Background

The ambition for digital engagement is driving business change programs globally.  The goals are to be where the customer is (their behaviour is underpinning this change), reduce OPEX, and enhance customer metrics such as CSat or NPS.  At a technical level these programs focus on 3 main areas:

• Contact deflection with Self-Service
• Conversational AI and Automation
• Assisted Service Effectiveness

The key reason is that customer service is a function of a brands budget and resources.  As this is finite it’s created a gap between customer expectations and the brand’s capability.  We call this ‘The Engagement Capacity Gap’.

This gap is not just a resourcing issue either, it’s about the channels and platforms our customers expect to connect on, and their ‘always on’ expectation of our business. 

It took 20 years to include email as a channel in our Call Centres.  It took another 15 years for significant adoption and integration of web chat.  Then another 5 years for messaging and social to emerge.  And almost immediately it’s been followed by AI driven automation and a huge proliferation of consumer devices, platforms, and channels. 

Due to this rate of change, the plans, technologies, and approaches we put in place, even 2 years ago, are now no longer appropriate if we want to compete for our consumer’s attention. To close the Capacity Gap brands can’t just focus on optimising what they have, they require a new strategy. 

As a warning though, when we look at our digital agenda, we need to consider this set of statements:

• Digital channels do not reduce operational costs
• Digital channels will increase contact volume
• Digital channels are critical for automation, which will reduce OPEX and improve service

What makes these statements true?

Latent Demand

This chart will help describe the situation which is both a problem with assumptions and changing consumer behaviour.

In our nominal contact centre, we have 100k calls a month on various queues with 500 FTE, running on an 8% abandonment rate. Then we added a web chat team of 20 to the mix and brought social channels in with a team of 5.

So far, so typical.  The challenge most operational teams have found though is that the promise of lower OPEX through the use of digital has never been realised, only complexity created. 

There are 3 key areas we need to consider in the above chart

• Web chat will eat into the abandoned contact volume, but because consumers are happier to chat than call (the Lower Barriers column) the total contact volume will increase. No saving
• The size of the web chat team constrains the contact volume. In this example, it isn’t that only 15k wanted to chat, it was that this is the maximum capacity of the chat team. 
• The Social interaction volume is limited because that’s often seen as an overflow for when I can’t chat, and the call queue is too long. For many they go to Social to shout, and constrained staffing prevents proper engagement.

This second chart outlines a possible approach. 

We have introduced 2 key elements to this mix:

• Digital automation
• Channel shift – Voice to Digital

We made a statement earlier that digital does not reduce costs and that was shown in our first chart.  However, it is a critical step toward using Bot or Virtual Assistant capably to resolve your customer journeys.  This automation can reduce assisted contact volumes by 40% for simple and perhaps some transactional customer intents.  It also helps alleviate the latent demand issue for digital and voice, eating into the voice abandonment rate and constrained digital staffing.

Once automation is deployed you then have the option to offer customers who called the chance to engage digitally on their mobile devices (‘Press 1 to Message with us’), and rely on the automation to address those intents at a lower cost.

Approach To Automation

As we approach our automation strategy, we need to take a view on where we start and where we stop. 

Simple informational intents and journeys, triaging contact, signposting, or authenticating customers automatically are typical first stage automation targets.  Even pulling data from transactional systems could be included in this category.

This chart outlines another challenge as we measure value of our automation activity – Not all intents are born equal!

It is typical that simple informational intents (‘Where’s my nearest store?) will result in much shorter calls / chats than complex conversations (‘How do I close my joint account following a bereavement?’).  It is also typical that the volume of these simple intents will be much higher than the truly complex.

This enables us to create ‘buckets’ of activity, starting with the most valuable.

When we build a business case, we need to understand this concept to ensure our calculations are representative of the reality.  We cannot say ‘Automation will reduce contact volume by 30%’.  30% of what?

The business case should reflect not the contact volumes, but the capacity created by the automation (e.g., AHT by intent x volume).  It must also include a requirement for a post engagement survey to ensure we are resolving our customer’s questions rather than ‘containing’ them.  Containment could just be abandonment and the customer has shifted to a call. 

This approach also needs to measure contact by customer and intent across all channels (digital / voice / email / social) to ensure we can accurately measure resolution rates.  This data will assist with optimisation of existing automation and targets for the next phase.

As we move from simple to transactional intent automation, we need to understand the:

• Transactional intents within your customer journeys
• The systems integration requirements and cost for each
• The value of automating them

For example, integration to an insurers Policyholder and payment systems would enable most ‘in term change’ intents to be automated, delivering significant business value and customer satisfaction. 

When we look at automating the more complex and contextual customer journeys, we need to take a view from a human perspective as well as considering the operational benefit.  There’s a cross over between humanity and technology and each brand needs to understand where to draw their line on this.

The most extreme example would be insurance claims processing where there’s a bereavement involved.  The question to ask?

‘Do I want technology to own this conversation when my customer is at their most vulnerable, or should this be something my well-trained advisors should deal with?’

By stepping through a customer journey and business process, we can consider how we would interact with our automation.  Take this example of a complex claims case where the policyholder also wishes to renew a policy.

One of the key benefits of this approach is that your advisors become the escalation point based on the customers’ intent, rather than the default as we are today.  This means they need to have the right information provided to them, in context, and personalised to that conversation.  They need seamless access to the whole conversation (regardless of channel) and the knowledge assets and information shared so far.

To achieve this holistic approach, we need to orchestrate our customer engagements and interactions, across all channels (voice and digital), business teams, and contact categories.  Only then can we deliver a world class experience.