Better Results through Smarter Analysis

Damage prevention gets more and more demanding every year as facility owners, regulators, and other stakeholders demand better and better performance. An organization either makes decisions by facts, numbers, and analyses of those facts and numbers or it will lose in the marketplace to competitors that do.

In fifteen years as an ASQ Certified Six Sigma Black Belt in senior executive roles at various leading damage  prevention organizations, I have seen many scenarios where erroneous conclusions were reached because of improper data analysis.

This case study of a simplified scenario is a good representation of many real world scenarios I have encountered. This scenario will make some simplifying assumptions to help keep this article within length guidelines. Real world scenarios will be more complex, which means doing the type of analysis that follows will be even more critical.

1) The Utility in question owns gas and electric facilities and both are present on every One Call notification  dispatched.
2) The Utility is split into two operating areas: A and B.
3) There are only two root causes for locating failure: Print Reading Failure and Locating Technique Failure.

The hypothetical case study is that the leader of a locating organization – contract locator or in-house locator – is concerned about locating quality and wants to figure out what, if anything, can be done to improve results.

The leader may take the initial step of looking at how many total damages there were in a given time period (e.g., last twelve months) in Area A and B (Figure 1). The conclusion from Figure 1 may be that Area A has poorer quality than
Area B because it had more damages.

Fig 1
Fig 2







The leader may then look at damages by utility type (Figure 2). The conclusion from Figure 2 may be that the number of damages to Gas and Electric are similar – they differ by only 16% – so there is no one type of utility locating that can be targeted more for improvement.

The leader may then look at root cause (Figure 3). The conclusion from Figure 3 may be that Print Reading is the
root cause of 750 more damages than Locating Technique, or 77% more, and an opportunity could be extra training for the entire organization on print reading.

All three types of analyses and conclusions are almost exact examples of analyses and conclusions I have seen from very good operations managers and executives in many  organizations – even organizations that are leaders in their respective industries. And all three are wrong.

Digging deeper into the dataset, the leader may create a graph (Figure 4) which shows number of notifications by area. What we see here is that Area A has twice the number of  notifications as Area B.

Digging deeper still into the dataset and using notification data from Figure 4, the leader may create a graph (Figure 5) which shows more details by root cause for Area A by likeli-hood (damages per 1,000 notifications).

Area A has a substantial quality issue with Gas locates and almost all of the root cause of the damages lies in Print Reading. The Damages per 1,000 for Area A for Gas of 2.20 is also more than twice as high as for Electric at 1.00, which contradicts the overall conclusion from Figure 2 that the Damages per 1,000 for Gas and Electric are comparable.

If the leader had not used notification data and dug deeper into Area A’s frequency metrics, we would have missed out on an opportunity to address the Gas locating deficiency.

The same graph for Area B (Figure 6) reveals that the  damages per 1,000 for Area B (2.20) is actually worse than for Area A as shown in Figure 4 (1.60), which contradicts the conclusion from Figure 1, which did not take into account notification volume.

If we had not used notification data and dug deeper into Area B’s relevant metrics, we may have used the conclusion from Figure 3 – need more print reading training all around – and missed out on the real root cause of a disproportionate number of damages in Area B: Locating Technique.

In addition, Figure 6 shows that Area B has a serious issue with Electric locates, which have more than twice (3.0 vs. 1.4) the frequency of damages as for Gas locates. An opportunity to improve the locating of Electric facilities in Area B would have been completely missed without Figures 4 thru 6.

This case study illustrates how critical it is to not only collect data but collect relevant data and then not only analyze data but analyze relevant data to get at actual root causes. The errors simulated in this case are not rare at all but are fairly common inside the largest stakeholders in the damage prevention industry.

Jemmie Wang is a consultant specializing in helping all stakeholders in the damage prevention industry achieve
significantly better results – whether it’s growing the top line, decreasing expenses, or improving quality or other
results. You can reach Jemmie at

Share on facebook
Share on twitter
Share on linkedin