Using OKRs to accelerate Relentless Improvement

Introduction

In my last 2 articles, I first introduced the concept of Objectives and Key Results (OKRs) with an illustration based on personal goal-setting then explored applying them to improvement of Product Strategy.

After experimenting on myself with personal OKRs, my first applications of the concept in a coaching setting have been in the area of improvement objectives.  I was particularly excited about this idea as I encounter so many teams and ARTs who lose heart with their retrospectives and Inspect and Adapt sessions due to failure to follow through on their great ideas.

I’ve lost count of how many discussions I’ve had in either the training room or a coaching setting with people who drag themselves to retrospectives feeling they are a waste of time.   The common refrain is that “we just keep talking about the same issues and nothing ever changes”.  A little probing generally exposes the following:

• Teams use the same format every time (my pet hate is the “What went well, what didn’t go well, what puzzles me” retro).
• Teams spend 80-90% of our session examining the past
• The (rushed) improvement idea generation yields concepts beyond their power to change or very fuzzy
• Few teams retrospect on the effectiveness or application of the ideas they do generate 

I’ve gifted many a copy of Esther Derby and Diana Larsen’s classic “Agile Retrospectives” to ScrumMasters and RTE’s.  Likewise, I’ve facilitated countless sessions on time-boxing your retrospective or I&A to leave at least 50% of it for the “improvement” part of the discussion.

Good facilitation will at least get you to the point of emerging from the session with some good ideas.  However, in the words of Michael Josephson “Ideas without action are like beautifully gift-wrapped boxes with nothing inside”.
 

Crystallizing your great ideas with OKRs

A couple of years ago, I started experimenting with the Improvement Kata to assist teams with actioning their ideas.  As I researched it, I came across a great article describing an “Improvement Theme Canvas” by Crisp’s Jimmy Janlen.  The canvas was beautifully simple, taking teams on a great journey with 4 questions:

• Where are we now?
• Where do we dream of being (What does awesome look like?)
• Where can we get to next? (And how would we measure it?)
• What steps would start us moving?

I used it again and again, in settings ranging from team retrospectives to executive workshops.  And it helped.     But it was still missing something.  There was a recurring trend – a lack of measures.
As I started experimenting with OKRs, there was an obvious synergy.   I had found the Key Result setting to be the moment where I did my deepest thinking myself, so I tweaked the canvas a little:

My first opportunity to apply it was the second day of a management workshop.  The first day had been spent exploring challenge areas and selecting focus areas for improvement.  We had a fairly clear view of current state and definition of awesome (which represented 18 months+ of hard work), and the job on Day 2 was to address the right side of the canvas with a 3 month view. 

We began with some objectives, then the hard work began – how would we measure them?  The result was incredibly powerful.  We moved from some very fuzzy, amorphous objectives into real clarity on how to generate momentum.  When you start to talk about how to measure movement, something changes in the conversation.  On the one hand, it grounds you in reality but on the other the fact that you are looking for a measure ambitious enough to only have a 50% chance of success stretches you out again. 

As an illustration, I’ve developed a sample based on a commonly occurring theme in Inspect and Adapt sessions – struggles to develop momentum with Test Automation.

As a final hint on the canvas generation, the order in which you complete it is crucial.  There’s a great temptation to rush to “Enabling Activities” because people love to begin with solutions.  

• Start with “Current state”: fishbone diagrams, 5 whys, and causal loop diagrams are all great tools for generating the data and shared understanding
• Move to “Definition of Awesome”.  The word “Awesome” is incredibly powerful in helping move minds into the art of the possible, and also encourages alignment around desired direction.
• Step 3 is to agree on your timebox (often 3 months or one Program Increment) and figure out how far you can move from your current state to the definition of awesome as you set your “Target State OKRs”.  You will start flushing assumptions, and more importantly discovering alignment challenges.  The movement to quantified measures with your Key Results will expose many an unvoiced insight.
• Only now do you set the “Enabling activities”.  They will come very quickly.  If you’ve set your OKRs well, your key activities will be very obvious.

So far we only have an Idea, what about Action? 

I can recommend nothing so strongly as the weekly priority-setting cycle from Radical Focus I described in my first OKR article.  The reality is, despite many schemes people use for creating space for follow-up on improvement ideas it is amazing how often they stumble.  Life gets busy, delivery pressure is high, and the first thing that gets sacrificed is time to work on improvement ideas. 

The beauty of setting yourself a few priorities each week on moving towards your OKRs and spending a little time reflecting on them at the end of the week is that it helps you “find time”.  A few months into applying personal OKRs I can attest to this.  I live a very hectic life.  Most weeks involve 5 long days facilitating and coaching with clients, 2-3 flights, and lots of “after hours admin” whilst coming up for air to be a dad, husband and human 😊  Its very easy to find excuses.  But every week starts with “how can I move forward this week?”  And in the back of my head every day are my weekly goals.  To be brutally honest, my learning cycle at the end of the week often focuses on “how to find time”.   I’m not waiting 3 months to look back and find out I made no progress and lose heart.   

My final hint is this: Make them public.  If you’re applying the technique to Inspect and Adapt,  your improvement OKRs should appear in your PI Objectives and progress discussed  in your System Demos and PI Demo.  Include the Key Result evaluation in the metric portion of your Inspect and Adapt.  There is both power and motivation in transparency about improvement objectives.