Engineering measurement must not be done in a vacuum

Looking at engineering performance in isolation doesn't give us a sense of proportion to make sound decisions based on data.

You might think that 15k miles per hour sounds like a lot, but getting to low earth orbit requires 17k mph.

It's important to know the context of what we're measuring. Using metrics like DORA without understanding the business context can cause big problems.

Yes, maximizing deployment rate can impact change fail rate, but if your customers expect new functionality regularly, prioritizing availability could lead to customer churn.

At the same time, if you ramp up your delivery speed in a business with relatively stable revenue, your costs will start to grow out of proportion to business needs.

Another mistake is not to include subjective measures in this balance. If you tune the quantitative engineering and business measures but have low employee satisfaction, your other measures won't stay on track.

It's equally important not to lose sight of why we are measuring ourselves and keep our goals balanced against engineering and business measures.