A Coaching Experience with OKRs

At Redgate, we try to guide our work using Objectives and key results (OKRs).

Our coaching team wanted to practice what they preach, and agreed some OKRs for the first six months of 2018.

Those six months have passed, so we looked back on our experience with OKRs to see what we’d learnt.

A good objective drove our work

Our first objective was to ensure “Coaching is an established part of the product division at Redgate“. This guided a huge amount of what we did in the first six months of the year:

  • Regular catch ups with our tech leads
  • Embedding in groups of teams, building stronger relationships
  • Pairing with dev leads to tackle problems
  • Raising the profile of coaching through internal activities and external speakers
  • Successfully introducing a new quality coach to Redgate
  • Working on cross-cutting initiatives, like use of OKRs within Redgate

We spoke about this objective a lot, both directly and indirectly, as we figured out how to best spend our limited time.

A bad objective didn’t…

We had a second objective; that teams would have plans for how to improve or maintain quality. Turns out this wasn’t useful for us:

  • What teams do isn’t (and shouldn’t be) in our control
  • We don’t think teams need explicit plans for quality; it should be part of what they do
  • This should be the result of established coaching, not a separate objective

We stopped talking about this objective. That helped us focus on establishing coaching now, rather than worrying about a second, less-tractable mission.

Our key results were garbage

With some exceptions, our key results weren’t useful. They didn’t show if we were making an impact, and we didn’t find them engaging.

They fell into a couple of categories:

  • Counting things we thought we would produce
  • Describing tasks we had already committed to

These felt real and meaningful at the time, but they wouldn’t tell us if the things we did had the desired effect or if we needed to change tack.

We did see key results

We felt our work was having an impact, but why? What actual key results were we reacting to?

We came up with plenty of signs, including:

  • People ask if we need more quality coaches
  • Teams approached us to work with them, and kept us up-to-date with their progress
  • Teams sought out a coach for assistance if the one they were working with was away
  • We organically developed better relationships, especially with tech leads and new starters
  • Teams have repeated activities we have introduced them to
  • We could articulate what “good” looks like for someone in a coaching role at Redgate

These are too granular to be good key results, but they give us great insight into where useful key results could come from.

What will we do differently?

Although we’ve learnt a lot, there are only a few big changes we want to make for next time:

  • Only have one objective
  • Reject KRs that don’t focus on an outcome
  • See how we can be creative with data when thinking of KRs
  • Be public with our objectives and progress

We’ll be updating our objectives soon, based on everything we’ve learnt over the last six months.