Lightning Talk: 12 Tricks to Good-looking Flow Metrics
Fun, thought-provoking, emotionally resonating talks presented by members of the DevOps community.
Hosted by Topo Pal and Jason Cox.
Presented by Sleuth
Chapters
Full transcript
The complete talk, organized by section.
Naomi Lurie
My name is Naomi, and I am here to teach you 12 tricks to good-looking flow metrics.
The goal is to teach you ways to make your metrics look good, so you look good. The source of my inspiration for this is Sarah Cooper's "10 Tricks to Appear Smart in Meetings," which she created based on her experiences working at Google, and it includes classics like "Draw a Venn diagram" or "Translate percentages into fractions."
Since then, she's created hundreds of these. One of my favorites is number 17: turn nouns into verbs.
So you can see where this is going, right? Let's get started.
Number one: design workflows with only active states. Eliminating wait states is a pro move. You'll be able to assure your customers that someone is working on it, and it's guaranteed to make you look like a top performer with 100% flow efficiency.
Number two: work exclusively on putting out fires. Your leaders will be so dazzled by your super-fast flow times, they won't notice it's only for fixes.
Number three: show 20 metrics at a time. When your dashboards look like an aircraft cockpit, people will think you have everything under control.
Number four: measure teams of three people max. The smaller the team, the better the flow. Lop off the time it took to get the work onto your board. That's not on you.
Number five: choose a new methodology before understanding what you're trying to fix. Lean heavily on your leaders' FOMO, and don't be shy about name-dropping Amazon if you think that's what it takes.
Number six: present important-sounding data. Insights are for newbies who can't interpret the data themselves. Your leaders will appreciate you letting them figure out the "so what" on their own.
Number seven: track flow to activities instead of outcomes. Don't mention it takes you nine months to patch a vulnerability. Instead, keep the focus on proxy metrics and bask in the glory of an OKR met.
Number eight: create separate metrics for tech and the business. Don't bother techsplaining to the folks on the top floor. Spin the bad news for them using phrases like "operational efficiencies."
Number nine: obfuscate problem areas by using averages. It's always a good call to talk about averages when dealing with data sets from very different value streams. You'll look smarter by keeping the focus away from the problem areas.
Number 10: ignore warning signs until you've measured a full year. If you're too busy to solve a glaring problem now, encourage a longer data-gathering period to buy yourself some time until it becomes someone else's problem.
Number 11: be prepared to cast doubt in case you're challenged. Say "data hygiene" a lot. This fancy euphemism gets you out of the hot seat and instead makes the entire department accountable.
Number 12: if you don't like your metrics, reorg.
Okay, this is a great opportunity to start from a clean slate. After all, you can't be held accountable for the flow under the old structure, right?
So, you know, guys, we've seen it all, right? But we're serious about using flow metrics the right way to improve flow and business outcomes. If you want to come and share your tricks with me, I'd love to hear them. I'm making a new calendar. Thank you.