not building an Uber or an Airbnb, so it’s a little bit more difficult to relate to sometimes. We just made sure that we crafted a narrative that would really resonate, not only in the enterprise side, which we’ve always focused on, but also on the consumer and the employee side.

Knowledge at Wharton: It doesn’t have to be the sexy idea, as you said. It has to be the one that meets a need and is the most effective.

Srivastava: That’s right. It’s very much a horizontal solution. Any company where we find 3,000 or more employees is when you start to get a network that’s complicated enough so you don’t know who to put into a new role. But if you think about it, there’s tens of thousands of those businesses out there. Right now, we’re focusing on a few different verticals, so data and analytics, financial services, med and pharma. But we hope and believe that this solution can apply to any big company. We really want to be that engine, that connective tissue that allows companies to maximize the value of their human capital by putting people in the right roles.

Knowledge at Wharton: You mentioned the 3,000-employee plateau. Is that the point when HR resources just are not enough to handle thinking about the roles that could be potentially filled within the company?

Srivastava: In many ways. Typically, when you are a smaller or medium-sized company, your primary HR focus is on growing and bringing in the best people as you’re scaling out the business. When you get to a slightly larger size, all of the sudden you have additional complexities. You might have different departments, you might have silos of people, you might have recruiters that don’t know who’s sitting on a different floor or in a different building even though they might be a great fit for a role. That’s when software and algorithms can really help to improve upon these human decisions of where to put people.

Knowledge at Wharton: Wharton is well-known for having these competitions, and some of the ideas that have come out of them are some of the biggest names out there. Warby Parker is one of the premier ones from the recent past. With something like this, there is that potential to have an impact that can be ground shaking going forward.

Quan: Absolutely. We see this as a really unique opportunity. There’s a lot of startups that say, “We’re changing the world. We’re affecting millions of lives.” It’s very rare for that to be defensible outside of this realm of human capital analytics software. Our software really touches human lives and meaningfully affects the career trajectories of millions of people, or has the potential to do that.

In the short term, what we really want to do is just nail this internal mobility niche, really make sure that we’re building the best possible software solution that will help large companies retain and redeploy their talent in the most effective way. Essentially, we want to help them establish more internal liquidity in their internal job markets, help them move people to the right openings inside their companies. And then on the employee side, empower these employees to grow with their companies rather than growing out of their companies and jumping over to competitors.

“We really want to be that engine, that connective tissue that allows companies to maximize the value of their human capital by putting people in the right roles.” –Nikhil Srivastava

Longer term, we see this technology branching out, much as you alluded to Warby Parker’s scale, to become a much more powerful product that helps any large organization algorithmically identify the best employees for any potential purpose. You can see these algorithms being used not only for recruiting purposes or internal hiring purposes, but for team formation. How do I find the best possible people to form this task force? How do I find the right people to fill into these different organizational gaps? How do I get different mentors and mentees to meet each other?

We’re seeing early inklings from the market right now of this potential to use software and this algorithm-enabled technology to help find the best people for any cause. That being said, right now we’re just kind of focused on one and we’ll expand from there.

Knowledge at Wharton: With any startup, there are potential pitfalls as you’re developing this process, correct?

Quan: Absolutely. One is the quintessential problem of finding product market fit in the first place for startups. For those uninitiated, it’s are you tackling the right market and is it a big enough market and is the market mature enough to take on the product that you have? And are you building the right product for that market in the first place?

A lot of the early challenges that we faced were really around customer discovery. We went into the market with a couple of hypotheses about the problems that we wanted to solve, but ultimately we needed to have 20, 30, 50 conversations with HR executives — people in the field getting their hands dirty in the HR function every single day — before we got enough fidelity and enough clarity into the problem that we were solving. What Twine is today has really been shaped by the product of dozens of conversations that we’ve had and a lot of feedback. It hasn’t been something that’s just come straight out of our heads.

Srivastava: The one thing I’d add to that is just the inherent pitfalls in the enterprise sales cycle, which are very long and often painful. Luckily, Nielsen is not only a client of ours but a strategic partner, so they’ve helped us build a product for the industry and for the market. But just going through legal and privacy and finding the right person in the organization and procurement, that’s a process.

Knowledge at Wharton: What were some of the best insights that HR representatives gave you?

Quan: One of the major insights that we gleaned from our initial conversations was this breadth of data that a lot of large organizations have that’s being underutilized today. Any large organization collects hundreds of records on every single one of their employees that’s captured throughout the value chain, so to speak, of an employee’s life. Before they join the company, when they’re applying, after they join, when they take courses and certifications — a lot of companies have all this rich data about their employees, but it’s disaggregated and disorganized across many different platforms and databases. One of the most insightful things that we learned that ended up forming a core thesis of ours inside of Twine was companies have great assets at their disposal and no good analytics or recommendation layer that sits on top of that and helps them make sense of it.

“We realize that it’s part of our responsibility to help create this market.” –Joseph Quan

Part of what we built Twine [for is to act as a] connective tissue that will help aggregate and consolidate all this data about your employees in one unified view, and then provide analytics on top of that to make it really intuitive and actionable to say, “Hey, now I know who are the five, 10 best employees I should be moving around my organization to fill open roles.”

Srivastava: I think another insight we got from going out to the market was that this is a cultural issue as much as it is a technology issue. You can build these predictive algorithms that will maximize how successful someone is in a particular role. You can figure out how likely they are to quit. You can suggest new jobs for them. But there definitely is a culture at the company.

Often, individual managers can be territorial. They might not want their employees leaving for other departments. What we’re focusing on initially is the early adopters, the companies who are forward-thinking in their HR departments, who believe in internal mobility, who believe in investing in the development of their employees. But that’s definitely a challenge that we are going to have to face down the road because you can deploy a technology, but unless you have the buy-in from the top, unless you have a culture to support it, it’s going to be hard to make it effective.

Knowledge at Wharton: As much as anything, this is the mindset of the HR department, of the managers, of the C-suite. If you don’t have the proper mindset, then it’s very hard to make this work and be able to improve your company.

Quan: It can be intrinsic and extrinsic, the way we see it. Absolutely, we’re targeting those first wave of companies that have this intrinsic mindset of employee development, career development, helping their employees grow with the company rather than growing out of their companies. At the same time, we realize that it’s part of our responsibility to help create this market.

What I mean by that is it’s our responsibility to educate the market, it’s our responsibility to put out this research and evangelize a lot of the great research that’s come out of Wharton, whether it’s from Peter Cappelli or Matthew Bidwell or Adam Grant or Cade Massey, the heads of the People Analytics department out here, to show that there are huge benefits to internal mobility and internal hiring. First, it’s kind of penetrating the early part of the market where people really get it, but very soon it’s going to become educating the market and nailing a lot of those extrinsic, external factors.

Srivastava: That’s a process that right now we are in the early stages of, but it’s a long journey. One of the things that Joseph and I have had conversations about from the very beginning is that this is a five-, seven-year road, especially coming out of school and working semi-full time for the last 12 months. But now we graduate and the rubber hits the road. We’re ready for the grind.