Memcachier: How to Gain Traction Through Partnerships & Avoid Becoming a Commodity

Interview with David Terei, CEO and Co-founder of Memcachier

By Poornima Vijayashanker

When Amazon launched ElastiCache, an in-memory cache web service, David Terei and a fellow Stanford PhD student named Amit Levy were fascinated with the technology. Rather than being intimidated by the giant, they were inspired to make a better version of the product that would make developers’ lives easier. They talked to potential developer customers, validated the need for the product, and then built MemCachier. Read on to hear how they distributed it successfully on Heroku, balanced startup life with their PhD program, and who, in David’s opinion, should be a startup’s first line of customer support.

MemCachier manages and scales clusters of memcache servers so developers can focus on their apps.

Poornima: “Hi David, thanks for taking the time to do this interview with me. I’d like to start by learning a little bit about your background. Tell me where you’re from and why you chose to move to the US.”

David: “I grew up in Australia and lived in Sydney until I was 24. I received an undergraduate degree in Computer Science, and during my studies I wrote a Haskell compiler. I always knew that I wanted to work overseas. Then I was presented with an opportunity to come and work under David Mazières at Stanford, working in the systems and security group.”

Poornima: “How far along are you in your PhD and how do you think it’s benefited you?”

David: “I’m hoping to be done in a year, and it’s actually how I met my co-founder, Amit Levy.”

Poornima: “How do you split your time between working on your PhD and building MemCachier?”

David: “We took time off initially, for the first 8 months. Since then I’ve been back at Stanford. I spend my weekdays running the business and the weekends doing research. MemCachier also provides opportunities for research.”

Poornima: “How do your PhD advisors feel about MemCachier, and were there any issues with Stanford and IP ownership?”

David: “My advisors are understanding, and the IP is not an issue. It would have been if an advisor was involved with building it. Stanford has been pretty amenable. MemCachier is Amit’s and my business, which we own. It’s not based on any research.”

Poornima: “What was the inspiration for MemCachier?”

David: “It was a little bit roundabout. Initially, we saw that Amazon released a product called ElastiCache, and we thought it was going to be a competing product. But we were actually pretty disappointed with it.

Then, Amazon released DynamoDB along with a publication and talked about how it was a key value store. The research paper was pretty inspiring. It talked about utility computing, distributed computing, and making things really plug and play.

ElastiCache was essentially EC2 (Amazon Elastic Compute Cloud) with a memcache pre-installed. But it was less than ideal because you still had to manage servers and there were a lot of options you had to configure. The decisions meant developers had to do a lot of work, and that’s when we realized we could do something better! It inspired us to build our own system.”

Poornima: “Once you figured out that ElastiCache wasn’t solving your problem, what did you do?”

David: “We didn’t jump into starting a business or building. We started talking to our friends who were in startups building software products and got feedback that a plug and play cache would be valuable.”

Poornima: “Great to hear that you did your customer development before diving into building! Once you had built your product, how did you go about distributing it on Heroku?”

David: “We got into Heroku pretty early. They were still building out their marketplace of apps. They reached out to us personally. Then we had to prove to Heroku that it was a product people wanted. We told them about the feedback we received from our friends who were in startups and said they needed it. Heroku had a fairly standard process for integration. We were initially in beta and then became available to the general public.”

Poornima: “What did you learn through the process of working with Heroku?”

David: “Working with Heroku early on was a great lesson for us. We learned the importance of building a relationship with partners, and how important it is to have ones that are straightforward and easy to work with.”

Poornima: “That is a great lesson! How far along are you now with the business?”

David: “We got traction pretty early on and have bootstrapped our way to profitability. We also have a good network of mentors that have helped us build the business.”

Poornima: “So has it been all smooth sailing?”

David: “No. I’d say I’ve had my best and worst experiences building the business. I’ve had to wake up at 3am because servers were crashing, which was not fun. But then we’ve had amazing customers sign up, and those interactions are really valuable. Managing a team and working with a co-founder is also tough. I’ve been fortunate to have a good experience with Amit.”

Poornima: “What is the ultimate goal of this business?”

David: “We want to build a business we are proud of. We aren’t driven to build a billion-dollar business. We used to have a third co-founder who wanted to build something bigger. Ultimately, he decided to move on to do that.”

Poornima: “Who is on your team now? Is it just you and Amit?”

David: “No, infrastructure is hard because we have to have high uptime. We cannot do a hack job. We’ve had to learn how to make tradeoffs like managing manual tasks or processes and then figuring out when the right time is to automate them. We’ve hired contractors who work in different time zones, and that helps deliver high uptime to customers.”

Poornima: “Who are your customers?”

David: “We’re a B2B company and mostly service startups. We don’t want to have to go through the enterprise sales cycle. But enterprises also aren’t on our platform because many of them aren’t on Heroku. It’s harder to integrate with their custom platforms.”

Poornima: “Final question for you: what should someone be thinking about if they want to build an infrastructure product and are just getting started?”

David: “Will your product become a commodity?”

Poornima: “Yes, I love that question! I am constantly asking people that.”

David: “You have to figure out how you’ll continue to add value because with things like data stores and PaaS (platform as a service). Most providers move toward providing a better interface, but the underlying technology is pretty much the same. You also have to think about what network effects you can bring to the table. I’d also recommend that for the first 6 months the CEO or founder manage the first line of support. It can feel like a distraction, but the support channel gives you a good sense of direction. You really start to understand and know where you need to take the business.”

Poornima: “That is great advice for our readers. Thank you David, and I wish you and Amit a lot of success in building MemCachier!”

Just to recap, here’s what we learned from David:

1. You want to talk to customers and understand their needs before you start building. David and Amit had a hunch that the existing product on the market, ElastiCache, was insufficient based on their own needs. But before they wrote a line of code, they talked to their friends who were in startups to validate the need for a product like MemCachier.

2. You want to partner with other businesses that care about building a long-term relationship and are easy to work with. MemCachier’s partnership with Heroku has been really beneficial as they got early traction and continued to build the business, and the key ingredient was that both businesses approached each other in a straightforward manner.

3. You want to think about how you add value and avoid becoming a commodity. Infrastructure products become commoditized quickly, so if you’re going down the path of building one, you’ll need to think about how you continue to offer services on top of the technology to meet the needs of your customers.

Get 7 more in-depth interviews with founders & early employees at startups like Pardot, Shopify, and Olark in How to Transform Your Ideas into Software Products, a step-by-step guide to bringing your product ideas to life.

Tweet

Pocket