Logic only leads to conclusions, it's emotion that leads to action

Question: Thanks for your series Glenesk…I run a data analytics team for a motor insurance firm.  We’re stuck in a catch 22 situation where staff don’t fill in data in our core system correctly and we know the data is not always complete and sometimes filled in wrong.  Although the business needs to track performance using the data, it’s understandable that our front-line managers dismiss the results as inaccurate but are also reluctant to spend time improving data quality because it can’t be compared over time in any case.  Have you ever seen this problem before and has it ever been solved do you know?  

We’re glad you enjoyed the series and please be assured that this challenge is much more common than you might think.  The generic argument goes as follows:

There is no point in creating metrics on this topic because the data is wrong and there is no point in correcting the data because it’s not used for anything important.

 The solution lies in effective staff communication and negotiation of the time required to fix the problem fundamentally.  Start by creating the case for why tracking this performance matters; and not just for the firm, but for the team and the individual too so they can know how they are contributing to something that is worth investing time and effort in.

If these conversations have been effective then the groundwork has been set for a negotiation with the business on how best to get the data into an acceptable state.  This is never about fixing all the legacy issues but understanding the trade off in time, effort and accuracy needed.  Often the analytics team can find ways to cleanse the data using a series of logical rules and they can often do much to isolate and allocate specific data fields for front-line teams to address in bulk.  If there is no capacity for this activity then the business may need to acknowledge that the data will only be valid from a given point onwards. 

In order to keep the data accurate in the future, involve staff who input the data in the review of performance that results.  People are more inclined to be diligent in their record keeping when they know the data collected is put to good use.

Question: Hi, I think we’ve followed all the advice in your series but are still struggling with the adoption of new KPI’s within our customer service teams.  We undertook a comprehensive piece of scoping work to work out what was important, linked to job responsibilities and regular departmental performance reviews.  Since implementation though the teams seem to be spending time working around the metrics rather than try and improve them.  As an example we are aware that requests for customer feedback are not being sent to all customers, despite this being company policy.  We are in the process of automating this request but wonder where else staff might be skewing our reporting.  There have also been a limited number of cases in which staff have cited our performance framework as a source of stress contributing to time off work and mentioned in exit interviews.  Do you have any suggestions for addressing these concerns?

Symptoms like these indicate that there is a challenge in how organisational performance is managed rather than how it is measured.

Both the symptoms you describe; manipulating the metric rather than driving the desired performance and reporting high levels of stress, result from staff members being responsible for delivering an important result that isn’t fully in their control to deliver.  There are a range of potential solutions:

  • Look at the way the performance is reviewed; is it in the style of mutual enquiry and problem solving or based on inquisition and blame? The difference in stress level that results is significant
  • What are team members explicit and implicit incentives? If career progression, status and remuneration depend on a limited number of key measures then they are potential sources of stress and targets for manipulation and it may be time to review them
  • Are inputs as well as outputs measured? The context for performance matters.  If consistent volume and quality of work (inputs) result in variable sales (outputs) then there may be a host of market, seasonal and firm specific factors that could be causative and a source of stress for the team
  • Have staff been provided with all the other things they need to get the job done? Are the training, tools, time, effective process, and consistent management support in place to make effective performance possible?