Why Operational Innovation is the missing ingredient to enable Generative AI


Share on LinkedIn


One of the key attributes of the digital age is that change programmes can be informed by data, using Generative AI based analytics to uncover the insights. The use of a test and learn agile approach allows a multi-disciplinary implementation team – operating as a squad that ignores traditional organisational boundaries – to achieve the following:

  • apply an iterative, task based approach – with measurable success criteria – to drive momentum and show business value quickly
  • encourage small focused groups to deliver incremental change
  • use metrics and reporting as a governance mechanism to allow changes to be tracked and validated so that processes and solutions can be adjusted based on observing the ‘real world’ impact

Continuous Improvement in contact centres is not a new idea, but when you have analysis of 100% of your phone and digital messaging contacts ‘at your fingertips’, then it speeds up the pace and agility of delivering change initiated results. That becomes the difference that makes the difference.

Concept of strategic agility

At its heart, strategic agility is the ability to gain business advantage through change. There are many examples of how this can be achieved that contact centre operations can learn from. For example, marginal gains theory is about improving and optimising your performance by a small amount across a number of different areas that leads to much more significant, noticeable improvements overall. If you look to improve everything you do by up to 1%, focusing on the smallest of details, then the cumulative gain adds up to a much more substantial outcome. There are proven examples in fields as disparate as elite team sports performance, wealth management investing and sustainability.

How do you create operational innovation?

Within a contact centre operation, strategic agility is all about creating a framework involving:

  • Test and learn – for example the ability to make small adaptations based on customer behaviour insight
  • Focus on creating simplicity in your processes – for example streamlining communications that cause unnecessary contact
  • Empowering teams through collaboration – creating cross functional squads who form to work on a specific and relatable challenge or problem – for example to identify the root cause issues where the symptom is seen in the contact centre but the remedy lies elsewhere in the organisation
  • Delivering small changes – micro improvements – as these lower the risk profile and reduce the uncertainty around achieving a positive outcome
  • Sharing the results – both failures as well as successes – this level of transparency enables other stakeholders to apply these experiences to their own experiments
  • Introduce a culture shift that supports the principle of ‘progress over perfection’ *

And from a generative AI perspective, having real time data is a game changer in terms of tracking these course corrections as it is the speed of receiving feedback that is key.

* Social science research suggests that having 15% experiment failures is the ‘optimal’ level of being agile – source https://news.arizona.edu/story/learning-optimized-when-we-fail-15-time – where your hypotheses only need to be right 85% of the time.

A structured approach to change

The following five step methodology is the ideal way to leverage the power of Generative AI and apply it in a structured approach to change.

Image credit: Multichannel Customer Experience Ltd

Start by select the improvement areas to investigate – think of there being a hopper of potential improvement ideas to pick the best candidates from – and use the insight from analytics to delve into the interactions themselves to understand what makes the good calls/chats good and the poor calls/chats poor.

If you have implemented an analytics platform with sentiment analysis and intent identification then this will speed up this learning process alongside any manual call listening that the change team do to augment their understanding.

This generates a set of hypotheses. The most actionable improvements are often those that are:

  • call handling structure related – that could be improved by changing the question and answer process inside the call
  • agent behaviour related – that could be improved by coaching agents in their communication skills
  • root causes of customer frustration – that could be product improvement or business process related for example

In the same way that agile techniques have revolutionised digital product development, so operational change can be implemented through a sprint based approach. This is particularly true for automation opportunities where the change team liaise with product management and technical teams to scope the impact of implementing the change. Regular iteration cycles engage users and make them more involved in delivering a positive outcome.

Once you have decided each improvement to make – and how it will be actioned – the next step is to separate that change from other variables. This means applying it to a target group that can immediately put it into practice. 

Don’t forget to observe the impact compared to a control group. Does the challenger beat the norm, and if it does then are there any unintended consequences. For example, the resolution rate might go up – which is a good thing – but what happens to the downstream repeat contact rate?

And is there a material impact on handle time that impacts capacity? If these stay constant, then the hypothesis is proven and the improvement can be rolled out from the target group across the rest of the contact centre teams, remembering of course to continue to monitor that the results are consistently delivered across the wider cohort.

The role of the central change team is to skills transfer this Continuous Improvement approach across operational teams so that these processes become self-sustaining and embedded across the contact centre.

Why now?

In the context of a cost of living crisis, a post pandemic talent shortage and with the risk of economic downturn, the time has never been better to harness the power of Generative AI. It meets the challenge of a practical business reality where contact centre operations need to deliver ‘more with less’ – higher revenues, with less capacity, at optimal compliance. 

Just like elite sports teams using marginal gains, contact centres that demonstrate operational agility will become the competitive front runners who break free from the chasing pack.

Paul Weald
Having worked for some 30 years in the contact centre industry, I have built up a vast array of experience across all aspects of people, process, technology, operations and Customer Experience. The latest innovations - such as customer service crowdsourcing and digital money saving apps - are now making it so much easier to provide new methods of customer engagement using digital communities.