A practical guide to blending bots, the crowd and live agents in your contact centre of the future


Share on LinkedIn

There is much hype around the notion that robots will be taking over contact centre jobs within the next five years. But for those of us who have been in the industry since the 1990s then there is a more measured response. We expect the digital channel to overtake voice based interactions within this time period, but the requirement for an emotional connection between a customer and the organisation will remain. Plotting a path that blends live support – either provided by contact centre agents or by crowdsourced brand advocates – with cognitive based tools will become a key requirement in any future roadmap, taking these employee and customer groups with you on that business change journey.

Impact on agents
At a recent eGain Digital and AI day conference , Gartner put up a slide stating that “by 2020 Artificial Intelligence will disrupt the jobs of 1,000,000 phone-based customer support agents”. That is quite some claim about both the pace and reach of AI disruption. Given that Contact Babel in 2016 have sized the US market at 3.5M agents and the UK market at 750k agents, and if we add a couple of million agents for offshore based staff (these stats are less reliable to source), then we are talking about circa 15% of all agent positions being affected. So we should definitely start taking notice.

But a more important question to raise is what form will this disruption take? If we think about customer processes across two dimensions – the complexity of the enquiry from simple to complex, and secondly whether the enquiry is information based i.e. could be answered from a knowledge base or is customer specific i.e. needs access to customer data system, then we can start to understand where different solutions could be applied:


Image credit: Multichannel Customer Experience Ltd

These scenarios could encompass:
Intelligent assistants – which exploit a knowledge base to self-serve simple information enquiries and to automate simple system based enquiries. Both these scenarios do not require an agent to get involved in the resolution of the enquiry but the configuration of any bot based solutions will require tuning and iteration to ensure optimal performance i.e. a better than 90% resolution rate. This may require agent supervision and support to achieve that outcome.
Crowdsourced agents – are a new concept based on the service model originally pioneered by the mobile phone operator giffgaff where they use existing customer experts to answer questions posed by other customers. This is now being exploited by UK start-up Limitless who are pioneering crowd based service . Key to the success of such initiatives are being able to moderate the quality of answers provided by the crowd sourced ambassadors and to ensure the timeliness of responses. Where both criteria are met then the customer who answered the question is rewarded financially which is a great way to recognise your supportive brand advocates.
Web chat – is the natural way to respond to digital customer requests through a two-way dialogue whilst the customer is online. The focus is about resolving the issue that the customer faces in completing their online journey and there is much case study evidence that demonstrates that web chat drives up sales conversion rates and service resolution rates as well as being highly valued by customers.
Home working – as digital support models mature then the traditional approach of contact centre staff sourcing could change – for example exploiting home based staff who can be rostered ‘on demand’ for those specific periods of the day when contact volumes spike.

It is also possible to assume that these use cases are more likely to be applied initially to digital enquiries compared to traditional voice. Whilst speech based interactions can be automated, a lot of the focus of AI vendors at present is on text based learning suggesting that messaging and email communications are a more likely starting point for a business Proof of Concept AI trial.

Digital first
Many organisations are now designing their business models as ‘digital first’. That means the primary method of customer contact is through a smart phone app, and online self-service. Falling outside of those channels by using the ‘offline’ processes of phone or written communication by email or post is deemed a failure. But just offering customers a digital alternative doesn’t necessarily change their behaviour – this takes time and education. It’s about seeing the solution. Many people don’t associate with facts and figures but they do get solutions when they are described visually. In the digital world then there is the opportunity to augment reality in this way to help customers to see different solutions to their current issues.

The Dimension Data 2016 Global Contact Centre benchmarking report describes the period since 2010 as being characterised as “telephone primed CX” evolving in the latter part of this decade towards “digital primed CX”. This recognises that the proportion of digital related contact will overtake voice over this period with “digital interactions currently accounting for 42% of all interactions”. However there are some barriers to these projections “with one on three survey respondents saying that digital tech is not meeting business needs”. Applying these implications to bot design then there is currently a gap of nineteen percentage points between the “desired level of self-service (37%) and the actual level of customer adoption (18%)” that these new sources of tech solution will need to fulfil.

Business change
Why is this so?

The inescapable fact is that both customers and employees don’t take easily to change. If we were to describe a formula for implementing change then it would have resistance to change as a strong force pulling in one direction with the opposing forces combining:
• Dissatisfaction with the status quo
• Positive vision of a potential future state
• With an achievable first step
The logic being that the higher the resentment of the status quo then the more likely we are to take a step towards an alternative that we perceive could lead to a better outcome.

When communicating business change, then it is really important to be clear on the ‘why’. Simon Sinek in his seminal TED talk stated this should always be the first question to ask. Although the driver for change may be external e.g. realigning the business cost base in response to austerity, or to the uncertainty of Brexit, or to new low cost digital competitors taking market share, the tactics the organisation uses to bring about change “must be aligned to the values of its staff as well as the beliefs of its customers”.

Only once the why is clear, can the hard work start in agreeing ‘how’ the change will take place. As far as customer strategy is concerned then there are the traditional questions to answer – should I insource or outsource my operation, how do I best support digital interactions through self-service and as a result, how many agent seats will I need in the future? But asking traditional questions limits the answers, so more radical alternatives should also be considered. For example should I look outside the organisation to the crowd to answer digital enquiries and/or could I support and manage an agent population who work from home?

So if you want to bring about fundamental change in customer behaviour then your customer contact strategy requires three things:
1. Why – A clear articulation of why the change is relevant – the business problem you are trying to solve – and why this matters for your staff and your customers
2. How – Defining how the change will impact your business operations – considering both the traditional questions around contact management as well as the radical alternatives
3. Visualising – seeing the solution through a digital first lens – understanding how customers need to be educated and supported so that they choose to adopt the new chatbot services and messaging apps

According to the behavioural science research led by Curt Coffman , it is only engaged employees that are the builders. “They use their talents, develop productive relationships, and multiply their effectiveness through those relationships. They perform at consistently high levels. They drive innovation and move their organisation forward.” This is the reason why new initiatives such as crowdsourcing for customer service should have a role to play as these brand ambassadors will also exhibit these engaged behaviours. The emotional connection of an advocate who really cares helping a person is really powerful for the brand, which simply is not there in a transactional robotic or self serve exchange. Businesses looking to differentiate their brand through service need to create such emotional exchanges – via crowdsourced advocates and engaged employees – to improve financial metrics like retention, revenue and profitability.

Likewise inside your contact centres, you need to find those individuals who see bots as the opportunity to take away mundane enquiries so that they can apply their empathetic skills to help customers resolve those issues too complex for self-service alone to achieve. To drive up completion rates of the early implementations, bots will need training and contact centre agents are best positioned to oversee this activity.

Only if you take both your front line staff and your advocate customers with you on this business change journey will contact centre leaders achieve the outcome that the AI futurists are predicting.


Please enter your comment!
Please enter your name here