Skip to content
Notifications
Clear all

Hot take: Mandatory training completion rates are a vanity metric, not a success metric.

2 Posts
2 Users
0 Reactions
0 Views
(@cost_observer_42)
Estimable Member
Joined: 1 month ago
Posts: 122
Topic starter   [#6884]

We've all been there. The quarterly FinOps newsletter goes out, proudly announcing that "95% of engineers completed the new cost dashboard training!" Leadership nods, the tool vendor gets a bonus, and everyone pats themselves on the back.

Let me be the skeptic in the room: So what?

A completion certificate in your LMS is about as meaningful as a reserved instance purchase without proper coverage tracking. It tells you nothing about behavior change, which is the only metric that impacts the bill.

I've seen teams with 100% mandatory training completion still generate six-figure waste because the training was a box-ticking exercise. They clicked through the slides, passed the trivial quiz asking "What color is the cost alert button?", and never logged into the tool again. Meanwhile, their unattended dev clusters run all weekend. 🎪

The real success metrics for rolling out a cost tool are buried in the billing data, not the learning management system. Did average instance utilization improve by Q3? Did the rate of idle resource alerts decrease month-over-month? Has the percentage of workloads covered by Savings Plans increased since go-live?

If you're measuring rollout success by training completion, you're measuring the cost of the input, not the ROI of the outcome. It's a vanity metric that makes the rollout *look* good while obscuring whether it's actually *doing* any good.

I want to hear from teams who skipped the mandatory training fanfare. What did you track instead to prove the tool was actually adopted and driving savings? Show me the billing data.


cost_observer_42


   
Quote
(@datadog_dave)
Reputable Member
Joined: 2 months ago
Posts: 157
 

Totally feel you on this. I've been guilty of celebrating those high completion rates myself. 🎉

But you're spot on - the real proof is in the cloud bill. I've started tracking what I call "training to tool usage" by piping LMS completion events into Datadog, then correlating them with actual login events and dashboard views for the cost tool. The drop-off is... sobering. Sometimes it's like 90% completion but only 30% ever even open the dashboard after.

The quiz example is painfully real. Better to have a "lab" where they have to actually find a waste alert in the tool and tag it, or something.


Dashboards or it didn't happen.


   
ReplyQuote