​Automation for the People

​Automation for the People

Artificial Intelligence (AI) is exciting, right? Even setting aside Hollywood's Androids and spaceships, the latest consumer manifestations - virtual assistants like Alexa and Siri, autonomous cars - do make it feel as if the future has finally arrived!

But the truth is that AI is more often far more mundane, and tends to be hidden from plain sight.

AI and customer experience

Customer satisfaction generates real value for utility companies - including regulated monopolies. Consumer trust and confidence means a more receptive customer base for new products and services, and also increases the likelihood of positive regulatory outcomes. For these reasons and more, improving customer satisfaction is high on the agenda for progressive utilities. So should utilities look to AI to help improve the customer experience?

AI is all around

AI is all around us, but most people aren't aware of it. When your cloud-based email service consigns a message to your spam folder, it is using AI to make that decision. Everyday, hundreds of millions of people engage with customer service agents via messenger apps, blissfully unaware that they're conversing with an artificially intelligent computer. The fact that such text conversations are possible without the client realising they're engaging with a machine demonstrates the power of AI to transform the customer service space.

Artificially intelligent customer service messaging apps, known as chatbots, are probably the fastest growing AI application. Facebook Messenger already hosts 34,000 bots, used by companies to engage with their clients. The reasons businesses are so keen to adopt the technology are numerous.

Fast, low-cost recruitment

It doesn't take months to recruit a chatbot, it takes seconds. Designed for use in large organisations with many thousands of customers, chatbots scale easily, meaning that a new resource can be added at the click of a button with very little cost. This is in stark contrast to the cost of hiring a new customer service agent, who requires agency fees, a salary, national insurance payments, office space, equipment and training before they even get started.

Personalised and pre-emptive

AI can learn from your own customer data. As it does so, it's able to improve its own customer service performance, moving beyond answering queries to acting pre-emptively and offering product recommendations based on user profile and usage history.

Integrating AI communications with customer data gathered from utility billing systems can delivery highly personalised services. Providing money-saving recommendations around usage and tariffs, for example, can improve the customer experience and overall levels of customer satisfaction.

Improving staff happiness

It may sound counter-intuitive, but AI is even able to improve the job satisfaction of your existing staff. While potentially AI will replace some jobs, current estimates suggest that the technology is only able to answer up to 35% of customer service enquiries. These are generally the most mundane and repetitive ones - in short, the enquiries that bore and frustrate human staff.

If the bot takes care of these, staff are left with the more complex queries to deal with. Devoting more time to these will improve the outcomes, as well as making their jobs more interesting and increasing their job satisfaction.

Ethics and moral judgements

Alongside the positives AI can bring to customer service, there are also areas where care must be exercised.

In an age of automation, there are concerns about the ethical impact AI will have on relationships. A good sales representative or customer relationship manager uses moral judgement and an understanding of the wider context to help resolve situations. Staff serve as an embodiment of a company's ethos, and if AI is to take on more responsibility of serving customers directly, there is a need to improve the empathy of robots to avoid offending or losing customers.


Accountability of actions will also require addressing, as AI is set to carry out more actions from self-learning and less through human programming. Alongside ensuring customer data is protected, the timing and frequency of communications and actioning tasks without acknowledging changes in the market could be damaging to both companies and customers. In a Business Insider article last year, David Hagenbuch posed the question “if a computer can't be held responsible, who should be?”

Detailed and responsive regulation for when AI goes wrong will be required across many industries to ensure the right parties bear responsibility. For now at least, an understanding of the potential risks will help to avoid an over-reliance on a technology still in it’s infancy.

The changing face of customer service

The future of customer service will almost certainly include greater use of technology, and particularly the adoption of AI into customer-facing roles. There are challenges to overcome, but certainly great potential too.

Currently AI suits larger businesses better, because the data set required to ensure continuous learning by the machine is about 5,000 conversations per month. However, as with all technologies these parameters will decrease over time, making AI more suitable for businesses of all sizes - including yours!

B a c k t o K n o w l e d g e