Improve Performance – Use Limits Not Targets

(Prefer to watch on video than read?  Then just click here – 5 mins, with captions.  Previous videos here BTW).

It’s very standard to set some targets when performance is in need.  Often smoke-screened by calling them KPIs (forgetting what the ‘I’ stands for).  Maybe even ‘aggressive targets’.

Yet…they don’t always get the hoped-for result, for a simple reason – targets aren’t how things work in the real world!

The Conventional Way

Here’s how we’d typically do it.  Take this graph, with Performance on the vertical and Time on the horizontal, with the horizontatal dotted target line.  And let’s assume higher is better.

On those time periods where the cross is above the target, we say “yep, that’s us”. When it’s below the target line, we carefully explain why circumstances out of our control caused that.  And for those receiving the information, they’re either thinking “why isn’t performance always like that”, or “those circumstances shouldn’t matter”.

Dissatisfaction zone for everyone.  Because we are ignoring that there is natural variation in every system.

A Better Way

A healthier approach is to acknowledge reality and use upper and lower limits.  Nothing fancy, just means our graph looks like this:

How do we set these upper and lower limits?  There are fancy statistical ways (Six Sigma anyone?), but the key is that you don’t set them on what you hope for, you set them on what is actually going on.  (If you want to get a bit more analytical about it, choose how many out of, say, 100, you’re comfortable with exceeding the limit, and draw the lines at the spot where only that many are outside it.)

A graph like this then allows us to put in a connecting line, which helps our minds to remember that all work can be viewed as a system that naturally varies…

Which means we only need to focus on those times when the variation goes beyond the limits.  And if we set our limits well, these will be genuine issues requiring resolution.  You could even get the whole team together to discuss, thus giving people ownership of their own work.  Who would have thought?

As for individual performance…this is more determined by the system overall than the individual, assuming competence.  Imagine we tracked your own arrival time at work, assuming you left the house at the same time every day.  The biggest determinant of when you arrive is not your skill as a driver…it’s how the traffic system functions.  Route options, weather, annual events, daily things like traffic light sequences…all make the biggest difference.  Individual skill definitely plays a part…but not as much as we think.

For situations where people aren’t able to get themselves within the lower limit…then we look more closely.  This is the equivalent to someone having genuine difficulty in negotiating their car and traffic.  Maybe they need glasses?  Maybe they need driving training?  All normal stuff, the key being that we’re looking to bring people into the natural variation of the system.

And not label them failures.

What About Improvement Then?

But remember, the point of measuring is to first make sure we’re keeping our promises to those we serve and who fund us…and second to get better.   Here’s how….

Take a typical performance graph over time, below, again assuming higher is better.

The natural instinct is to want to get that line higher (higher is better on this one remember), but there’s something we need to do first, which is to get the thing more under control.  Or, in fancier terms, reduce variation.  Why?  Because it’s easier to improve things if they don’t vary as much

So what we might do is examine the higher and lower points, asking ourselves what happened there, and ultimately using both analysis and intuition to arrive at ideas which can help reduce the extremes. 

When successful, it looks like this…

…where you can see the line going green and the limits changing when the variation was reduced. 

You’ll notice that the best ones aren’t so great anymore, but remember, what customers and funders are after more than the occasional brilliant performance is reliability, which in turn builds trust.  The green part of the line is simply more reliable.

Now, not only have we learned about bit more about how the show works through our discussions, we can now turn our attention to lifting the whole thing up.  Once again, a combination of people in a room (especially those who do the actual work) doing analysis and sharing intuition to arrive at ideas, which ultimately leads to….

…a higher green line that is still more reliable.

And remember the real benefit – we created a situation where those involved in the work, plus perhaps those who have insight to add from other areas have been able to come together in a room and work on their own system.

Bringing it Home

It takes more courage than we’d expect to not set a target, and instead create the upper and lower limits that reflect actual system performance.  Much safer to use conventional methods which allow for easy deflection of accountability.  And it also takes a combination of courage and skill to navigate the early conversations and show the group how to create quality dialogue that leads to real improvement.

But you don’t really expect results to improve by simply saying a new number out loud, do you?

I didn’t think so.

😊