Stanley Milgram conducted a famous experiment in 1963, the year before I was born. No, he wasn’t a Frankenstein de nos jours, and I wasn’t his creation, whatever you’ve heard from my school chums.

His experiment measured the appetite and capacity of study participants, men from a diverse range of occupations with varying levels of education, to obey an authority figure who instructed them to perform acts conflicting with their personal conscience – specifically, to inflict pain on other people.

Migram

The experiment found, unexpectedly, that a very high proportion of men would fully obey the instructions, albeit reluctantly. It also established that the further removed the teacher was from his students, the more likely he was to continue inflicting pain.

In the increasingly online world, providers of products and services are increasingly being removed from customers. In retail, finance, even government, we have gone from selling face-to-face or in person to distributing through third parties – and are increasingly driving the customer to online self-service for information, advice and support. We are removing human assisted customer services and implementing logic based automated triage of customer problems and with the advent of next generation predictive analytics, we are removing the human element.

Machine learning techniques are not only improving, but the barriers to and price of employing predictive analytics are lowering. This increase in the use of unsupervised machine learning will move us further from a paradigm where human analysts can review data through the personal touch.

The inherent risks are easy to see and could be underestimated by managers more focused on obeying orders to adopt the latest in AI adorned Emperor’s new clothes or drive profitability, EBITDA and service levels, oblivious to the increased distance from, and impact on, customers.

So what can we do to minimise the risk of creating a distance so great, that we inadvertently inflict pain on customers we never in good conscience intended?

The first, golden rule must be – listen to your customers!

Now we’re not Luddites and advocating a return to 100% human-based customer contact management. Many customers are happy to self-serve – to a point. Many analytics technologies produce great insight at rapid pace and low cost, driving genuine service improvements as well as efficiencies.

What we are advocating is to hedge the rapid advancement in automated contact management and analytics with strong mechanisms to capture customer sentiment and feedback at every touch point in every journey.

And to complement these transactional customer insights with what we call relationship surveys, designed to establish customer satisfaction and recommendation/repurchase intention based on overall experience with your brand, not just a specific interaction or journey.

And ensure you ask customers to voice feedback in their own words, not just as a set of scores – verbatim feedback which can be converted to text and analysed in real-time to identify key trends, gripes, etc. thanks to increasingly powerful text analytics.

But remember to have a skilled human, ideally customer management practitioner with real-life experience of running or supporting contact centre and customer services operations, review the machine-driven analysis and conduct their own deep dives.

What else can we do to minimise customer distance and pain in the pursuit of following orders? Here are a few ideas:

  • Solicit and bring to meetings the customer stories generated by feedback mechanisms
  • Put yourself in their position and ask yourself if you’re happy with how you’re being treated – take a good, personal look at your customer journey maps, systems and processes to see their impact
  • Co-create – can you design your processes with your customers?
  • Make sure the machine does what it says on the tin – we can easily be led into making bad customer decisions by adopting every feature of new tech, so ensure people are in control of the design, implementation and review of what it does
  • Where you do deploy people to handle customer touchpoints, ask their opinion about whether processes and systems are delivering value or pain to customers

We can’t complain about customer disloyalty if we continue to distance ourselves from them any more than if we put up prices or change ingredients without justification – or increasingly, without their consultation. If a brand shows it doesn’t want a proper reciprocal relationship with customers, why should they feel anything good about the brand?