How Travis CI Is Fixing Company Culture By Taking On ‘Culture’

Photo of author

By Blake Thorne

Travis CI
Travis CI

Here’s a loaded phrase in the startup world: culture fit.

It’s a term with humble early intentions that has grown weeds and sprouted out of its container. It started as a simple way of talking about whether a new hire and current team would work well together. It’s grown into a loaded gun of baggage and misappropriation. It’s used to hire unqualified people and fire great ones.

Mathias Meyer, CEO at Travis CI, started to notice a problem with “culture fit” and the way it was implemented at many companies. It seemed to him like “culture fit” was doing the opposite, and holding company cultures back. Companies, if not careful, would create a monoculture, with everyone acting and thinking the same way. This is terrible for creativity and growth.

Or as Meyer put it in an excellent blog post:

“There’s one fundamental mistake in both using and looking for culture fit as a means for hiring: You’re assuming that your current culture is healthy and doesn’t need to be changed.”

I chatted with Meyer about his thoughts on culture fit, growing Travis CI and what they’re doing to create an authentic company culture.

Can you briefly tell me the story of how Travis CI got started? Where did the initial idea come from and what were the early steps in getting it off he ground?

Travis CI started out as an open source project in 2010. One of our founders scratched an itch to have a continuous integration system that he could run on Heroku and that would run tests for contributions to his open source projects.

Other projects and their maintainers got interested in using it too, so Sven worked on turning it into a multi-tenant platform that could support many open source projects. Josh joined in on the efforts, and we slowly saw more projects adopting it and more programming languages being supported. It started out supporting Ruby, but by now supports two dozen different languages.

Back then, the entire build infrastructure ran on Heroku and one or two sponsored Hetzner servers. The platform was adopted by popular projects like Ruby on Rails and ended up running several hundred builds every day. Word of mouth and our own roots in open source have been very helpful in growing usage.

Mathias Meyer
Mathias Meyer

By the end of 2011, a company was incorporated with the idea of turning Travis CI into a business that would allow us to work on the project full time. Companies approached us wanting to use it for their products and private projects too, so that’s what we focused on in 2012, eventually launching a paid product in September that same year.

Our open source platform, which continues to be free to this day, has been an important driver in terms of growth. The companies initially approaching us have been an important part of getting our product off the ground, but since 2012, our growth has been fairly organic.

A curious bit of our initial growth is that our product remained in private beta until June 2013, a total of about nine months. We were initially afraid of not being able to cope with capacity demands from customer growth that we can’t control, so we restricted the amount of customers that we’d invite, and mostly just invited people who’d email us about getting access.

It wasn’t until late June 2013 that we put up a very basic landing page (I can say with modest pride: “I designed it myself!”) and increased the number of people we’d let sign in every day. We hid the sign in button after ten people clicked it every day. We increased the number as we grew more confident about being able to meet capacity demands.

The big spike in growth, though, came when we finally opened up the product for everyone. We realized that our fears were mostly irrational, and that basic landing page and just letting people try out our product turned out to be the simplest way to grow our customer base. It sounds so obvious in hindsight, but it made an incredible difference on our bottom line.

How quickly has the company grown? How many people now and where are they located?

We’re now three years into running the company with a paid offering. Our initial goal was to be able to pay ourselves, which also put pressure on all of us to focus on growing our revenue. Without any outside funding other than a crowd-funding campaign to help us get off the ground in the beginning, we had to focus on being profitable in a way that allowed us to pay ourselves a minimum salary at least.

We had about a hundred paying customers by the end of 2012, so we could acknowledge that we may be on to something that we can grow further.

There were and still are five founders in the company, and we made our first hires from the open source community, people who contributed to the project or that we met at conferences. By the end of 2014, we had 11 people, with only one team addition in 2014 in total.

In 2015, we ramped up our growth to cover some important areas that were causing us growing pains, specifically customer support, but we also hired more engineers and people supporting the entire team.

In July 2015, we have 23 people on our team, almost evenly split between remote and Berlin-local employees. We’re distributed across Europe and America at the moment, with the majority of our local people being in Berlin, and a big part of our US team on the east coast, but we’ve grown as far as the LA and San Francisco areas.

What specific challenges with scaling up and adding people did you encounter?

One challenge is related to the distributed nature of our team. We’re pretty evenly split between people working at our Berlin office and people working remotely. We pushed hard for a more remote team this year, in part so that our engineering and customer support coverage across the globe fits our customer distribution (> 50% are from the USA).

We’ve been struggling with ways to keep everyone in the loop with the goings-on in other teams and with the company in general. A big part in that is that we now have so many people in the US vs. people in Europe, that time zone coverage has become a problem, in particular when it comes to valuable face time with each other.

We’re still trying to find the best solution there. iDoneThis is one of the tools we’re using to increase visibility across continents, and it’s a great tool for this purpose. But we’re constantly on the look-out for things that help make our remote employees feel more included.

You’ve written about ‘culture fit’ and your problems with the term and how it’s implemented at companies. What made you first start questioning “culture fit?”

Diversity is one of our core values at Travis CI. We try to foster that value on many levels, in part through our efforts supporting RailsGirls Summer of Code and setting up the Travis Foundation with the goal of making open source a better place.

We found that keeping a hiring discussion around culture fit kept us from identifying why we think someone wouldn’t be a good candidate on our team. Talking about culture fit implicitly had us assume that we know exactly what our culture is and it had us believe that our culture is good the way it is.

That in turn hampered the possibility of extending our horizon as a company and as a team. The discussion would move away from whether a person would bring something entirely new to the table which would help us improve our culture.

Talking about culture fit kept us from living one of our core values, diversity. Because when you look at culture fit as a hiring requirement, you may end up keeping people out that might not be the same as you, but who might make your team, your company and your product great.

Related to culture fit, how do you want Travis CI to be different? Is culture still important when hiring?

Culture is still very important when hiring, but we look at it from a different angle. What we’d like to have represented in the discussion with candidates are our core values as a company. They alone don’t define our culture, but they do define what we stand for as a team and what we expect of each other.

Being aligned on values is very important in the hiring discussion. I find that’s what people tend to mean when they talk about “culture fit”, but focusing the discussion around alignment of values is much more explicit in our view.

Diversity is a big differentiator for Travis CI, and we’ll continue to push for fostering a diverse and inclusive work place. As of July 2015, our team is evenly split between men and women.

What specific steps have you taken to address culture fit in a different way?

Our team grew by 12 people so far in 2015, and one thing that we focused on was increasing our gender ratio. Given that we’re advocating diversity so much in public, we found that internally, our own team structure didn’t reflect this value well enough.

With our most recent hires, we explicitly focus on building a team that’s split pretty evenly both between women and men, but also between remote and local people.

Increasing the diversity on our team has forced us to rethink how we hire, how we write our job descriptions in an inclusive way, it’s forced us to think about what we really value and making those things explicit.

Things like work-life balance, taking time off, fostering family life are important to our company, and have been from the very beginning. Making those explicit has helped us build a much more inclusive work place. It also helped us realize that while we do have a culture, we can continuously improve it, but the key to do that is hire more people who aren’t all similar.

Are there specific questions that you ask hires (or yourself) during interviews that reflect your different way of thinking about culture?

During our initial application phase, we only ask two questions. One of them focuses on why they want to work at Travis CI. This question has turned out to be an important differentiator to filter out candidates. Based on our openness regarding our culture, for instance writing about changing our vacation policy, has lead to a lot of applications mentioning that, or that people want to work from home. While those are great reasons to work with us, it also shows that the candidate hasn’t really dove into what we stand for as a company, what values we have, and why we appeal to them.

The other question reflects the importance of valuing a life outside of work. We ask candidates if they have any noteworthy accounts around the internet, or a blog that they run, possibly even on a topic not related to technology or their work. We mention Etsy and Ravelry as examples alongside GitHub to make it clear that we value this kind of diversity.

During our first application process to build out a customer support team, one person mentioned their Ravelry account. While this wasn’t a deciding factor, this made her profile stand out to me, and she’s now leading our customer support team. Subtle differences like that matter a lot to us, and they have helped us to build out a great team.

Any specific books, blogs, other resources on this topic you have found helpful?

My most mind-opening experiences on the topic of culture fit and diversity have come, without a doubt, from Model View Culture.

Just recently I read two great books, from the same author, who are very focused on dynamics in teams, one is “Five Dysfunctions of a Team” and the other is “The Advantage“. Both have been eye-opening for me, and they’ve helped shaped my thinking about culture.

P.S. If you liked this article, you should subscribe to our newsletter. We’ll email you a daily blog post with actionable and unconventional advice on how to work better.


Boost Your Productivity In 5 Minutes

Get daily tactics, insights, and tools to get more done.