Based on my own recent experiences, I'd expect at least one of the below statements to ring true for your organization:

  • Covid accelerated your digital plans for customers but increased complexity and strain in your operations (more channels, new products, different inputs etc.)
  • Hybrid working has made it harder for your operations leaders to sense the pulse in their teams, making accurate performance reporting more important than ever
  • Data and AI/ML projects have struggled with poor data quality, some of which is driven by employee data entry that isn't improving as fast as you'd like
  • New technology solutions that could solve some of this are either too expensive / complicated for your needs, or have made the situation more complicated not less

Assuming that you're still with me, let me set my stall out further and assert that achieving and maintaining accurate operational data is mostly NOT a technology problem. In fairness to the first bullet above, improving customer experience can increase data quality where it eliminates data selection / entry errors.

A well thought out form or journey can do wonders for the accuracy of volumes or productivity metrics by case type or call reason, so investing in UX is usually money well spent. What next though?

The trick is to have management systems (i.e. people making decisions with other people based on performance insights) that trend towards the truth. What I mean is that continuous improvement is built in, being curious about trends and root causes is prompted by meeting agendas and dashboards, so it's just part of the culture.

Some concrete examples of where things often go wrong would be:

  • Having too many KPIs, not being clear on what we're ultimately managing to i.e. the few metrics that truly matter. Sometimes this can happen as strategy changes and new metrics are added but previous ones aren't adjusted or removed;
  • Having a few KPIs but using different ones at different levels in the organisation
  • Not measuring and reporting on the key drivers of those priority KPIs so we can understand why we are where we are;
  • Following on from the previous point, asking colleagues to update too many data fields so they have to choose between data quality and their own productivity (hint: productivity wins)
  • Not using performance reporting to benefit front line workers. This is arguably the most important fail, as if accurate data entry doesn't help those doing it then they won't invest time to get it right. By truly "closing the loop", the benefits for front line workers would be better training / coaching, fairer performance management, a sense of achievement as performance improves etc. etc.

All in all, gathering better operational performance data that doesn't help teams and individuals to work better is pointless.

By being clearer on what metrics matter, enlisting the teams in their design and data capture approach, and enabling effective data-driven review sessions at all levels, an organization can fix its data problem with an overall system designed to continuously trend towards truth. (And yes there might be some data and BI work to support all of that).

Originally Published 15 December 2022

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.