Building a Sustainable, Capital-Efficient Business: Backblaze CEO Gleb Budman (Part 4)

Sramana Mitra: You started this in 2007?

Gleb Budman: Exactly.

Sramana Mitra: You self-financed it with your colleagues?

Gleb Budman: We did. It was definitely a long, crazy journey. Five of us committed to a year without salary. That would be the seed round. We then had two more people who joined us. We call them demi-partners. They weren’t complete co-founders but they did join very early on. They also committed time without salary, and they also put a little bit of money into the company.

Sramana Mitra: All of you had worked together before?

Gleb Budman: Yes, we had all worked together before. The five original founders worked for a full year and a half without making a single penny in salary. Then we raised a small angel round of around $370,000. It was largely to fund server purchases. At that point, revenue was also coming off of the company and we started paying ourselves minimum wage. It was a very exciting moment because after a year and a half of making zero salary, making minimum wage is great.

It was the first time we became legitimate. What I mean by that is we would go for things like healthcare. They would say, “Great. What’s the W2?” We would say, “There’s no W2. We’re running a company and not making a salary.” Even though it was minimum wage, it allowed us to be in a category that other companies understood. We made minimum wage for another year. Then, we paid ourselves double minimum wage for another half year. It was a long time of making far, far under-market salary.

Sramana Mitra: You all had exits before coming into this. You all had a little bit of money from your previous exits.

Gleb Budman: It was some combination of savings and exits. We all had some money saved up. For the most part, it wasn’t anything super tremendous, but it was enough to allow us to have the luxury to go for a year and a half without salary than to go for a sub-market rate and continue living in the Bay Area, which is not a cheap place to live.

Sramana Mitra: How long did it take you to get the product out?

Gleb Budman: We started actively working on the product in October 2007. We shipped it in private beta in June 2008. It was a little bit less than a year for the initial private beta. The first public launch was in September 2008. Basically, one full year from active development to when it was fully opened in its first incarnation.

Sramana Mitra: When did the first revenue hit?

Gleb Budman: The first revenue hit before that – probably six months before that. By revenue, I mean $5. We started charging as soon as we had a billing system that worked. Our first customer signed up and paid us money before we actually announced the private beta in June of 2008. A few months before that, it was in a state where we could charge money, we could back up, and restore. It wasn’t fleshed out enough to where we felt we could launch it into a lot of user’s hands. But, it worked. We started collecting revenue immediately.

This was a discussion and debate internally whether we should be charging people money while the service was in private beta. We decided that we would do it for a couple of reasons. One was, we needed revenue. Two was, you can’t really fully beta test a product if you don’t test the billing both from a business model perspective and from a product perspective. Are we charging correctly? So we started charging. We were a little nervous that there would be a backlash against it. In reality, only about three people said, “Why are you charging me while I’m giving you feedback?” We had a thousand people at that point who were using and paying for the product. Out of all those, there were only three who asked a question.

Sramana Mitra: It was only $5 a month, right?

Gleb Budman: It was only $5. It wasn’t some astoundingly large sum.

Sramana Mitra: Not exactly breaking the bank for anybody. How long did you keep the private paid beta open?

Gleb Budman: There were two main private betas. Initially, we had the private beta version of the Windows online backup service. We launched that service in June. Then we announced it publicly in September. Then we went into private beta later that year with the Mac version of the product. That was four months or so during which it was in private beta.

Sramana Mitra: How much revenue did you make that year?

Gleb Budman: In 2008, we made a mind-blowing amount of revenue. It was $35,000.

Sramana Mitra: You were charging $5 a month. That’s not a bad revenue.

Gleb Budman: That’s true, but at that point, there were seven of us that had worked full-time and put money into the company. We were paying for servers and everything else. We were looking at it and saying, “If this doesn’t ramp up quickly, we’re going to have a very painful situation on our hands.”

Sramana Mitra: So you closed 2008 with a product that was pretty much working, right?

Gleb Budman: We did. The Windows product was fully functional. The Mac product was fully functional but it was still in private beta because we wanted to make sure that there weren’t any bugs of significance.

Sramana Mitra: Talk about 2009 customer acquisition strategy.

Gleb Budman: It took us a while to really figure out customer acquisition. Frankly, we’re still working on it today. What happened in 2009 was a lot of one-off experiments. At that time, we were a very small team. We did round robin customer support. The founders took a day each during which they would be responsible for support. When it was my turn to do support, one of the things that I would do was when somebody emailed in and they had a signature on their email, I would Google them and see who they were just to get a better understanding of who our customers were.

One of the people who emailed in was a guy by the name Amit Gupta. I Googled him and saw that he ran this site called photojojo.com, which I had never heard of. I started chatting with him a little bit and he said, “I love your service. I want to email my newsletter about it. It’s a newsletter for amateur photographers who want tips and tricks of what to do.” I thought that it sounded great. We talked about it a little bit and he included us in his newsletter. It turns out that he had something like a quarter million in his newsletter. The next day after he sent it out, we had thousands of people who signed up. It was very successful. It wasn’t completely repeatable because he couldn’t send out newsletters every day. We introduced an affiliate program around that time where we could pay people a percentage of the revenue of customers they brought to us. Over time, that became a meaningful, but not dominant, percentage of the business. What really hit for us was that in late 2009, we published a blog post that just went insanely viral.

Sramana Mitra: What was it about?

Gleb Budman: We had been writing blog posts since the start of the company. I think the only person who read them was me and maybe my mom. This blog post had over a million people read it. Initially, we were planning on using Amazon as the storage. What we ended up doing instead was we realized we couldn’t afford S3. We would lose money on every single customer because at $5 a month unlimited, it was just going to be too expensive to store data on Amazon. Then we thought about using equipment from EMC or NetApp. All of that was going to be expensive.

We ended up going into this crazy path of designing our own storage, building our own equipment, writing our own software, and running our own operation. Basically, we built a cloud storage service for ourselves. No one knew how we were able to afford $5 unlimited. People had all these theories. One theory was we must have tons of venture capital and that we were burning through it. The other theory was, “They’re going to sell all of our data and monetize it,” which is not true. The third theory was we were not keeping the data at all. We were just hoping that no one would ever do a restore, which is not true as well. We had built this very interesting and low-cost storage. No one knew about it but we were actually profitable in providing our $5 a month unlimited. We decided to write about that. We specifically focused on the hardware itself – what we call the storage pod.

A storage pod was a storage server that held 45 hard drives and connected it to the Internet for the lowest possible cost. Then we decided that we would share a little bit of history behind it. Then we decided that we would actually share the design. Finally, we decided that we would open source the hardware design itself. We did that in this blog post. We had a chart in there that showed the relative cost of a petabyte stored on Backblaze storage pods versus Dell, EMC, NetApp, and Amazon S3. On Backblaze storage pods, storing a petabyte costs $170,000. On NetApp, EMC, and S3, you were looking at a couple of millions of dollars. It was wildly less expensive to use the storage system that we had designed. Here we were giving it away. The first day, the press wrote about that and about a quarter million read the blog post. On the second day, it went completely viral on Hacker News, SlashDot, and Reddit. Over half a million read it the second day.

It was just eye-opening for us that there was so much interest in storage if you were willing to share and give things away. It became part of our customer acquisition strategy to think about what people want to know about storage and can we provide insight into that. Over the next couple of years, we continued to iterate the storage pod design and improve the way we store data with that. We open sourced and shared the design of 2.0, 3.0, and 4.0. Just about a week ago, we shared the design of 5.0. It’s certainly not a direct translation. It’s not like someone reads one of those and goes, “I love this humongous storage server that Backblaze designed. Now, I’m going to sign up.” A certain percentage of people do. What we found was by continuing to publish these blog posts, a percentage of people would go ahead and sign up for the online backup service.

This segment is part 4 in the series : Building a Sustainable, Capital-Efficient Business: Backblaze CEO Gleb Budman
1 2 3 4 5 6 7