3 Things I Learned at LinkedIn (and 2 I didn’t)

It’s a neat feeling when a story about your grandfather helps you contextualize your own journey.

When Poppop, my grandfather, was a young man, he worked in his father’s store. He would keep detailed records of what people were buying. Poppop tried various promotional strategies to improve sales, then closely examined the results. He found that he had a natural talent for gaining insight from looking at numbers; he realized that gave him a big advantage in business (he didn’t need lots of data or fancy A/B testing).

Later on, Poppop would bring those skills — and a natural ability to relate to people — to a business he started with his brothers. Their clothing company, Rockower Brothers, would go public in 1962; it was acquired by Woolworth in 1978.

That progression — from data-obsessed junior employee to entrepreneur using data strategically — sounded much like my own. For both me and my grandfather, mastery of data and math would be one step of several on the trail toward starting a meaningful company. Unfortunately, he died when I was young, so I never got to directly ask him about those experiences.

It’s tempting to paint Silicon Valley as a unique haven for progress with no historical precedent. As with many similarly bold pictures, that painting contains elements of truth, but doesn’t tell the whole story.

In the first post in this series, I wrote about my relatively confined role at PayPal; it made me comfortable with startup life but didn’t really prepare me for entrepreneurship.

My experience at LinkedIn — where I was the first analytics employee and led the analytics team — did much more to prepare me to start a company. That difference is more about my role than about the company: despite having a strong vision, LinkedIn was not tactically great and the team was not as strong as PayPal’s (it’s unlikely we’ll ever see a substantial “LinkedIn Mafia”).

Specifically, my time at LinkedIn allowed me to grow in three areas:

1) Product ecosystem and distribution

Silicon Valley’s priorities are changing. In the past few years, we’ve seen a collective appreciation for the importance of user experience on product. In the last year or two, the term “data scientist” has become popular, but almost no one understands what big data really means(here’s my answer). Very recently, the term “growth hacker” has become popular (I’ve been so branded), thanks to an understanding that getting people to use your product isn’t just something to hope for.

But almost nobody talks of the importance of understanding an Internet company’s usage ecosystem. Take a new user, and project her activity out a few months. Is she going to get through the signup flow? Engage with the product in different ways? Invite friends? And more important, how does this behavior affect other users on the site? Growth hacking requires an understanding of a subset of the usage ecosystem, but the usage ecosystem is much more than that. A company that does well on growth hacking but poorly on creating an ecosystem ends up providing very little value for its users. That was where we found ourselves with Circle of Friends — Circle of Moms’ predecessor — and a big part of why we decided to pivot.

At LinkedIn, a big chunk of my time was spent understanding the ecosystem. Before that experience, I had only a superficial understanding of virality, never seeing how small changes can make big differences. I knew nothing of the effects of different channels — transactional emails, marketing emails, social emails, SEO, partner referrals, press — on a product’s ecosystem. At PayPal, I crunched numbers. At LinkedIn, I wrote email copy to boost numbers and realized the importance of understanding user psychology. Those skills and interests would transfer nicely to the experience building out Circle of Moms, arguably becoming my strongest asset as a founder.

2) Leadership

LinkedIn was my first experience hiring and managing. There’s a lot of (mostly superfluous) writing on whether leaders are born or made. I won’t generalize, but I will state that I certainly wasn’t a born leader: 26-year old me was a much weaker leader than 34-year old me is.

26-year-old me was a very awkward manager. I felt bad asking my employees to do things for me and I was overly deferential to both my own team and others in the organization. To some extent, that’s my personality and will never change completely. But in two years managing people I gained a degree of confidence and authority as a leader that is absolutely essential for a co-founder. If in 2004 I’d started a company instead of joining LinkedIn, I would have ultimately ramped up and gotten used to managing people, but it would have been a bumpier process (and Circle of Moms was bumpy in its own right).

At LinkedIn, I hired two full-time employees and one part-time employee. All three were older than I was and more educated (two with PhD’s). Hiring was a new experience, and an educational one. In the hiring process, I saw a contingency recruiter misrepresent both sides to increase her own take; I also had a candidate accept an offer from us, only to change his mind when his then-employer made him a counter-offer. When I started (what became) Circle of Moms, I was hardly a hardened and savvy hirer, but I’d at least been around the block once or twice.

Of course, management experience is a double-edged sword for entrepreneurs. We all know stories of those who easily managed hundreds of people in big companies, then struggled to be hands-on when part of a tiny startup. My pre-startup experience — managing a couple of people at LinkedIn but still being mostly hands-on — was about the right prep to start a company. For those who more naturally gravitate towards management, leading a team pre-startup is probably unnecessary.

3) The hubristic belief that I could do better

I continued to build out my longtime side project, Team Rankings, while at LinkedIn. It was becoming more and more of a business. In 2004, I started to charge for premium NCAA Tournament content. In 2005, I started publishing predictive models on the site and charged people to access them. Later that year, Tom Federico started working on the site with me, and we raised the bar on product polish and the branding. Come 2006, we were really cranking on our March Madness product: we built an awesome tool called BracketBrains to help with office pool picks. That March, I was constantly tweaking the site’s SEO, scrambling to keep the server up as traffic surged, and optimizing our marketing messages in real-time. It was exhilarating.

Meanwhile, things were moving along pretty slowly at LinkedIn. I was firmly planted on the analytics side of things. I couldn’t write production code, and had limited places to make user-facing changes. A few months earlier, my friend Jawed had pinged me about joining his new startup called YouTube. I hadn’t been interested then, but after that March Madness experience, I saw things differently: YouTube’s early growth, quick speed of iteration, and small size (around ten people) had some appeal. I had significant discussions with YouTube but didn’t wind up joining them; I probably would have if I’d found online video as interesting and important as a professional network.

In the months following, I became more frustrated with both the slow speed of development and what I thought were some poor tactical decisions. Because I was so close to the numbers, LinkedIn’s decision-making bugged me, and made me think I could do better. Because I had built up Team Rankings and been able to quickly iterate on it, I figured I could also build products better and more quickly (to those who have had the misfortune of working with my code: insert a snide remark here).

What I didn’t learn at LinkedIn

When I started Circle of Moms, I was a crappy software developer (I’m still no superstar). I could write code pretty quickly and solve scalability problems in clever ways, but I didn’t do a very good job architecting a system. That wasn’t the end of the world — we had lots of good engineers who eventually made our code base fairly solid — and I’d argue that for very early consumer Internet startups, an A+ engineer is only a nice-to-have.

I also didn’t have much of a business focus at LinkedIn; I didn’t know or understand much about raising capital, equity, finances, or the role of sales on LinkedIn’s success. At Circle of Moms, my co-founder Ephraim — who was CEO while I was CTO — focused more on these areas. My lack of knowledge here certainly didn’t help, but probably wasn’t the end of the world: it’s much easier for a good product person to learn about business than for a good business person to learn about product.

Moving on

In January 2007, it was time to move on. I told Sarah Imbach (my boss) and Reid Hoffman that I was planning to resign and ultimately start my own thing. Reid sat me down, seemingly to convince me to stick around. In a similar conversation nine months prior, Reid had convinced me that LinkedIn — and not YouTube — was the place to be (it’s nice when your fork in the road leads to honey on one side and sugar on the other). But the dynamic for this discussion was very different: I’d found the honey for myself at the end of the LinkedIn road and wanted to move on to climb a big but as yet undefined mountain.

In that conversation, Reid revealed to me that LinkedIn was very close to bringing on a CEO to replace him, describing the candidate as a very well-respected #2 person at a public company. I was somewhat taken aback, but it made sense: Reid understands his skills (product vision, many different types of strategy) and weaknesses (operational details) as well as almost anyone. It was probably a last-ditch effort to get me to stay, the implication being that LinkedIn would soon be better led and better run. The new CEO, who it turned out would be Dan Nye, would hopefully address the weaknesses I saw in the company and its culture.

Alas, I’d made up my mind and knew where I wanted to go. I’d grown in three very important areas. Like my grandfather who moved on from his father’s store, I was ready for a new challenge.

Mike Greenfield founded Bonafide, Circle of Moms, and Team Rankings, led LinkedIn's analytics team, and built much of PayPal's early fraud detection technology. Ping him at [first_name] at mikegreenfield.com.