“What gets measured, gets managed.”
This well-known quote is often applied to business situations like marketing or product creation.
But the basic truth is applicable to almost any situation: When we really want to focus on something, we pay close attention. We want details that can help us make good decisions.
And at Buffer, we want the details that can help us grow as an inclusive and diverse team.
As we grow, we want to be conscious of how we grow: With a focus on the 10 core values that guide us, and with an awareness that a diverse team across a variety of perspectives not only helps us represent our audience better but also makes us more innovative.
We’re not where we want to be yet in regards to a truly diverse team, but we’ve taken some small steps in this direction, including:
- Modifying our language in job descriptions
- Exploring our own biases
- Reaching out to groups and individuals working on diversity initiatives
- Formalizing a family leave policy and sharing more about how Buffer supports families of all types
As we’ve shared these changes and efforts, one of the major comments we’ve heard in response has been: “Great! So what kind of difference has all of this made?”
What gets measured, gets managed.
We’re grateful for the companies and individuals who’ve created frameworks to share, measure and understand tech and startup diversity data, including initiatives like Tracy Chou’s Women in Software Engineering Stats and DoubleUnion’s Open Diversity Data project.
Today we hope to add our own small contribution to this big effort with the Buffer diversity dashboard.
How the dashboard works
You can explore this self-reported data in a few different ways—view the makeup of the Buffer team and/or our candidates through these three lenses with either bar graphs or pie charts, or check out the raw data and comments as they come in.
We also have filters for each subset to allow for drilling down deeper into any part of the data.
For a little extra context, we’ve also identified the most and least diverse areas for each of these three demographics.
How we’re collecting the data
We began collecting this information on a voluntary basis on May 18. All the applicant information comes from a Wufoo form that those who apply for a role at Buffer are invited (though not required) to share with us after their application is submitted. Here’s what it looks like:
(The open comments field is new as of this week; we’re interested to see how it will be used!)
In order to avoid bias in any particular direction and to comply with the guidelines of the Uthis information is collected totally separately from all applications and doesn’t travel with an individual in any way.
UPDATE 6/25: Thanks to some great feedback, we added a new field to the voluntary spreadsheet that asks users to choose between the options of publishing the data to the dashboard or keeping it private. Both are great options!
We use Zapier to feed the Wufoo information into a Google Spreadsheet that’s published to the web and feeds the “Raw data” section of the dashboard. Then we used the R shiny server project to create the various graphs and comparisons. The code is fully open source, in case any other organizations might want to give it a try. (More info on the technology behind the dashboard is over at the Overflow blog, with a full report from Michael.)
All Buffer team members were invited to take an identical survey in order to compare our team versus our candidate pool. We’ll re-survey the team every quarter to keep the information current.
What comes next
Sharing the dashboard is the very beginning of what we hope to do with this data. We’re only just now getting enough information to begin to make interpretations and discover possible disconnects in our recruitment or hiring process that we might address.
We want to make the data more interactive so it’s easier to interpret and understand change over time. We’d love to add graphs representing the demographic makeup of the workforce of San Francisco (Buffer’s official home, though we work all over the world), the United States, and even the world in order to see how our team makeup compares to these benchmarks, kind of like this graph from GigaOm:
And we want to add new areas of study over time, because we know that the areas of gender, ethnicity and age are only a small part of understanding true, intersectional diversity.
But the main thing we want to do with this dashboard is use it to better understand how we can grow as a diverse and inclusive startup.
Right now, we don’t have a team makeup that feels representative of our community, and this dashboard could be key to figuring out why that is and making changes. Are our job opportunities reaching all kinds of people and communities, including underrepresented groups? Could we change the way we hire to eliminate more bias and create more inclusivity?
We hope the dashboard will be a way to benchmark all our future efforts and discover what, if any, changes we make will have an effect on these numbers. What gets measured, gets managed.
We want your thoughts
Making Buffer more inclusive is an ongoing effort, and members of the Buffer community have been amazingly kind to share their thoughts with us.
I’d love to hear feedback on how the dashboard looks and operates. By open sourcing the code, we hope that you’ll share your ideas or improvements with us. If you make any great changes, send us a pull request so we can add it our dashboard! And if you’ve got any additional resources or ideas, I’m keen to hear all of it in the comments.
The diversity dashboard is thanks to the hard work of lots of members of the Buffer team, especially Michael, who worked on the data, and Julian, who worked on the visualizations. Niel and Kevan offered great guidance.