Who knew that one could develop warm feelings for a German Federal Employment Agency chatbot? If you own a business and wish to apply for state funds to supplement your employees’ reduced salaries, then UDO will fill in the application form for you. “Let’s go!” the digital assistant declares, launching into a series of questions. The system displays reassuring expertise; the queries—about the size of your workforce, the extent of the reduction in working hours, and so on—are simple, clear, and sensitive to previous responses, and the interface offers soothing blue tones and rounded edges. UDO goes on to ask why the workers are on reduced hours: for economic reasons, such as the cancellation of a large order due to the coronavirus, or because of an unavoidable event, such as a measure to mitigate the spread of the pandemic? And by now, a powerful and comforting thought may well arise in the citizen’s mind: UDO really cares.
In this article, we argue that smart use of automation can enable governments to provide outstanding levels of customer experience, driven by innovations that are as sensitive to people as they are to technology. We begin by considering the challenges and rewards of enhancing customer experience for governments. Then we discuss the benefits to governments of using automation to improve customer experience. Finally, we turn from why to how, identifying three key practices common to successful automation initiatives in public services.
Smart use of automation can enable governments to provide outstanding levels of customer experience, driven by innovations that are as sensitive to people as they are to technology.
Public services customer experience: Challenges and opportunities
Government leaders face major challenges as they work to improve their customer services. They must compete for talent despite budgetary constraints. Their data might be stored in isolated silos, even though citizens now expect an integrated offering. And they are expected to maintain resilience amid the growing complexity of citizens’ needs, including aging populations that put increasing strain on health and social services.
Given challenges like this, it is no surprise that there is space to improve the customer experience provided by government. Across a range of countries, McKinsey has found that public services users rate them on average at 5.5 to six out of a maximum ten points. By contrast, the industries that best satisfy their clients enjoy an average score of about eight out of ten. These include supermarkets in Canada, Mexico, and the United Kingdom; e-commerce sites in France and Germany; credit-card providers in the United States; and pharmacies and supermarkets in Australia (Exhibit 1).1
Yet public-sector leaders globally are recognizing that outstanding customer experience has become an imperative. It is driven by the need for governments to maintain the trust of their citizens—trust that has been eroding in many countries. It is also motivated by citizens’ everyday expectations, which are shaped by the offerings of leading firms. And it is impelled by the intricate and urgent needs of entire populations in crisis. Take a recent demand on the German Federal Employment Agency. In the two-month period of March and April 2020, in the midst of the COVID-19 crisis, the agency processed short-term allowances for employees of more than 788,000 firms—over 385 times more than the same period in 2019. Given the severity of the crisis, it was crucial to provide services not only at scale but also of excellent quality.
By strengthening their services to provide outstanding customer experience, governments can generate several benefits (Exhibit 2). The main one is increased trust: on average, across a range of nations, satisfied citizens are nine times as likely to trust governments, and nine times more likely to believe governments are achieving their mission. This is a stark advantage in turbulent times. In addition, satisfied citizens are far less inclined to return for follow-up appointments, having received the service they require the first time, so a well-run system processes fewer unnecessary visits. By contrast, dissatisfied citizens are not only more likely to return with their unresolved problems but also to complain publicly or take legal action. Finally, quick and effective assistance, free of needless bureaucratic processes, can boost the morale of government employees, which in turn reinforces the quality of service they provide.
How automation can help to transform customer experience
Leading organizations across industries are already turning to automation to improve customer experience. For instance, more than nine out of ten companies with world-leading brand recognition and high levels of customer satisfaction use artificial-intelligence (AI) solutions to raise customer satisfaction, compared to the average of four out of ten companies.2
How can successful automation improve the public services customer experience in particular? We have identified three kinds of benefits. Perhaps the most obvious of these advantages is that of reliability and simplicity. It is crucial for public services to minimize inaccuracy and other mistakes, given the fundamental impact of these services on citizens’ lives. Well-designed automated services are unimpeded by human frailties, like the fatigue or distraction that causes people to enter data incorrectly or make calculation errors. It is also important for services to avoid needlessly extensive efforts on the part of customers, as citizens have limited time and energy to engage with government. By offering lucidly devised engagement processes, automation can support simple service delivery.
The Australian Tax Office (ATO) offers an example of automation-driven reliability and simplicity that has helped to build outstanding customer experience. ATO is a leading agency with respect to orienting around customer satisfaction. In a qualitative survey, citizens strongly approved of the ATO’s highly automized tax-return process, and rated it among the best Australian federal government services on CSAT score. 3
Before automation, tax forms were complex and arduous; many Australians had to work through a tax agent, rather than tackle the process themselves. By contrast, the automated system features many benefits. Data provided to the ATO are used to prefill annual tax returns, so citizens won’t have to locate and reenter information available elsewhere. Errors are reduced by means of “nearest neighbor” techniques: figures entered by citizens are automatically checked against those of people in similar circumstances, and users are prompted to review anomalous entries. The ATO’s tax returns can now take just a few minutes to complete.
The ATO also applies automation to customer feedback. Call-center volumes are automatically monitored, allowing for rational allocation of resources based on informed forecasts of future traffic. Inbound calls are transcribed using speech-recognition software and then automatically reviewed for emerging trends, allowing the ATO to identify potential problems in its services, and also to provide call-center staff with the latest information to deal with trending customer queries, ensuring that citizens enjoy reliable service, receiving adequate and accurate information.
The ATO had already automated its tax returns before the COVID-19 crisis. But the pandemic has given further impetus to public services globally. Given the global surge in welfare applications unleashed by the COVID-19 crisis, unemployment-related services provide one highly topical domain of opportunities to enhance service using automation. The German Federal Employment Agency, an innovative body with a policy of continuously expanding its online offerings, has in recent months eased the way for citizens to register themselves as unemployed. Until the crisis, registration required in-person visits to the agency. But because of pressure on the agency from a wave of new applicants, desire to limit physical encounters, and lower availability of civil servants to staff offices during the crisis, the agency has instituted an innovative “Selfie-Ident” program. Applicants may download an app that guides them to record a video of themselves, and to make images of their identity document. This allows citizens to register with the agency as unemployed without an in-person visit. Here customer experience has been enhanced precisely by not requiring citizens to make an arduous and inconvenient in-person visit, but rather to enjoy a remote application as a far-simpler convenience.
A second way in which automation can boost customer experience in government services is by enabling civil servants to offer more complex and caring service provision. Because citizens’ expectations are rising, and populations are aging, public services increasingly require skilled people to perform tasks that call on a broad knowledge of processes and an empathic attitude to citizens. Given budgetary constraints and the difficulty of expanding the workforce, automation can be of great help by relieving employees from repetitive tasks and allowing them to support nonautomatable areas. Civil servants can personally help citizens navigate through intricate processes and systems effectively and efficiently, resolve complex cases, or make decisions on difficult applications. In addition, they can take the time to connect with citizens seeking emotional engagement, such as an unemployed person hoping for a conversation that includes not only formal service but also some encouragement. In short, automation may enable governments to deploy greater numbers of client-facing staff who can provide high-value, dedicated service to citizens who need it.
More extensive attention to citizens requiring special attention may improve not only customer experience but also the experience of civil servants themselves. This, in turn, can strengthen government workers’ commitment to serving members of the public.
Indeed, if governments encourage their employees to take on roles that tap into the full range of human capacities, the result could be a significant uptick in motivation. The chance to help citizens directly, using one’s institutional and procedural knowledge to improve their lives, is what inspires talented people to join public service. Bob McDonald, former secretary of the US Department of Veterans Affairs, makes the point powerfully: “When it comes to attracting talent and fostering passion in your workforce, I think that ‘mission’ is the one thing the public sector has as an advantage over the private sector.”
There is considerable potential for a transformation in the roles of civil servants. In the Australian Public Service, for instance, about 40 percent of employee time is dedicated to collecting or processing data.4 Automation can absorb many of these tasks, such as linking customer information to internal or external databases, allowing civil servants to migrate to more valuable roles in public service. Of course, such changes in role will require reskilling for many civil servants—and automation could be a spur for governments to think through and actively shape the future of work in their organizations.
A third enhancement to customer experience that automation of government services can offer is personalized service delivery, including AI-enabled service. Here we have in mind two striking features of automation: the ability to simulate aspects of human engagement, and the capacity to perform tasks that are beyond the ability of any human, enabling a service to be customized to individuals. UDO, the German Federal Employment Agency chatbot, displays the former feature, being designed in such a way that it offers a calm, friendly, and reassuring experience. AI offers the latter feature, providing pattern recognition in data sets that are too large, complex, and dynamic for humans to detect an underlying order. Firms like Amazon and Netflix leverage this capacity to make recommendations for individual customers, based on patterns in their databases of customer interactions.
Similarly, governments can use AI to suggest services to individual citizens. For example, a disabled citizen interacting with government may receive particular service recommendations that they would likely find valuable, based on patterns in preexisting data for citizens with similar disabilities, family support, histories, and other factors. In short, AI allows public-service automation to propose services that are sensitive to a citizen’s personal circumstances. This can work in powerful combination with the other aspect of personalization offered by automation: engagement that reflects human warmth and empathy.
Automation can provide a powerful boost to customer experience of public services—but is a global crisis really a good occasion to take a potentially dramatic step? We would suggest that leaders who are weighing the viability and desirability of automation may find that the pandemic is a window period during which profound change has become possible: some entrenched practices have been loosening their grip, and people globally are more prepared to reevaluate their attitudes.
How to automate successfully: Journey, satisfaction, and change
There is no single way to automate; rather, governments considering the automation of public services can choose from a suite of technological options. In robotic process automation, routine tasks—such as data extraction and data cleaning—are done by machine. Smart workflows integrate tasks performed by humans and machines, such as month-end processes. Optical character recognition converts handwritten, typed, or printed characters into machine-encoded text.
In machine learning, algorithms are employed to learn from experience, without explicitly being programmed with a set of rules. Natural language processing involves the automated analysis of text and speech to derive useful interpretations. And cognitive agents serve as a virtual workforce to support workplace tasks.
How can governments achieve their aim of adopting such technologies in ways that bring sustained outstanding customer experience, without generating issues that impede the success of a project? Leaders of successful automation projects recognize that automation requires changing existing systems and processes, while at the same time retraining and redeploying employees for new tasks. The complexity of implementing such changes is often underestimated. Decision makers who steer effective public services automation projects are smart program managers, ensuring—for instance—that they communicate their vision and progress, and get the most from digital talent. In our experience, successful leaders of automation projects also devote attention to these three areas:
- They view automation from the perspective of the end-to-end customer journey. Seeing the journey through a holistic lens allows automation to be properly integrated through an approach we call automation experience design (AXD), which combines human-centered design and automation.
- They focus automation efforts on three drivers that matter most to customers: simplicity, reliability, and consistency. When resources are constrained—a familiar condition in public service—targeting these drivers realizes the greatest return on investment.
- They invest deeply in change management from the outset. Automation can cause substantial disruption across all parts of an organization and requires careful and considered support for employees and clients throughout the process.
Designing for end-to-end customer journeys
A customer journey consists of a series of steps to complete a given task, including the channels and touchpoints with which users interact and their needs and questions at each stage. For most organizations, mapping these journeys through interviews with customers and analysis can be eye-opening—and for governments, it is especially important. Citizens do not differentiate between different parts of the same government agency, or across distinct government agencies: from their perspective, it is all one experience. Deploying isolated automation interventions without stepping back and understanding the end-to-end journey is unlikely to improve the overall customer experience much. Adopting an AXD approach starts with understanding customer needs, and mapping the journey across touchpoints and channels, including business processes and technologies. McKinsey has recently described a complementary approach, in which civil servants collaborate with citizens in “agile labs” to reimagine service journeys.
This all-embracing attention to the user journey can be seen in an initiative by Australian Unity, a company that offers health, wealth, and living services to more than one million Australians. Australian Unity aimed to improve customer and employee experience, reduce manual labor, and enhance data accuracy. They combined human-centered design and automation technologies to develop customer-centric journeys. The company made an up-front redesign of its processes to determine the greatest opportunities for improvement.
One such opportunity was the creation of accurate customer records. With many customers across diverse services, each record created significant amounts of personal and financial information that required review from skilled employees. For eight months, robotic-process automation handled 42,000 transactions, eliminating 22,493 hours of manual work, and reducing processing time on average from 30 minutes to four minutes, without errors. By automating where the greatest benefit would lie, Australian Unity created capacity for six employees to be redirected to higher-value work, serving customers and designing improved customer experiences. The company’s detailed focus sparked various benefits that a less detailed attention to the user journey might not have revealed.
Understanding the underlying drivers of satisfaction
While there are many drivers of customer experience, McKinsey’s global surveys reveal that customers care most about simplicity, reliability, and consistency when they employ the services of government (Exhibit 3). Consistency involves employing the same standards and approach across channels and services—for instance, where all government services accept the same payment methods. Simplicity is the ability to understand how to complete the service without the need for external support, such as legal advice. Reliability amounts to trustworthy performance across a range of conditions.
Any efforts to improve customer experience should first focus on these areas for maximum impact. Attention to these underlying drivers helps governments establish priorities. There is little value in automating processes and redirecting staff to areas that have a weak impact on overall customer experience. By contrast, it may be well worth automating for the sake of simplicity, reliability, and consistency.
The Department of Welfare in Trelleborg Municipality, Sweden, offers a case of focusing on the drivers of satisfaction when automating critical welfare-support decisions. The department’s mission is to ensure openness, respect, and responsibility to help Swedish citizens regain self-sufficiency.
McKinsey’s global surveys reveal that customers care most about simplicity, reliability, and consistency when they employ the services of government.
By means of a series of measures over the past decade, the department has made extensive changes in its management of welfare-payment applications. Prior to these measures, it took an average of eight to 20 days for the department to make payment decisions. In the interim, citizens would continuously contact the department to inquire about their application status or to ask for help. The department identified this as an opportunity to improve customer experience and decided to use machine learning to support Trelleborg employees’ decision-making process. After this implementation, the period for welfare-support decisions went down to just 24 hours. Through additional capacity, employees could provide higher-value services, helping 22 percent more people than the previous year to find jobs, and to support their efforts at becoming self-sufficient.
In the case of Trelleborg Municipality, we can see the power of a focus on simplicity, radically driving down the time needed to make welfare-support decisions so that employees could focus on higher-value services for citizens. The municipality’s automation-assisted decisions were also consistent, using machine learning to treat citizens’ varied circumstances coherently. And the automated new services worked reliably, allowing important decisions to be made in a trustworthy way.
While simplicity, consistency, and reliability are the key drivers of customer satisfaction, successful public services aim to glean further insights by establishing comprehensive views of customers’ responses, among other indicators of service quality.5 On the most granular scale, operational KPIs and indicators help to track the processes that drive a customer journey, such as the time taken for an individual step in the process. At a higher level, journey level metrics involve customer feedback, including satisfaction scores, at the level of the end-to-end journey. Beyond an individual journey, experience drivers measure the broad elements of customer experiences, at a higher level than an individual journey: how satisfied are customers with the public service more generally, and how do they perceive its brand? And at the highest scale, outcome metrics assess the outcomes that the public service ultimately aims to drive, such as an increase in applications processed in a single visit and higher public trust.
Such metrics call for a range of methodologies, from customer surveys to automated measurements, in which AI is increasingly featured. As these methods continue to develop, one fundamental feature remains: the metrics serve to measure what matters in order to turn these insights into action.
Taking change management seriously
Achieving sustained benefits from automation requires change-management discipline. During the planning stages, governments need to identify which employees, processes, and citizens will be affected, and collaborate with line managers to support stakeholders from inception, not merely after implementation. As a public project moves ahead, leaders can produce highly informed change-management practices by keeping an eye on what we call the five Cs of public-service transformation: committed leadership, clear purpose and priorities, cadence and coordination in delivery, compelling communication, and capability for change.
In many instances, we have seen more than half the budget for new automation and AI initiatives being spent on change management, a sign of the seriousness with which leaders regard this function. Employees can perceive the full value of automation when it is preceded by extensive change management.
As a system forges ahead, it is vital to consider the consequences for everyone who interacts with it, including the experience of all its users. Thus, when Australia’s Medicare stopped accepting cash, it also noted the fact that Australians had been slow to provide bank-account information for payment purposes and anticipated that some citizens would be unable or unwilling to receive direct payments into their accounts. The organization therefore announced special arrangements for such citizens, who could have funds transferred to an account of a family member or friend. For example, Medicare was prepared to provide citizens with checks, or mail them to nominated friends or family.
Automation can be a powerful tool to improve the efficiency and quality of customer engagement, free up civil servants to solve citizens’ problems and add value to their lives, and ultimately strengthen public trust in government. And as the pioneers of public-service automation are showing, public institutions are capable of building cutting-edge automated offerings that provide customer experience comparable to that of lean start-ups or leading private-sector corporations. Those automated offerings embrace technology not for its own sake, but to meet the growing expectations of citizens for reliable, simple, and personalized service from their governments.