Originally published on the Clara Labs Medium-hosted blog (now defunct). Thank you Maran Nelson and Jason Laska for many helpful insights and comments on drafts of this essay.
Just as we closed our Series A financing at Clara Labs, we articulated a set of operating principles. Though there is a lot of helpful material online about principles at a high level, this post is meant to give you exposure into the raw creation process: why we did this project, how we did it, and how you can do it too. Lastly, here is a template document you can use to run this process on your team: Operating Principles Formation Template.
Internships are great places to study how companies instill culture. My favorite story to share about this is from my first internship, at Apple. During orientation, the recruiter spoke to us about the care that Apple took in figuring out the wording for interns’ offer packets. She relayed how her team had debated the cover letter copy for hours, and ultimately chose the opening line of “Ah, paperwork” to reflect Apple’s “necessary evil” attitude toward paperwork. Her message was clear: at Apple, we painstakingly sweat the details.
My experiences at Apple made me believe in the power of reinforcing a well-articulated cultural identity and set of beliefs. However, when in a company’s lifetime is the right time to begin formalizing cultural identity and beliefs? Too early, and you won’t have enough experience as an organization to reflect on. Too late, and you may lose the cultural cohesion cultivated by the early team and founders. At Clara, we decided this inflection point was right as we raised our Series A.
In advance of this organizational growth spike, we wanted to lay down a framework that would enable us to build a great team, fast. In our view, the goal of principles are to help incentivize consistent behaviors (hiring, rewards, discipline, decisions, time investments, interactions with others, etc.) in a scalable way (i.e., not hundreds of ad-hoc interactions). Our hypothesis was that this would enable more autonomy for current team members and empower us to communicate externally who we were, helping us recruit great people.
Before we get into how we designed our principles, let’s articulate some reasons not to take on defining your principles. Doing it poorly can be more harmful than helpful. Your biggest enemy is apathy from the team: if people don’t buy into the principles, you can discredit yourself.
Once we determined we were prepared to define our principles, we needed to first figure out what success would look like. Here’s a list of the most helpful articles we found:
From this research, we determined that successful operating principles would fit the following criteria (henceforth referred to as The Essential Criteria).
In short, they would need to be concisely articulated, already true, differentiating, non-obvious opinions we have conviction for. To achieve this, we outlined and executed the following five-step process:
To develop principles that were credible, authentic, and distinct we needed to start by surveying ourselves, our team, and our time.
Organizational Audit: The easiest way to figure out what you value is to look at what you make time for outside of “doing work tasks.” We wrote down a list of everything we do at the company, from how we organize our space to events we make time for as a team.
Founder Survey: While we didn’t want to design our values by committee, we did still believe team input was crucial to the success of this process. To enable this balance, we first cropped the questions from Molly Graham’s post (copied below) and both the co-founders — Maran and myself — answered them in a document.
Team Survey: We then designed a simple Google Form (see template for the text) and asked the team to fill it out within a week so we could integrate their perspective and check overall alignment between our answers and theirs. Giving folks a clear deadline and communicating that they should spend a nontrivial amount of time on this is important for the feedback to be useful.
With the above three data sources now recorded, we began the synthesis phase. The goal of this phase was to cluster the recurring ideas from the three sources of feedback into potential principles, and it gets messy. We focused first on clustering concepts before prose-writing to ensure we were capturing the key ideas.
Clustering: This involved a lot of copy/pasting from the three inputs bullets in a doc. We focused on breadth-first, as opposed to trying to prematurely decide what felt “right” and “important”. At this point, we had eight common clusters. If you’re interested, we’ve included the raw form in our template.
To keep the principles memorable, we wanted to get the number of clusters down to seven. We also needed to make each cluster more focused, as the above list was a bit chaotic. So I sat down with my co-founder to talk through what felt strong and what felt weak about each cluster, simultaneously testing against The Essential Criteria.
Prose-Writing: Once the ideas were down, language became critical. Developing language for principles is hard because you have to turn an abstract concept into an actionable, memorable phrase that highlights the most important points to the team. We settled on a couple hard and fast rules:
Each principle should have two parts, a short “slogan” (< four words) and a one to two sentence description. The slogan should start with a verb (helps make it feel actionable), and the length helps make it memorable. You can see what our prose ended up looking like here.
Once we had prose for each of the principles, we were ready to share them with the team to get feedback. We had a 14-person team at the time, so we coordinated an offsite that culminated in an initial principles presentation for feedback. This is what our initial presentation to the team looked like.
While we did start out by talking about the goals of the principles, my co-founder and I underestimated the importance of making sure everyone fully understood, at the beginning, what we wanted the principles to accomplish for our team. We ended up needing to clarify halfway through the presentation that these were not intended to inform product decisions, but rather how we think about solving problems.
We then went through each principle and its prose one by one: how we arrived at it, what it meant, and examples of how team members have exhibited them in the day-to-day. Lastly, we presented a slide that had all the principles on one screen, so folks could connect them all together to see how they worked as a whole. We collected feedback from the team one person at a time, listening carefully and taking verbatim notes to review later instead of trying to defend our presented principles.
Now we had two assets: the current working version of the principles and three pages of near-verbatim notes on the team’s feedback. We grouped the notes into clusters of feedback, and then figured out which clusters we wanted to address by changing prose, removing/adding principles, or by better framing the principles. Here are some highlights of feedback from the team. Based on the team’s feedback, we made the following modifications to the principles:
Once we addressed all feedback, we did a final litmus test against The Essential Criteria. They all passed, and Maran and I felt confident in representing and fighting for the principles. We decided we were done, for now. Here was the final product:
It’s been two months since we completed our operating principles exercise, and we’ve already noticed some encouraging impact.
One thing we haven’t done is stick posters with these principles displayed all over our office. We’re wary of getting gratification without actually integrating them into our work.
Overall, the operating principles process has driven a lot of clarity for the team at Clara. If this is something you believe you could benefit from, we hope this post serves as a helpful reference, in addition to the Operating Principles Formation Template. If you end up using this template or a variant of this process, we’d love to get your feedback.