Customer service automation in the Public Sector

Customer service automation technology is commonplace in every sector, from artificial intelligence (AI) engines used to front-end a customer interaction and problem diagnosis, to ‘chatbots’ and  robotic process automation (RPA) of transactions. 

The efficiency benefits are clear enough, but there are differences between the public and the private sectors that makes the adoption of these new technologies a little different.

The key differences lie in complexity and risk. Most public sector organisations are a composite network of relational services and the risks are often far greater due to the sensitivity of data and support for vulnerable people.


‘Bots vs People’ – the balance for public services

It is tempting to view these new automation technologies and the new digital operating model for customer service they enable as being primarily about efficiency (or not getting left behind in a rush to ‘go digital’).

There is no doubt that many of these customer interfaces, such as ‘chatbots’ can drive efficiency and productivity – eliminating the need for people to answer calls or emails in contact centres. But this is the wrong motivation, because it is all from the point of view of the service provider, not the service user.

Whilst there may be savings, unless the new automation service models are designed from the point of view of the customer/citizen/client, the results are likely to be that the technology is frustrating for the user, and ultimately damaging to trust, confidence and customer satisfaction.

In the private sector, customers will simply go elsewhere if they are unhappy. Users of public services may not have this choice. But it is not that difficult to design public services around citizen need, and then to implement these new technologies. Putting people not technology first in the adoption of automated and integrated customer service built on the latest technology is the only way to longer-term efficiency.


Service Automation – where to begin?

Start by analysing existing customer-facing processes and the top customer activity,  breaking down common service journeys into components where possible. Many of these ‘building blocks’ will be relatively simple functions or ‘microservices’, which link together to form recurring and repeatable transactions. 

Then prioritise the simple areas rather than the more complex ‘relational’ services that require sophisticated integration. The more complex areas in public services may well offer greater return on investment, but require experience of building an omnichannel and using tools that can allow true integration of channels -email, phone, etc.

The next step is to develop your AI capability. Aim to start in areas with a definable customer interface that is fairly stable. For example, by all means use an AI engine, such as Alexa to build a voice activated virtual assistant to help citizens seeking social service (say), pointing them to suitable self-help and online resources. But take great care before allowing the artificial engine to diagnose and deliver complex service unaided, such as care packages and ensure that there is a join up with your customer relationship management systems.

Over time, machine learning and AI capability will be able to tackle the most complex needs in the public sector, as we are seeing in health services, but today it is still essential to have the ‘safety valve’ of a professional human agent, especially to verify appropriate intervention in complex areas.


Automated Customer Service the right way 

There are a few simple rules to follow in order to use these new technologies effectively:  

  1. The customer must be the primary motivation, not efficiency or productivity, with a focus on the user need, preference and journey
  2. Customer channels should be designed coherently, so that there is a commonality in their ‘look and feel’ for the citizen
  3. Simplify to create efficiency
  4. Review your customer relationship strategy
  5. Data is key
  6. There must be ‘safety valves’. This means there must always be the ability for human intervention

These points may seem obvious, but it is easy to be driven by the technology potential rather than the customer service need. 


Building Empathy

Although humans are not always good at empathy, machines are in general still not incapable of adapting to human emotion. Designing an avatar with friendly-looking characteristics, perhaps with a multicultural appearance or ‘demographic sympathy’, is not the answer. 

Having systems that focus on the customer expectation is key: Could the reason for contact be resolved in one visit? Was it quick, easy and intuitive? Did the service response use known customer data, such as from previous interactions? 

By integrating channels of communication and engagement, using microservice componentry, public services can create a true omnichannel structure, designed around the citizens need and preference – this, with human intervention when necessary, create empathy. This does not only require trust in the service provider, but also trust in digital services:


In this way not only are services automated where possible, but where this is not completely possible, the intelligence supporting a CRM can inform a professional agent getting involved, where risks or complexity require it. 


Key Messages

Some public service organisations feel they are getting left behind – they sense they need ‘live chat’ and a ‘chatbot’ avatar on their website, and they know that these tools could be more efficient and productive. 

Successful customer service design requires public bodies such as councils to re-adjust current working practices if the technology is to fulfil its potential and citizens are to feel comfortable and trusting of the new styles of interaction. 

There is no magic short cut to an automated, self-service customer experience. Bots cannot replace people, but they can, with the right planning, design and ethics, enhance our ability to deliver exceptional customer service in 2021

Then these new methods can significantly improve public services, especially for vulnerable groups and the complex interactions that local councils often encounter. And it is this improvement that will drive efficiency and productivity. Far from disenfranchising, these new tools can help public services to reach those who need them most, quickly and effectively and ,ultimately, more efficiently.


Freshstart banner


10 Action Tips for Mastering Customer Service Automation


1. Do not introduce technology piecemeal.

It is important to plan for an integrated set of customer “channels” and tools, so that the customer experience is consistent. This requires a meld of digital and customer strategies.

2. Design around the customer, not the technology.

This is hard to do and depends on culture and board leadership, blending IT, service priorities and risk control. Set some basic principles and get the customer involved.

3. Review your customer service strategy.

The new tools and ways of working cannot be “bolted onto” and existing customer service and strategic ambitions.

4. Learn and train.

The new technologies and methods such as chatbots, AI needs to be understood, to avoid risk “blind spots” or reliance on expensive external resource.

5. Start with the easier areas.

It may be tempting to start with the service areas yielding the biggest ROI, but it is better to start with the areas with lowest risk, complexity and scale, to learn and grow.

6. Track and trace the risks, before, during and after introducing change.

New customer service models which have complex and powerful technologies at their heart, such as AI, bring new risks, especially for public services.

7. Have a customer service performance framework.

This will help to track the value and benefits of new technology and methods, against a measured starting position for customer satisfaction and fullfillment success.

8. Focus on data: have a data strategy.

It is not enough to have control of customer data. There is a need to ensure careful data classification, handling, sharing, integration and insight methods to avoid bias or mistakes.

9. Do not ignore the need for human intervention.

Things will go wrong, and complex problems will usually still need a human to help – people still make a great “safety valve” and complement automotive customer service technologies.

10. Select technology partners who share your ethos.

There are many IT suppliers and solutions out there, some traditional, some radically new. Not only does the IT need to be flexible and intuitive, but suppliers need to understand the wider public service challenges.