To get the dashboard project moving again, five simple steps are all you need:

STEP 1: Focus only on your top 2 to 3 priority performance results.

Stop trying to do it all at once, perfectly. It’s expensive, it’s time-consuming, it’s momentum-dampening. You’ll get to success sooner by taking the quick and easy route to set up simple, bare-bones dashboards that do the job with inexpensive applications like KPI Dashboard, Tableau and even Microsoft Excel. Once you’ve finished these 5 steps for these first priorities, focus on the next 2 to 3 most important performance results, and keep building and improving in bursts.

For my own business, I use a Results Map to capture all the performance results that make up my strategy, and it helps me to work out the most important ones, the results that are pivotal to success and need to be measured first and foremost.

STEP 2: Choose a few performance measures for those priority results.

There’s no prize for using all the KPIs you can think of to track a particular performance result. In fact, there’s a penalty and it’s your attention being spread too thin to make any kind of performance improvement at all, or your attention being pulled away from your unique priorities. Think narrow and deep. Go for the truly relevant and insightful performance measures.

If I ever get stuck and can’t figure out what the best measures are for a performance result, the five-step Measure Design template works a treat. It makes you think about what your result really looks like in tangible terms, and that’s the key to finding the right measures.

STEP 3: Define exactly how those priority measures are calculated, and from what data.

Lame measure names won’t cut it. “Customer Loyalty” isn’t a performance measure. “The percentage of customers who have purchased from us more than 3 times in the past 6 months” is a measure. One of the biggest time-wasters in setting up a performance dashboard is trying to figure out exactly how the measures should be measured. That sounds tautological, and it is. If you can’t figure out how to measure something with real data, then it ain’t a measure yet!

Before I build my dashboards, I define each of my performance measures and keep these definitions in a Performance Measure Dictionary.

STEP 4: List the data sources for priority measures, and set up links in your dashboard to that data only.

A lovely man named Siva once had dozens (some say hundreds) of special spreadsheets where very key performance measure data was kept for a freight business. Only he knew how the spreadsheets worked, and where they were saved on the server. And I think he has sinced retired (or won’t be far off). Doesn’t it make you shudder?

My performance dashboard is built with Tableau software, inexpensive and easy, but most importantly for me, able to link directly to my core business database. When my assistant updates my business data each month, my dashboard updates automatically. Love it!!

STEP 5: Systematically create the graphs in the dashboard for each priority measure, and arrange them on a single dashboard.

Forget silly dials and gauges. Forget pie charts and 3-D bar charts. (Even though it’s near impossible to find an example dashboard without these useless gizmos!) The most effective way to monitor your performance measures is with a line chart that tracks changes over time. That’s what you’re managing when you manage performance: moving performance closer and closer to target as time goes by.

I arrange my performance measure line charts onto my dashboard in segments that correspond to my priority performance results. After all, you’re using the measures to know if you’re acheiving the results. That’s one of the 5 principles of designing useful and usable performance reports.

TAKE ACTION:
Whether you choose Tableau, KPI Dashboard, your existing dashboard/scorecard application, or even just Microsoft Excel, use these five steps and bring your most important KPIs to life *this week*. As Dr Poh, one of my readers, recently emailed me, “If not I, then who? If not now, then when?”

IN THIS EPISODE:

Feature: 5 Steps For a Fast Performance Dashboard.
Q&A: Should dashboards be used only for KPIs that can be reported on daily or weekly?
Quick Tip: A review of KPI Dashboard.

Subscribe at iTunes or listen here:

Download mp3

IN THIS EPISODE:

Article: 3 Things You Need On Every KPI Graph.
Q&A: How to get Executives understanding different methods of reporting performance results.
Quick Tip: How to choose the right graph type for your KPI or performance information.

Subscribe at iTunes or listen here:

Download mp3

There are seven important performance signals you’ll want to look for, and be prepared to respond to when you see them:

Signal 1: Performance is unpredictable or chaotic

When performance is unpredictable you’ll see it fluctuating wildly with very large variability from week to week or month to month. This chaotic behaviour is symptomatic of a business process that is out of control because it lacks standardisation in the process steps or the inputs used. Don’t try to improve performance. You need to get it under control first.

Signal 2: Performance is worsening

Depending on your performance measure, worsening performance might be evidenced by an upward trend or shift, as in the case of Expenditure or Rework Hours, or by a downward trend or shift, as in the case of Customer Satisfaction or Profit. If you can pinpoint when the worsening in performance began, you have a better chance of finding out why, and taking successful action to turn the trend around.

Signal 3: Performance is stable and not changing

Performance values will always vary to some extent, and variation does not necessarily mean change. If your performance measure values are varying consistently within the same band or range, and you don’t see any values breaking away from this consistent but random pattern over time, the measure is signalling that nothing is changing. Sometimes that’s good, but it’s not good if you expect to see improvement.

Signal 4: Performance is improving, but not fast enough to reach the target

All improvement in performance is good, when it’s planned, but sometimes improvement is not big enough or fast enough to reach the planned targets. If the targets still matter, and your measure is signalling that the improvement rate won’t be fast enough to reach the target in time, you need to intervene.

Signal 5: Performance is improving at a rate fast enough that the target will likely be met

This is a good signal! That your measure is indeed tracking confidently toward its target level of performance is a great sign that your strategies and improvement projects aimed at achieving the target are working. But such signals are not a sure sign, because just as extraneous factors can cause a deterioration performance beyond your control, they can also cause an improvement in performance beyond your control too (just take the economy as an example in both cases). Check before you celebrate.

Signal 6: Performance has reached target

The best signal to get from your performance measures is that the target has been achieved. As with the signal of performance tracking confidently toward its target, this is a great sign that your strategies and improvement projects have worked. But again, such signals are not a sure sign, because of those extraneous factors outside your control. Check before you celebrate.

Signal 7: Performance has exceeded the target

Contrary to what you may think, exceeding a target is not better than meeting a target. It is a potential waste of resources and time that could have been better spent fixing more important performance shortfalls.

The best way to be sure you can see these signals in your KPIs and performance measures is to present them in a time series, using a line chart, with at least 20 historic values, and add your target as a point above the future date it should be achieved. Then you’ll have a clear view of what’s really happening with performance, and you won’t make those rash and wasteful decisions from drawing conclusions based on “this month to last month” comparisons.

TAKING ACTION:
Are your KPIs and performance measures displaying graphically in a way that can highlight these 7 important performance signals? If not, display your measures in a time series line chart, with at least 20 historic values, and add your target as a point above the future date it should be achieved. What insights does this give you about actual performance?

PRACTICE #1: Use two or three related measures, not just one measure to understand a specific result. Staff Turnover is not a sufficient indicator of staff engagement, but combine it with measures like the percentage of staff intending to leave within the next 6 months and staff satisfaction with their worklife, you get a less biased picture.

PRACTICE #2: Don’t tick & flick – consider all your measures as important characters in the same performance story. Don’t you think that optimising on-time running and safety performance is much wiser than trying to maximise each of these, independent of the other?

PRACTICE #3: Consider qualitative indicators as well as quantitative measures to add more life to the story of performance. Just like a series of video case studies or photos of smashed up cars give sensory context to improvements over time indicated by road crash severity metrics.

PRACTICE #4: Look at your data in more ways than one - the average is only one small part of the plot. A fuller understanding of customer satisfaction comes from examining the variation in satisfaction ratings, the change in satisfaction levels over time, the correlation between satisfaction levels and revenue – not just the average satisfaction rating.

PRACTICE #5: Report causal analysis and hypotheses along with the performance results, to give the story somewhere to go. If cycle time is blowing out, then find out the events that likely triggered the shifts in cycle time, or find out on which activities most of the cycle time is being spent.

PRACTICE #6: Ask your data questions to draw out more of the story. Why aren’t costs staying within budget? Which parts of the budget are blowing out most? Where is budget being managed well? What if we cut back spending on this?

When you think about putting your performance measures to use to improve performance, think systemically: do what you can to draw out the story in the performance data. Stories engage people, and when it comes to numbers and statistics and measures, we usually have to do all we can to get people engaged!

TAKING ACTION:
Think of one performance measure for which you have plenty of historical data. Collate that data, and start looking at it using different tools, to uncover its story. Try a time series graph to see if it’s changed over time, a histogram to see its variability, a scatter plot to see how strongly it is associated with other measures.

And do you wonder why people aren’t really getting much value from measuring performance?

Because performance measures are feedback, you need to track them often enough for the feedback to be of any use in guiding your decisions and actions to improve performance.

So here’s a checklist to help you decide whether you should measure a result on a weekly, monthly, quarterly, or annual basis:

__ How SENSITIVE is the performance result to change? If you were to go make an improvement now, how long would it take before you saw the effect? For example, if it takes years to see change happen, like a staff safety culture or community attitudes to water use, you can probably get away with measuring six monthly or annually.

__ How URGENTLY do you want to know if the performance result has shifted? Is this a result that needs to be improved in the next six to 12 months? If so, it probably needs to be measured at least monthly!

__ How COSTLY would it be to measure the performance result more frequently? Is this worth the benefit of having more frequent feedback?
Could you improve performance so much that the cost of measuring more frequently is far outweighed by savings or added value?

__ How ACCURATELY do you need to measure the performance result? By taking smaller samples more frequently, you can reduce the cost of measuring often. But with smaller samples comes less reliability, so take care. If you just need a good indicator of the trend or change over time, smaller samples can work a treat.

YOUR CHALLENGE:
Think about the kind of response or action you designed your measure to inform. Are you measuring it frequently enough to inform you before you act, or only frequently enough to tell you after you act?