My Journey Building 3 Companies in 2 Years

Post Mortem of FriendHire Inc.

Recently I had to shut down FriendHire, a company I built to improve the job seeker’s experience. It’s taken me a few months to write this, as I needed time to process and reflect on what happened. I hope this write-up can help others who may be considering starting their own business. I worked on FriendHire and the result of two pivots, AppMeDaily and toptalent.network, for almost 2 years. I left Facebook in 2014 to build something amazing, hoping to make a difference in the world of recruiting.

It’s no secret many employees decide to leave companies when they have time to think about what’s next in their career. In my case, it was after my son, Leon, was born and Facebook granted me a generous four month paternity leave. I had been building software for 20 years both as an Engineer and PM at great companies (Lawrence Livermore National Lab, Microsoft and Facebook) and had the ambition to make job seeking awesome. I decided to take a year off and build my own startup.

FriendHire — Original Idea

FriendHire was inspired by the realization that the process by which candidates find jobs is fundamentally broken. I wanted to fix the following three things:

1. Current job seeker tools don’t reflect successful job seeker habits: Stats show the best jobs are always found through people you know. Interestingly enough, almost no platforms existed that allowed job seekers to find jobs easily at companies where their friends worked.
2. A company’s culture is not transparent: As a job seeker, I would love to have insights into why employees leave a company. Companies keep exit interview data to themselves. But even if they made this information public, the data is often inaccurate because departing employees are seldom forthright in their reasons for leaving a company.
3. Finding new roles internally is hard: I had colleagues at Facebook who looked for new positions internally, didn’t find anything, and then left the company. Networking is the number one tool to help internal job seekers find new opportunities. At Facebook, I built ChowRoulette, a tool allowing employees to sign up for randomly assigned 1:1 lunches in order to network and learn about new career opportunities within Facebook. I wanted to leverage what I learned from building this tool for my startup.

FriendHire started as an iOS app to help passive candidates find jobs through their friends. Once logged in, you can see jobs available to you. The job postings are associated with people you are directly (or indirectly) connected with at that company. You can use the inbuilt messenger to chat about opportunities of interest to you.

Within six months, FriendHire was released. It consisted of ~40k LOC, job listings (scraped), a messenger, push notifications, a networking feature, and information on why people have left a company. With help from a friend who was a designer, the app looked professional and was easy to use.

What happened?

I launched the iOS product in January 2015 without funding, and without having a strong user base. What I did wrong:

1. Built engineering focused: I fell into the typical engineering trap, where I had fun building the app and focused less on tackling the product. I didn’t spend enough time validating the idea and the need for the product. Instead of building an MVP, I focused on building too many cool features from the start.
2. Assumed getting funding would be easy: I was under the false impression that having an awesome idea and a tool to show would result in funding. After many discussions with VCs and Angels, I learned the idea doesn’t matter; it’s all about traction. Not having funding at launch impacted my ability to invest in user growth and scale the business.
3. No co-founder: Building a startup without a co-founder is miserable because you have no one to talk to about ideas or next steps, no one who challenges you and checks the sanity of the project, and no one there to motivate you besides yourself. I had no one aside from my astoundingly supportive wife, who shouldered the load of caring for our infant son in addition to her demanding biotech job — and moonlighted (literally) as my sounding board in all aspects of the business.

AppMeDaily — My own Advertising Platform

I had very few users, an unvalidated product, no funding, and no co-founder. Where do you go from there? Ultimately, I wanted to make my startup work and needed to overcome the impossible, fast.

I tried to focus on FriendHire’s user problem, which stemmed from the fact that user acquisition was very expensive (for me ~$5/user on Facebook). At some point I realized I couldn’t be the only one having this problem. There had to be others? I started reading the apple store RSS feed to search for apps written by individuals like myself who had experienced scaling issues due to advertisement cost. I developed the idea that if I could provide a cheaper acquisition for these app developers, I could become my own advertising platform — and ultimately solve the problem I faced with FriendHire.

Next I began working with my newly-recruited co-founder to build AppMeDaily, which allowed app developers to acquire users for a mere $2. Users would earn $1 in the process of installing and testing apps. Concurrently while building the website in AngularJS, I tried to confirm the need for this platform. I sent surveys asking whether people would test apps for money and how much money they would need to do this. I also sent over 700 personal emails to app developers to inform them of AppMeDaily and how they could acquire users more cheaply than through other channels.

The idea seemed solid because users received money for testing interesting new apps, which required minimal effort, and app developers got feedback on beta products without racking up bad reviews in the app store.

What happened?

The response rate from my >700 emails to app developers was very low. Turns out there are a ton of competitors in this space jockeying to help developers acquire users cheap. In this space, the solution didn’t matter so much as being able to get your message through to users who are bombarded with advertisements and marketing emails. This was not a battle I wanted to fight, so we stopped implementing after just two months.

We pivoted once more, this time reusing the AppMeDaily codebase. toptalent.network was born.

toptalent.network — Help Companies Hire from their Network

After a detour into user acquisition, I returned to my original inspiration to improve the recruiting world. The idea behind toptalent.network is to help companies source great talent through their employees’ 1st and 2nd degree connections. Think of it as LinkedIn with two primary differences:

• The ability to find 1st and 2nd degree talent based not only on your connections but the connections of everyone in your team or company
• The ability to distinguish between the quality of connections, as all connections are based on personal recommendations

How it works is that hiring managers prompt their teams and team members to invite 5–15 past colleagues with whom they enjoyed working. Within minutes, a team graph of 1st degree connections is established. Once these individuals join and invite their high performing former colleagues, 2nd degree connections are added to the team graph. With more and more teams participating, a company graph of talent is established.

Why is this interesting to companies? Companies receive candidates via the applicant pipeline, sourcing, or referrals. With every company I spoke with, I learned that the highest candidate to hire conversion resulted from referrals. Companies are very keen to see more referred candidates, hence the referral bonus programs at many companies. However, referral programs have a big limitation as they (usually) only reach 1st degree connections.

Toptalent.network was received favorably by a lot of companies because:

• It provides a tool for the otherwise manual approach (mostly done via whiteboarding) to bring a team together and gather their connections.
• It’s cheap in discovering 2nd degree employee connections.
• It’s based on recommendations and not resumes, which can yield “hidden gems” overlooked by other companies. This is also ideal for campus recruiting.
• It can help increase quality with regard to diversity hiring (think recommendations).
• It’s not limited to the tech space.
• It’s a really great tool to engage with passive candidates.

Why is this interesting to users? Now why would you or someone you know join toptalent.network? I researched what it would take for users to try a new recruiting platform, and here is what I found:

• Money is not that important, i.e. referral bonus.
• Being able to establish meaningful relationships with hiring managers in a similar domain (e.g. designers, database engineers) on an informal basis is somewhat important.
• Receiving regular insights to announced (and unannounced) jobs at companies where your friends work is highly valuable.

I incorporated these ideas into the design of toptalent.network. Ultimately, toptalent.network is like an elite alumni organization that includes all the people you trust and not everyone from LinkedIn. This is achievable because it’s invite only. I believe the sign up conversion rate for toptalent.network was quite high (20–30%) because people: 1) respond to a personal invitation from someone they trust professsionally and 2) appreciate the opportunity to see unique job opportunities at companies where trusted friends work.

What happened?

Mistakes always teach you something, and in a startup you need to learn fast. This iteration, we gathered feedback from both companies and users before we even started coding. We benefitted from being able to reuse most of the AppMeDaily code to build a prototype rather quickly.

The big push with toptalent.network was in business development, which was not my forte when I started this journey. I built and worked my network to score meetings and present MVPs at dozens of companies. The perception has been incredibly positive, even in unexpected domains such as biotech.

So where did I fail? It feels like there is an ocean between having an executive fall in love with a concept vs. actually use a tool. It takes more than business development to obtain commitment, and it takes real sales professionals to close a deal. Convincing a company to give you money to experiment with a new product is really hard. Oftentimes you won’t have a chance to speak with the decision maker, and more often than that you won’t hear back to find out what people really think. My best guesses as to why company participation was low: the business model was too complex (% of 1st year base salary), decision makers (often HR) were focused on reducing recruiting tools/budget, we didn’t have an integration story (into ATS systems), and HR departments have tried tools in the past and many have failed (so why try more?). However, all of these hurdles didn’t stop us. Our nemesis was time; we ran out of time!

Additional lessons learned:

• Have funding from the start, enough for 2 years to experiment
• For business applications, have a sales expert
• Spend enough time figuring out the business model up front

If I was to do it again

You learn such an incredible amount by running your own startup because, ultimately, you fail a lot! If I was going to do it again I would:

• Start with a great team with diverse skills
• Test the idea over and over (chats, surveys, MVPs, etc.) and the business model before writing any code
• With an MVP and a few successful customers, secure funding
• NOW is when the startup starts (and you leave your full-time job)

Why it was an Awesome Experience despite the Failure

This was one of the toughest and most rewarding experiences of my life, and here is why…

… it was tough:

• Startups are hard. Everything is unknown. I worked many late nights and pretty much every weekend knowing my time to experiment was limited.
• I ran an unfunded startup for almost two years. At the beginning I had a newborn son. Sometime in the middle my daughter was born. If you have kids, you get the idea.
• Startup days are filled with lots of ups and downs. It’s an extremely emotional ride. It’s up to you to maintain motivation for yourself, your team, and your family.

… it was rewarding:

• I failed and learned a lot in a small amount of time. I want to fail more because I get better at what I am doing every time I fail.
• I overcame barriers. I never imagined I’d be doing business development, meeting with executives at small and large companies across the Bay Area. And it was fun! I also met with dozens of VCs and Angels, an experience in itself.
• I made hundreds of new connections and friends, such as a former Apple Executive who shared his story of getting kicked out of Steve Jobs’ office. I am most thankful for the investors who sincerely believed in me and my vision and supported me throughout my journey with advice and introductions.
• I am incredibly thankful for the three awesome guys who allowed me to convince them to join my startup — without pay! I’ll find a way to make it up to you.
• I am thankful for having so many supporters who listened and provided feedback, who cheered me up on particularly rough days, and who invested their FB credit for this cause.
• I got to see my kids every day. Playing with them in the morning and evening was the highlight of my day and gave me the motivation to keep going.

Finally — thank you, Anna (my astoundingly supportive wife), for allowing me to do this!

Feel free to share …