The importance of a good team to build data solutions can’t be underemphasized. If you’ve read anything like Francois Ajenstat’s recent Forbes article or Roger Pen’s e-book on building data science effective teams you get many of the key points; however I would argue that in addition to these points you need to invest time with your Human Resources (HR) team and make them an integral part of the success. Developing their data literacy should be part of your objective to building a successful team.
The following isn’t a prescription for a single conversation, presentation or analysis for your HR team, but a way to develop their data literacy around what constitutes a great data team.
Skill Diversity
Your HR team will focus on inclusion and diversity, but may not understand diverse skills and experiences and how they contribute to creating great dashboards, models, etc. At Juice we’ve found on numerous occaisions that Zach’s experience in digital marketing has opened new insights or ways of designing valuable healthcare analytics solutions. It could be just asking a different question or offering up a solution to a similar problem in a different industry (btw, funnel visualizations in healthcare are amazing). If everyone on the team is from the same industry or has similar experiences how do we get the HR team view this as a concern or red flag? How do we convince them to find someone with complementary skills and background?
The right way to think about skills is less about an individual’s skills, but about the team’s overall skill set. Your series of HR conversations should be an understanding of what the team is good at, where are their blindspots and what skills are needed. Also, HR needs to understand how to ask questions like, “Describe to me some of what you’ve built in Python and how users were impacted” vs. “How many years of Python experience do you have?”
Fit
One of my favorite quotes I’ve read recently on building data teams is “Hire people, not experience.” It comes from this piece on Medium by Murilo Nigris. How do we get to know people? Rather than ask them to talk about the tools they’ve worked with convince them to tell you story. Its in those stories that you’ll understand their values and priorities.
For the HR team to be able to assess fit you need to decide what your team’s value are. Do you value speed, creativity, production quality code or collaboration? All of the above is not a valid answer. Give your HR team 3 to 5 values to screen for. Take the time to explain why these values matter. Include examples of how someone on the team currently exhibits these values and how they makes you successful.
Defining and describing values will sound like a lot of work; however it will be a fraction of the effort of having to let someone go because they weren’t a fit.
Data-Driven Job Descriptions
Many of the job descriptions I read for data positions are painful to read. The biggest miss in my mind is I never really know what this person will be doing exactly on Day 1 or Day 500. Your job descriptions should read more like these on the Salesloft website.
WITHIN ONE MONTH, YOU’LL:
Build a prototype application that will be posted on our website.
Completely data visualization online class to bring your data literacy vocabulary in line with the team.
WITHIN SIX MONTHS, YOU’LL:
Conduct product feedback interviews to gather feedback on existing features, and speak to new features coming.
Successfully lead a scrum team by running planning meetings daily.
A nice benefit is that a job description written like this becomes the individual’s performance plans and goals if they are hired. Here’s a template from the Google offers a way to think about job descriptions as another example. https://hire.google.com/job-description-template/
Skills Assessment
Most new candidates have to go through some technical assessment. Make sure your HR team is involved with the assessment. Don’t let them punt involvement in the skill assessment because its “too technical”. If you can’t explain the skill assessment or if they don’t understand its desired goals then you have a problem. Use the opportunity to explain the assessment as one way to develop their data literacy. They can also see if you have any blindspots in the assessment and to make sure there isn’t bias in your assessment.
Also, make sure that they know skills assessment changes as new technologies are adopted and implemented, so it's never a static test.
Recruiting Talent
Often you are sharing the HR team with other departments. As a result, the amount of time that HR will actively recruit new candidates is limited. In my experience the HR team will send you 3 to 5 candidates or resumes and if you elect not to choose any then you’re completely dependent on whoever finds your website.
When discussing recruiting efforts with your HR team ask the following questions:
What is your time commitment on this opening?
What kinds of efforts will we make to find candidates that are probably already employed?
What can our team do to supplement your efforts? (What are we allowed to do?)
Are there any monetary incentives for us to find our own candidates?
Are OPT candidates a viable option? Do they understand OPT?
https://www.internationalstudentinsurance.com/opt/what-is-opt.php
After your disappointment diminishes, here are some items you can take to supplement their efforts:
Have your team share the job posting link with their social networks
Volunteer to present at local meetups, events, universities and conferences. Try to do at least 2 per position.
This will seem like a lot of work, but building models, visualizations and data solutions without a full team is time consuming too. Note that the lessons above are very applicable to bringing on contractors or consultants to your data team as well.
The initial HR conversations will be hard, but keep the dialogue going even when you don’t have openings.
To learn more about data culture and teams make sure to get your copy of Data Fluency, Empowering Your Organization with Effective Data Communication. If your timeline for your customer facing data project doesn’t include time to get HR on board and fluent, reach out to us to learn how the Juicebox platform can handle some of the challenges with getting the right data team in place.