As you are aware, I try to take a mix of heart driven and data driven approach to working at SEER.
So I decided to look at our Holiday Policy with a little data after years of looking at it with heart.
In brief our holiday policy lasts from Dec 23rd - Jan 3rd (usually).It is a don’t come INTO the office, stay home with your family, but be available for your clients and co-workers if they need you, policy. If you wanna be “off-off” then just use your vacation days which are unlimited anyway.
That could be 1 hour a day, 0, or 10 depending on if a fire comes up or not. As a big fan of Speed of Trust - we don’t monitor people’s time, we just believe that they’ll do what’s right when it comes to helping teammates and clients. So this year many people tacked on a day or two and were out of the office working a heavily reduced schedule for 2 weeks. Perfect, it gives people time.
For some reason I wanted to see if I could get a pulse on how many emails were sent during the holiday as a one barometer of are we giving a “false” perk, but people are working all the time? So I asked our IT manager if Gmail gives reports on how many emails were sent, upon which I then found that we could go all the way back to July 9th to look at emails sent by day.
I found some very interesting data, by simply converting the dates in the CVS export from MIke. I was able to get this data set . A little pivot table magic, and voila!
First, my assumptions:
- July-January SEER had an average of 90 employees.
- Sent emails is a better gauge of what you have to respond to vs what you receive
- This obviously would miss things like working on projects / hours and updates in basecamp, but we work so much in email that this is a decent barometer.
- Let me say again, a lot of our work can be outside of email. So I am using email more as a relative scale (how much does it drop) over the typical daily workflow.
Warning: I don’t even know if we can do this at the individual level, but DONT DO IT! This is about a barometer on the company, if you start looking at individuals and how many emails they sent, you have LOW TRUST in your team and you’ll have a lot bigger problems.
First thing I noticed:
Holy shit people’s emails sent drop by 95.81% from the peak (wednesday) to saturday at an average of .68 emails sent per employee, and 1.168 on Sunday.
Average emails sent: Saturday = 61.2
Average emails sent : Sunday = 104.4
What inflates this number is the 6 person leadership team. Anecdotally I know we work more on weekends, because I get emails from them on weekends more than anyone. To check that I used conspire to look at my personal weekend emails as you can see on Saturday Dec 20th, I sent 34 emails
The whole darn company only sent 58!
So on Saturday Dec 20th, 20 emails were sent by 89 people (of which most of the 20 were probably leadership team). That makes me pretty happy.
Since that was the holiday break I picked another random date October 18th. I sent 13 emails, the company sent 66. Again knowing that the exec team is probably 50%, not bad. Especially on Sundays when Crystal does her updates, she’s been known to put out 15-20 emails starting on Sunday 9:00pm EST, as she gets ready for the week :)
I also decided to look at hours, even though I don’t like using hours b/c its too dirty, as getting 100 people to track their hours the same way is impossible, like I didn’t clock any hours on Oct 18th, but obviously I worked, you saw the email sends above (I’m, bad I know), so that means this number is under reporting in terms of hours, but it was a check against the issue of e-mail only work vs reporting, strategy work.
Saturday was 98.09% less hours than the peak of Tuesday
Sunday was 95.91% less hours than the peak of Tuesday
How can you use the data to check your company trends, you can also look at days of the week benchmark this against yourself.
Hours Trend
Hours Trend
(note, someone let their timer go ape shit and forgot to stop it!)
Individual day peak email sent (holiday week) = 470
Individual day peak email sent (random selected week) = 1388
66% drop in emails during the holiday break, I like that!
Individual day peak hours work (holiday week) = 266
Individual day peak hours work (random selected week) = 775
66% drop in hours during the holiday break, I like that!
There are so many gray areas in perks, that maybe a check against them from time to time is one way to make sure you are staying true.
The next thing for us to do is interview our clients and see if we did a good job covering their bases and getting things done when they needed us.
At a time when company after company is releasing its diversity data
http://www.google.com/diversity/at-google.html
http://blogs.cisco.com/news/cisco-shares-workforce-diversity-data
http://www.microsoft.com/global/en-us/diversity/RenderingAssets/Microsoft_EEO-1_Report_2014.pdf
Maybe we could start looking at other important metrics too, vacation days used, sick days, emails sent, you name it.
I also realize that sending email is not the only way to measure but again its a barometer and a start, its why I also looked at hours. Maybe we can evaluate every quarter @ SEER.
The goal is not to compare yourselves to anyone, the goal is to look at the numbers and as a leader ask yourself, is that what I’d like? Am I happy with it? Then based on that stay the course or change direction.
If you find other ways to look at this data let me know, I’d love to use it to help make SEER just slightly better tomorrow than it was today.
Thank you!
–Wil