Why Start the New Year with Blank Charts? A new year doesn't mean a new system...

3

In my travels, it's more common to see daily, weekly, or monthly performance measures posted on boards in hospital departments.

I like it best when I see time series data displayed as a line chart (run chart) or, better yet, a “process behavior chart” (a form of a Statistical Process Control or Control Chart).

When we get to December, it's not uncommon to see 10 or 11 data points on some monthly metric chart. That's helpful context, even if the chart is drawn as a column chart instead of a line chart. We want to see trends. We should look for outliers in the data. More data points make it easier to do this.

Charts like this always make me wonder why we throw away the additional context of data from 2016 when we start the 2017 chart.

It's unlikely that January 1st in any given year is the magical creation of a new system or a new process. In other words, January 2017 was likely the same process as we had in December 2016. The annual cutoffs are arbitrary. Sure, people are accustomed to thinking in annual cycles, but is this helpful for improvement?

The problem is when February 2018 comes and the 2017 data gets thrown away. We'd have a chart with one data point (January). We can, at best, compare that data point against the target (which might have changed for 2018 even though we have the same system as 2017… and we should expect the same performance without changes being made to the system).

Comparing individual data points against a (usually) arbitrary target is not the best way to manage improvement, as I've written about before.

Here is a chart that Amazon shows me, with weekly book sales from 2017 as a line chart:

Now that 2018 has started, I don't want to see a blank chart with just one or two data points as January starts and progresses.

Amazon does what I would prefer… they add new data points to the chart. I can choose to view the last 52 weeks. As we add a data point, one data point is dropped. I can also view “all historical data” for even more context if I choose (which is what I normally look at).

This chart, with the first data point of 2018 added and the oldest data point removed, would be helpful:

This chart, with just one data point for the new year… not as helpful:

If part of the purpose of having a metric is to ask “are we improving?” then you shouldn't just throw out the previous year's data in a big batch at the start of the year.

Your new year is most likely the continuation of last year's system. I'd suggest your charts and metrics should reflect that.

What do you think?

What does your organization do? What do you think they should do?

Please post a comment and join the discussion. Subscribe to get notified about posts daily or weekly.

Mark Graban is an internationally-recognized consultant, author, and professional speaker who has worked in healthcare, manufacturing, and startups. He is author of the Shingo Award-winning books Lean Hospitals and Healthcare Kaizen, as well as The Executive Guide to Healthcare Kaizen. His most recent book is the anthology Practicing Lean that benefits the Louise H. Batz Patient Safety Foundation, where Mark is a board member. Mark is also a Senior Advisor to the technology company KaiNexus. His latest book has been released as an "in-progress" book, titled Measures of Success.

3 Comments
  1. Mark Graban says

    Comment from LinkedIn:

    Phil Weihl
    Retired Operations Executive…Serving my Community and Advising Select Organizations Pursuing Operational Excellence

    Mark Graban, spot on! If you don’t see historical actual performance, it’s impossible to gauge if a single data point is improving or deteriorating, represents a trend, requires countermeasures to address, or requires actions to reinforce. Can’t answer the question of ‘what changed’ if you can’t see that the latest result is a change from the past.

    1. Mark Graban says

      Another LinkedIn Comment:

      Rodrigo Bernal
      Adjunct Professor en UTEL Universidad

      It drives me crazy when orgs do that. As if January magically resets everything back to zero. A good management system adds more data to look for special variation and common cause. 12 data points “at best” won’t let you know much. Especially if you are using a bar chart.

  2. J. Paul Everett, IE, ret. says

    FULLY agree, Mark. As long as the System itself has not been changed, continuation of data is best. And I would argue that even if the System has been changed, better continue with the data to see if things are better or worse after a suitable time.

    I am reminded about that need (period of time to go by) by the wild weather the world is having and the brutal cold on the East Coast right now and the hottest summer on record out here in Western Washington State. Complexity Science teaches that when there is a disturbance in the System, there will be greater initial oscillation until a new level of System equilibrium is achieved. But, every System has some volatility, just much less when stability has been achieved.

Leave A Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.