Throughout his career, Joel Spolsky has seen many sides of the technology industry: from serving as a Program Manager on the Microsoft Excel team, to founding Fog Creek Software, to penning the popular developer blog “Joel on Software”, and now, as CEO of Stack Overflow, a question and answer site that has become the essential destination for programmers. We sat down with Joel to talk about being the leader of a fast-growing company, how (and how not) to use data in decision-making, and where Stack Overflow is headed.
There are a lot of programmers that dream of being a CEO one day. What was the spark that caused you to make the jump?
I’d like to see more programmers as CEOs, for sure. For me, I was mostly just trying to solve a problem. Before we started Stack Overflow in 2008, there was a good amount of programmer discussion on the internet, but it was scattered across various locations. The biggest site at the time was actually a paid offering, which I thought was ridiculous. I kept asking myself: how hard can it be for a developer to create a new discussion group for everyone? So, I got together with my co-founder Jeff Atwood, and we built Stack Overflow.
It looks like you found a good problem to solve—the site took off like wildfire.
It was fast—we were dominant within six to nine months. We’re now the top developer website in the world and in the top 50 of all websites on the Internet. 800 million pageviews from 50 million unique users per month. We’ve got 250 employees, growing bookings at about 65% per year.
There were a bunch of factors that made the site successful. We knew we were on to something as fast as the second day when the numbers had doubled from day one. After about a month or two, I was presenting at a Microsoft conference with maybe five thousand programmers in the room, and I decided to ask the audience if they had used the site. About a third of them raised their hands. When I ask that question these days, I just get laughter in return. It’s like asking “do any of you use gravity?”
From a funding perspective, how did you evolve the business?
The first couple of years were self-funded and profitable from advertising. We focused on recruiter ads because they are the most lucrative, and everyone wants to hire programmers. It’s still our main source of revenue. We decided to take a few rounds of funding in part to make ourselves more disciplined as a business. Fairly recently, we’ve turned the corner back to profitability, so now our destiny is in our own hands.
As a leader, how important is data in your decision making?
Everything is about data for us. We try to be as data-driven as possible, but there’s always the risk of falling in love with the wrong metric, especially with incentive programs. For example, as a frequent flyer, I’ve noticed certain airlines have started measuring their performance through the metric of on-time departures. We can all agree that being on time is a key part of any travel experience, but if you sacrifice everything else to push away from the gate on time—no longer taking the time to help passengers with a bag, or processing upgrades at the gate—the program is going to backfire. These airlines have made the mistake of assuming on-time departure is a proxy for good customer service, but it’s actually ruining every other part of the experience.
Have there been times when data has changed your mind or the direction of the company?
Probably so frequently that I can’t just give you one example. For me, the biggest eye-opener has been getting better access to financial data. Back when we started, I looked at the financials at the end of every month, asking “how much did we make?” But as your business grows, individual months tend to matter less and less. I’ve learned to lean into quarterly data because the monthly data can show too much noise and variation.
What’s been the biggest learning for you as a CEO?
That being a good CEO is focusing on your people, especially when you’re growing as fast as we are. You’ll run into instances where the right person to do the job at one stage of the company can’t fill that role at a larger level. It puts a lot of stress on the organization, and that’s something the CEO needs to actively plan for and manage.
What’s the next step for Stack Overflow?
We’re launching a whole bunch of products for people with private questions that they want to ask internally on their own teams—stuff they don’t want out in the public eye. There’s a version of the product called Stack Overflow Enterprise for big companies, and shortly, we’ll also launch a product for teams. There are other team-based products available out there—like email and discussion groups—but not with the Stack Overflow brand and the ability to store and archive specific questions and answers. Really, any way that we can help developers learn and share their knowledge with others is where we want to be.
*Article edited for clarity and brevity.