UK Power Networks, an organization that has to make sure electricity gets through to 20 million UK customers, says that its use of cloud-based ‘contact-center-as-a-service’ has boosted both customer outreach and satisfaction ratings.
Its Head of Customer Contact, Alex Williams, also says that the recent use of AI has already shown huge potential for reassuring vulnerable users at crisis times like power cuts.
UK Power Networks is a major British energy distributor. The UK energy market is split into two kinds of operators: energy suppliers and energy distributors.
An energy supplier is the brand that consumers and businesses pay their bill to. Buyers can choose which one of these to buy and swap between as much as they like.
An energy distributor, on the other hand, owns and operates the physical equipment that constitutes the electricity network in any given area - i.e., the power lines, underground cabling and local substations that transmit the power. A proportion of each bill goes to the energy distributor serving a specific area.
In its case, UK Power Networks maintains the electricity networks for London, the Southeast and East of England. That means it is responsible for nearly 190,000 kilometers of network, both overhead and underground, infrastructure carrying nearly 30% of all the electricity distributed in Great Britain. In its last full financial year, it recorded turnover of £1.8bn.
Williams explains his role and responsibility as:
I'm in charge of all of the ways our customers can do business with us - so all of the different communication channels, from social media to the website to email or them calling us, the journey will start with me.
Back in 2013 and 2014, he says, he and his customer contact team started to see several problems with outages that were starting to affect perception of the brand.
That timeframe saw significant spikes in calls due to weather events, which included severe storms and floods.
It looked like these severe weather events were becoming more and more common, and indeed have been since.
The problem was that customers were having trouble getting through to find out when their light and heat would come back on.
Call volumes were hitting very high spikes, but we didn't really have a platform then that could cope.
Specifically, the firm was using an older, on-premise call routing system which lacked self-service facilities for callers and , because again, it was on-premise and used a very manual configuration.
Another limitation that Williams could see - even in 2014 - was that there was no support for any kind of automation for dealing with routine queries, to free up agents to deal with more complex customer power queries.
I didn’t see that as a ‘robots take over the world’ moment, but we've got lots of older customers who we really want to support as much as humanly possible during a power cut, which can be a very difficult time.
The team decided to go to market to find a way to start to automate what he calls “some of the low hanging fruit” to allow agents to be available for anyone trying to get through during high impact events, like future weather-related brownouts.
Williams says he found the answer he was looking for in the shape of cloud-based call routing software called ‘storm’ from supplier Content Guru.
This promised very high availability to ensure customer requests and issues get quickly and accurately resolved.
This, says Williams, was at first perceived by the company as a much needed telephony solution - allowing the contact center to scale up from a normal day’s 1500 calls to the peaks of more like 100,000, when there’s a break in service delivery.
The platform first and foremost gave us capacity, but since then we’ve started to build in more and more useful features.
These include the first elements of that missing self-service via a new IVR (Interactive Voice Response) feature which allows callers to say their postcodes so that the firm can immediately map that location detail to what it knows about possible power cuts.
Useful information can also be played to callers while they're waiting to reassure and inform them. He said:
Very often, that makes them very happy as people's primary concern in power outages is to see if we know their power is off and then if we can tell them when it’s coming back on.
Now, we're able to provide all that information for them in the IVR—and if they choose to go through to an agent, which they always get the choice to do, the software passes on all their information to the advisor.
That’s important, he says, as customers often get very irritated when they are re-asked about information during a crisis that they have already punched into the phone.
AI for smarter routing
All this is very useful, he says, but even better is the system’s recent ability to use AI in the form of natural language processing to quickly and better route calls that need sensitive handling.
A prime concern here is callers who rely on medical devices at home that have potentially stopped working due to power loss.
If you’ve used key words like ‘sleep apnea’ or ‘oxygen machine’ in the IVR, you get immediately flagged up as a priority and moved out of the normal queue, regardless of what option you had been pressing and you need to get through to the advisor quicker.
And this isn’t about me using AI to employ 15 fewer people or anything like that. This is me using the technology to make sure the right customers get through to the right people.
The software is also now embedded into UK Power Network’s overall network monitoring system, which alerts the company in under four seconds if there’s a power outage and which can even inform them as to how many current users may be affected, plus who they are.
That means that as soon as you ring in, we'll be able to tell you we know there is a power cut in your street or nearby because we’ve done all the integrations to make sure everyone’s talking to each other at our end.
Finally, use of IVR and AI has also reduced the class of emergency calls humans need to deal with by 30%.
Burden of customer satisfaction proof
Williams says all these benefits are very useful - but that the most important metric is how all this helps customers.
And here, he has some hard numbers to show impact. Prior to implementing contact-center-as-a-service, the firm was seeing customer satisfaction scores under 7.9 - but at the end of 2022, the figure had skyrocketed to 93.4.
These figures come from the UK Customer Satisfaction Index, which the Institute of Customer Satisfaction publishes every six months, and which rates several major brands - and which shows that a rather obscure energy distributor is now consistently at the top of these rankings and was even first in the July 2022 questionnaire.
Next steps for the company include expanding use of AI across more of the firm’s services and telephone lines.
Summing up, for Williams:
To me, this is a story about using good technology to change company culture and achieve great customer service.
What I mean by that is we now will call you, and by the same agent you called first, if we have any news - including the bad sort, which people appreciate so much more than no power news at all.”