We’ve been reading Accelerate as an engineering org recently as we focus our reliability - particularly as our usage continues to increase. This has resulted in fascinating conversations as a team as we try to figure out where to direct our attention and what to measure.

The authors identify a series of capabilities that drive four key measures. The measures are useful because they’re predictive¹ of outcomes that organizations care about (financial and non-financial measures). The thing about the measures, however, is that they’re all lagging indicators. Which means they’re informative, but aren’t great places to look if we want to understand what to do today.

One of my favorite illustrations of a lagging indicator is quarterly results according to Jeff Bezos. In his view, the results aren’t lagging only a few weeks or quarters, but years:²

When somebody … congratulates Amazon on a good quarter … I say thank you. But what I’m thinking to myself is … those quarterly results were actually pretty much fully baked about 3 years ago.

Jeff Bezos

The point is not to say that lag measures are not useful. They are. But only in a post-hoc analysis. The bigger point is the importance of understanding whether what you’re measuring is a leading measure or a lagging one. Mistaking one for the other will be problematic. This is as true in a professional context as a personal one.

Footnotes

• ¹ Three of them are predictive. All four are correlated. The authors go to great lengths to make this point.

lag time measures vs lead time measures and how you can use them differently

“a good quarter is the result of work a year ago, 6 months ago, etc.” - bezos

understanding whether a measure is lead vs lag is important for understanding how to interpret it