How a PHP Library made Segment a Unicorn

With Calvin French-Owen, Co-founder and CTO

Segment helps businesses connect their customer data to the hundreds of different business tools and applications they use. Today, the Segment team is comprised of over 550 people with its HQ in San Francisco, and offices in Vancouver, New York, Denver, Dublin, London, and Sydney.

Segment helps businesses connect their customer data to the hundreds of different business tools.

Calvin French-Owen is the co-founder and CTO at Segment. While Segment is approaching 10 years as a company, Neel and I chatted with Calvin to understand Segment's origin story and their strategies for focusing on the needs of their customers in the early days.

For those that may not know, what is Segment?

Segment is a customer data platform (CDP). Today, companies are always trying to gather data about all their interactions with their own customers — from their websites, their mobile apps, support tools, etc. so they can understand who they are, what they want and make sure they have a consistent experience. Segment helps them get that all that data cleanly and consistently into a single place, so they can then leverage that data across all their different customer touchpoints — whether that's personalization in their application, sending data to Salesforce, updating tickets in Zendesk, analyzing user data in Mixpanel or Google Analytics, or anything else.

Can you tell us how you got your first hundred users?

We had a bit of an odd journey. When we launched, we made the product free — but we quickly realized that we needed to figure out some way to monetize it.

As we were building out more features, we added them to our pricing plans, but we never actually provided a way to pay. Throughout the first seven or eight months, we were onboarding more and more people to the product for free. By about eight months in, we had maybe 1,000 people or so sending us data every day! At that point we agreed that it was time to button ourselves up and start making money somehow.

The first step we took was sending our users a pricing survey based on the Van Westendorp method. It gives you these four pricing curves that you can look at based on all of your responses. From there, you can figure out where the sweet spot is in terms of pricing.

Based on those survey results, we launched our initial pricing and emailed our users asking them to start paying for our product. Many of them did immediately, because they were getting great value from the service.

Editor's Note: From Wikipedia, the Van Westendorp method asks these four questions:

  • At what price would you consider the product to be so expensive that you would not consider buying it? (Too expensive)

  • At what price would you consider the product to be priced so low that you would feel the quality couldn’t be very good? (Too cheap)

  • At what price would you consider the product starting to get expensive, so that it is not out of the question, but you would have to give some thought to buying it? (Expensive/High Side)

  • At what price would you consider the product to be a bargain—a great buy for the money? (Cheap/Good Value)

Generally speaking on the subject of product-market fit, when people truly like the way you solve a problem (no matter how narrow), they immediately start asking for the next six things that they want. And so I feel that once you establish some of that credibility in the relationship, it's very easy to start discovering more problems over time.

In those early days, what was your strategy for getting feedback from users?

One of the most impactful things we did first was set up live chat on the homepage and on the signup pages.

At the time, we used Olark. People would just drop in and start chatting with us — they gave us feedback and were interested in what we were building. So as a piece of advice, I'd say no matter what, get live chat on our site as soon as possible.

As we started growing, we created an official help desk. At first we used HelpScout, but then switched over to ZenDesk. In the early days, all of the co-founders also did a round robin every day, responding to support issues. This was great because it meant each of us could see where people were running into issues.

The last thing that we did is send an automatic email about 45 minutes after sign-up from Peter, our CEO. It had a simple message: "Hey, I saw you signed up for Segment. Do you have any feedback? Let me know!" It was surprising to us, but a lot of folks were immediately interested and responded to our emails.

Generally speaking on the subject of product-market fit, when people truly like the way you solve a problem (no matter how narrow), they immediately start asking for the next six things that they want. And so I feel that once you establish some of that credibility in the relationship, it's very easy to start discovering more problems over time.

Are you actually making something people want?

Organize and quantify customer feedback to make better products.

What was your strategy for deciding what problem to tackle next?

In the early days, like any startup, our strategy was pretty slapdash. People would submit requests for integrations or libraries, and we'd just tally them up.

A good example was the PHP library. At the time, we were super "trendy" developers who used Node. We thought no one uses PHP anymore, so there'd be no point in building a PHP library.

We were wrong. It turned out, because so many people have built their apps with PHP and used to it, it ended up being one of our most popular requests. So of course we built it!

About a year in, we got more disciplined about how we were going to build our product. We already had a decent amount of sources and destinations, but we wanted to do more. We started asking ourselves: what are the things that we can build for our more strategic customers?

For example — to figure out that out, our CEO took a trip to New York, where he met with many of our big customers. In all of our conversations, we noticed nearly everyone mentioned they were using Redshift or a data warehouse of some kind for their data. We realized at that point there was an opportunity to build that pipeline for them, so they didn't have to do it themselves manually.

Fast forwarding to today, how is this process different?

At the end of each quarter, we'll compile a list of feedback from our go-to-market teams — what we're seeing in the field about competition, the sales process, etc. Typically, there'll be a handful of representatives from each of the different sales teams: someone from enterprise, someone from commercial, someone who is a SDR, etc. Those people will all give feedback to someone on our marketing team, who then puts it all together and does a stack rank of what our customer want. From there, the list gets circulated to the product org and individual PMs to figure out what changes or features to add to our product.

Calvin, thank you for talking with Userstand. It was great to learn a lot more about Segment and how your team met the needs of customers in the early days.