Relevant automation potential in almost all sectors

Key messages from the McKinsey study "A Future That Works" (2017)

Opportunities and risks of automation

Industry-specific insights

Implications for corporate decision-makers

Conclusion

The Oops, an error occurred! Code: 20251209163316797f2c3e provides a sound framework for the discussion. This study analysed the technical automation potential of various industries and activities worldwide as well as the possible effects on productivity, economic growth and employment. The key findings of the study include

Less than 5% of all current jobs could be fully automated with the technologies already available. In other words, there are hardly any jobs in which all tasks can be performed by machines.

However, almost all occupations can be partially automated - i.e. a significant proportion of their activities can be automated. McKinsey quantifies this partial potential on a global level: around 50% of all work activities (measured in terms of time spent) could theoretically already be automated today using existing technologies. Around 60% of professions have at least 30% of their tasks that could be automated.

These figures illustrate the enormous technical potential - and that automation mainly affects specific tasks within jobs, rather than replacing entire jobs at the touch of a button.

The study analysed activities at a granular level and found that the degree of automation strongly depends on the type of activity. Routine activities in predictable environments or those with clear rule-based processes are the most likely to be automated. Specifically, physical work in predictable environments, for example, has an automation potential of around 81%, data processing activities of around 69% and data collection tasks of around 64%.

In contrast, activities that require high social or cognitive skills - such as managing employees, creative problem solving or interpersonal interaction - are much more difficult to automate (often less than 20% automation potential with current technology).

This discrepancy explains why almost every profession has some automation reserves, but can rarely be completely replaced by machines: In one and the same job, there are usually both repetitive routine tasks (automatable) and tasks that require human dexterity (not automatable for the time being).

The proportion of potentially automatable activities varies considerably depending on the economic sector. Industries with many structured, repetitive processes - e.g. manufacturing - naturally have higher automation rates than those in which human interaction or complex expertise are the main focus (e.g. education or healthcare). The following list provides an overview of selected sectors.

• 73%: Hospitality industry (accommodation & catering)
• 60%: Production / Manufacturing
• 57%: Transport and storage
• 53%: Retail trade
• 44%: Wholesale
• 43%: Finance & Insurance
• 36%: Health & social services
• 27%: Education

As the list makes clear, activities in an industrial environment or in retail are much more threatened or characterised by automation, while in areas such as healthcare or education, many jobs are characterised by human interaction, expertise or unpredictability - which current machines are less good at.

However, McKinsey emphasises that there can be major differences within each industry. For example, in manufacturing, certain job profiles such as welders or plant operators, which predominantly consist of predictable physical tasks, have an automation potential of over 90%, while in the same industry, for example, customer service activities are only less than 30% automatable.

The same applies across all sectors: higher-skilled (and better-paid) jobs are on average less automatable than simple, repetitive jobs - but in principle, almost all occupations have at least some tasks that could be automated.

The widespread introduction of automation technologies could trigger a strong surge in productivity. McKinsey estimates that automation could increase annual global productivity growth by around 0.8 to 1.4 percentage points.

By way of comparison, productivity growth in many developed economies has often only been in the region of 1-2% per year over the last few decades. An increase of up to 1.4 percentage points would therefore be significant - and would come at just the right time, as the labour force is shrinking in many countries and economic growth is difficult to achieve without productivity gains. Automation could make it possible to generate further economic growth and increases in prosperity despite an ageing workforce.

In addition to pure productivity at a macro level, the study also cites economic benefits: Companies can not only save on labour costs through automation, but also, for example, shorten throughput times, increase output, reduce quality defects and increase plant running times. These performance benefits have a positive effect on turnover and competitiveness - a decisive incentive to invest in automation.

Perhaps the most discussed topic is how automation affects jobs. The McKinsey study paints a differentiated picture here. On the one hand, the theoretical potential (50 % of all working hours) corresponds to the work of over 1.1 billion employees and wages totalling around 15.8 trillion US dollars worldwide. These figures illustrate the scale of the possible shifts - however, this is a long-term potential, not a short-term scenario of mass unemployment.

McKinsey emphasises that historical experience (e.g. the transition from agriculture to industrial society) shows that technological progress destroys certain jobs: Technological progress destroys certain jobs, but also creates new ones - often different ones than before, but on balance, employment losses could be offset in the long term.

The transition phase is important: workers whose tasks are automated have to switch to other activities. The study assumed that redundant employees find other jobs - only then can the productivity gains be fully converted into economic growth.

The challenge is therefore not so much whether enough new work will be created, but how the transition will succeed. In the short term, friction could arise on the labour market: certain professions could shrink while demand in others increases.

The study draws a comparison with previous major shifts in the labour market (e.g. the decline in agricultural jobs over decades) and sees the coming wave of automation on a similar scale - with the difference that the change could possibly take place more quickly today.

How quickly will automation catch on? McKinsey provides an interesting outlook here: In a medium scenario, half of today's work activities could be automated by around 2055. Depending on the framework conditions, however, this date could be 20 years earlier or later - i.e. somewhere between the 2030s and 2070s.

This wide range shows that the speed of adoption is uncertain and depends on many factors. Among other things, the study cites technical feasibility in the respective use case, the costs of development and implementation (vs. labour costs), labour market conditions (are there enough qualified personnel available or does a shortage of labour favour automation?), the economic benefits (business case for companies) and social and regulatory acceptance as decisive influencing factors.

In other words: Just because something is technically possible does not mean that it will be used everywhere immediately - economic considerations as well as ethical, social and legal issues will help determine the pace. Accordingly, the forecast "50% by 2055" should not be understood as a fixed date, but rather as an indication that the change will take place gradually over several decades.

On the one hand, companies and society still have time to adapt; on the other hand, the scenario emphasises that the transformation is inevitably progressing and that significant changes may already become noticeable in the next 10-20 years.

Automation has enormous disruptive potential for the world of work, but its actual impact depends on how we deal with it. In the next step, we take a look at the opportunities and risks arising from this development.

Like every technological upheaval, the wave of automation brings with it both great opportunities and challenges. It is essential for decision-makers to understand both sides in order to successfully shape the transformation.

From an economic perspective, the benefits of automation lie primarily in productivity increases and efficiency gains. Routine tasks that previously tied up manual labour time can be automated to run much faster and with fewer errors. This speeds up processes, increases throughput and capacity and reduces costs. Repetitive tasks in particular, where there are sources of human error, benefit from the more consistent quality of automated processes (e.g. machine precision work or algorithmic data processing). Companies can thus achieve competitive advantages - be it through lower unit costs, higher product quality or faster delivery times.

Automation can also lead to completely new business models (such as "smart" services that are provided around the clock without human intervention).

At a macroeconomic level, automation promises to boost growth: it can deliver a significant proportion of the productivity growth required to compensate for the demographic decline in the labour force in many countries, for example. Higher productivity as a result of automation tends to contribute to gains in prosperity - for example in the form of rising economic output, potentially falling prices or new investments.

Last but not least, automation offers the opportunity to relieve workers of dangerous, monotonous or physically demanding tasks. Robots can, for example, work in hostile environments (mining, chemical plants) or take over tiring assembly line jobs, while human employees are given more demanding and varied tasks.

Ideally, automation will allow people to utilise their creative, strategic and social skills to a greater extent instead of being trapped in repetitive routines. Especially in times of skills shortages (in some industries), automation can help alleviate bottlenecks - by having some of the work done by machines, scarce human resources can be better targeted to where they add the most value.

On the other side of the coin, there are considerable labour market and social challenges. The central question is what will happen to jobs whose tasks are increasingly automated.

In the short term, there is a risk that jobs will disappear or change radically, which means uncertainty for the employees affected. Studies such as McKinsey's forecasts or the often-cited Frey/Osborne Report (Oxford 2013) have cited some high numbers of potentially affected jobs - even if these figures do not mean that all these jobs will actually disappear, they do emphasise the scale of the problem. Without countermeasures, automation could lead to rising unemployment in certain occupational fields, particularly in routine-intensive activities. Historically, such transition phases have often been accompanied by social stress (e.g. structural change in former industrialised regions).

Another risk is the polarisation of the labour market: if mid-level qualifications in particular (e.g. clerks, factory workers) are rationalised away, income disparities could grow - well-trained specialists benefit from technology, while lower-skilled workers compete for simple jobs. Further training and retraining are key to counteracting this (more on this in the implications section). Regional disparities can also be exacerbated if entire regions are dependent on a particular industry that is heavily automated (e.g. automotive production), while growth sectors (tech, services) are located elsewhere.

Questions of acceptance also arise in society: will people welcome automation if, for example, they are cared for by robots as patients or only served by chatbots as customers? Trust, ethics and legal framework conditions play a role here - such as liability issues in AI decisions or data protection.

Furthermore, the integration of automation systems initially requires high investments and the return on investment is not always immediately guaranteed. Small and medium-sized companies may find it difficult to keep up financially and in terms of know-how, which can lead to a digitalisation gap (large pioneers vs. laggards).

Finally, it must be remembered that new technologies also bring new risks - e.g. IT security risks (cyberattacks on automated systems) or dependencies on certain providers or platforms.

An often overlooked opportunity in the face of these risks is the emergence of new jobs and occupational fields. Technological upheavals have always created new jobs in the long term - just think of the boom in IT professions since the 1990s, which hardly anyone could have foreseen 50 years ago. Automation will also create professions, some of which do not even exist today. Examples are already emerging: data analysts, AI specialists, robotics engineers, machine learning trainers - these roles will be increasingly in demand to develop, monitor and optimise the new automated processes.

At the same time, automation is giving rise to new services and business models that require personnel (e.g. experts for robotic process automation in process organisation, or AI ethicists to monitor fair algorithms).

Activities mediated between humans and machines could also grow outside the tech sector - for example, maintenance technicians for collaborative robots, coaches who teach employees how to use AI systems or specialists in change management who accompany the digital transformation.

McKinsey predicts that the overall demand for more highly qualified profiles and for social skills professions (e.g. care, education - professions that require empathy) will increase in the coming years, while the demand for purely routine employees will decrease.

The challenge for the education and training system is therefore to equip workers at an early stage with the skills that will be in demand in the new world of work. If this succeeds, many employees can switch from their old job to new roles instead of becoming unemployed. If, on the other hand, it does not succeed, there is a risk of a shortage of skilled labour in growth areas and, at the same time, unemployment in shrinking jobs - a paradoxical imbalance.

Automation is neither a blessing nor a curse per se - it is a tool. Its net effect depends on how we humans deal with it. If it is managed proactively, the opportunities (productivity, new value-adding activities, relief from stressful jobs) can outweigh the risks. However, if active management is neglected, the risks (job losses, skills gaps, social tensions) can become real problems. For decision-makers in companies, this means setting the course now to successfully and responsibly usher in the era of automation.

Automation does not have the same effect in all sectors. As already indicated in the core statements, there are major differences in automation potential from sector to sector. Here we take a focussed look at some sectors - especially those highlighted in the McKinsey study as being particularly affected or predestined, and which are relevant for corporate decision-makers in areas such as manufacturing, retail, healthcare and the financial sector.

Industrial manufacturing has always been one of the pioneers of automation - from mechanical machines during industrialisation to assembly lines, industrial robots and today's smart factory approaches. The theoretical automation potential in this sector is around 60%. This means that more than half of the working time spent in production consists of activities that could be automated using current technology. This mainly includes repeatable physical tasks in a predictable environment: e.g. assembling parts, welding, painting, packaging, quality inspection or transporting materials within the factory. Many of these tasks are already carried out by industrial robots or automated systems - and the trend is rising as robots' capabilities increase (e.g. visual inspection using AI cameras, flexible robotics for smaller batch sizes, etc.).

The opportunities in production are correspondingly large: higher quantities with lower labour input, lower error rates, 24/7 operation without fatigue. However, implementation often requires considerable investment in machines, IT systems and the reorganisation of processes.

Employee profiles in production are shifting: while the need for purely manual assembly line workers is decreasing, the need for technicians, mechatronics engineers and data analysts to set up, monitor and maintain the automated systems is increasing. Flexibility remains a challenge - people can often adapt to product variants or changes more quickly than a rigidly programmed machine. This is why modern manufacturers are increasingly relying on collaborative robots (cobots) that work side by side with humans and can be reprogrammed relatively easily in order to remain flexible.

Corporate decision-makers in industry should examine which production steps are still manual today and can be made more efficient through automation, without neglecting the workforce - ideally, employees should be trained for higher-value tasks (e.g. monitoring robots, process optimisation) instead of being completely replaced.

In the retail sector - from supermarkets to wholesale warehouses - the potential for automation is also relatively high (retail approx. 53%, wholesale approx. 44%). Standardised processes are particularly affected here.

In retail, we are already seeing automation in everyday life: self-service checkouts and scan & go systems are partially replacing cashiers, digital price tags and ordering systems are reducing manual price labelling, and fully automated warehouses are being used in e-commerce.

Companies such as Amazon operate logistics centres in which robots transport shelves and pick orders. There are also initial trials with robots in bricks-and-mortar retail (such as inventory robots that scan shelves or service robots that show customers the way).

Wholesalers and distribution centres have long relied on conveyor systems, sorting systems and automated high-bay warehouses. The transport of goods (logistics) is closely linked to retail and will also be automated in the future by autonomous vehicles or drones.

The effects: Routine jobs such as shelf replenishment, warehouse work, order picking and cashiering could decrease significantly. At the same time, retail staff could take on more advisory and management tasks - e.g. customer advice, problem management, store organisation.

One opportunity lies in cost reduction and speed: with automation, retailers can operate around the clock (process online orders immediately), shorten delivery times and deploy staff more specifically for peak times.

However, there are risks if simple tasks are no longer performed: Retail was an important employer for low-skilled labour in particular - there is a risk of unemployment here if there is no transfer to other areas. Customer acceptance also plays a role: not every customer wants to interact with a kiosk terminal instead of a person.

Retailers therefore need to find a balance between using automation where it brings real added value (e.g. faster processing, relief from heavy or tedious work), but continuing to offer human service where it is appreciated.

Strategically, retail companies should now invest in digital infrastructure and process automation (keyword omnichannel, seamless integration of online and offline) in order to remain competitive, but also train their staff for new tasks in customer service, for example.

Banks, insurance companies and financial service providers are considered to be predestined for software-based automation, as their core business consists of information processing. The McKinsey potential here is around 43% - at first glance lower than in industry or retail. However, this is somewhat deceptive, as almost all processes in the financial sector can be digitalised in principle.

Many institutions have already implemented extensive RPA (Robotic Process Automation)-initiatives to automate repetitive back-office tasks: Forms processing, data entry into legacy systems, reconciliation of records, reporting, etc. Such software robots can book transactions around the clock, check invoices or answer customer enquiries via chatbot.

Advanced AI is also making inroads: for example, in fraud detection (recognising conspicuous patterns in transactions), credit analysis (scoring borrowers) or automated investment advice (robo-advisors).

The opportunities for financial companies are obvious: automation saves human resources, reduces errors (e.g. due to incorrectly recorded data), improves compliance (automated checks) and shortens processing times for customers (e.g. credit decision in minutes instead of weeks). At the same time, employees can be relieved of monotonous tasks and deployed for higher-value activities - such as advisory services, complex case processing or customer care.

However, there are also risks/hurdles: Legacy IT systems in banks sometimes make it difficult to introduce automation, as legacy systems cannot be easily integrated (RPA bots that use screen scraping can help here - but this is an interim solution). In addition, strict regulatory requirements must be met, which requires high quality and control of the automated processes.

In terms of employment, the number of clerks in traditional administrative processes (policy management, account management, etc.) could decrease, for example. Instead, the need for IT specialists, data scientists and compliance experts will increase. Overall, however, the industry expects a transformation of job profiles rather than a clear-cutting: tomorrow's bank clerk will probably spend more time analysing and advising and less time typing data.

Corporate decision-makers in the financial sector should now examine which processes can be made more efficient using RPA or AI - competitors are already doing this. At the same time, it is important to take employees along on this journey: through retraining (e.g. as data analysts) and a culture that welcomes innovation in order to reduce fears.

At around 36%, the healthcare and social services sector has one of the lowest automation potentials. This is due to the fact that many tasks in this sector have difficult sensory or interpersonal components - such as patient care, medical diagnostic consultations or therapeutic activities.

Nevertheless, there is also significant potential for automation in the healthcare sector, particularly in administrative procedures and diagnostic processes. Example: Hospital administration (scheduling, billing, documentation) can be digitalised and automated to a large extent. Modern clinics are increasingly relying on electronic patient files, automatic appointment scheduling systems and AI-supported documentation assistants to relieve doctors and nursing staff of paperwork.

In diagnostics, AI systems are already demonstrating impressive capabilities - for example in recognising abnormalities in X-ray and MRI images or in pathological findings. Such algorithms can support doctors as a "second opinion" and independently analyse routine cases (e.g. eye screenings for diabetes patients). In surgery, robotic systems (e.g. the da Vinci surgical robot) help surgeons to perform operations more precisely and minimally invasively, which ultimately contributes to better patient outcomes and faster recovery. However, the robots remain tools in the hands of the doctor, not autonomous actors.

The opportunities offered by automation in the healthcare sector lie primarily in relieving the burden on skilled labour and cushioning bottlenecks: Nurses and doctors can concentrate more on direct patient care if bureaucratic activities are automated. Telemedicine and digital monitoring can also increase the efficiency of care, especially in ageing societies with an increasing number of patients.

On the risk side, there are aspects such as liability (who is responsible if an AI misses a finding?), ethical questions (does the patient want to be diagnosed by a machine?) and, above all, the interpersonal aspect: health and care rely heavily on trust and empathy. Fully automated care robots could take over some technical tasks (there are robots that lift patients out of bed or hand them food), but they are no substitute for human care. Therefore, the following applies here: Automation yes, but only to support people, not as a replacement for core care.

It is important for decision-makers in the healthcare sector to understand digitalisation and automation as a means of making the best possible use of scarce resources (staff, time). Investments in hospital IT, networked devices (IoT) and AI diagnostics make strategic sense, but they must be introduced cautiously and always in the interests of patient well-being.

In addition to the examples mentioned above, there are of course other sectors:

• Transport and logistics, for example, are undergoing a profound transformation as a result of self-driving vehicles, drone deliveries and automated warehouses - in the long term, there could be a significant reduction in lorry drivers or warehouse workers, while fleet managers or maintenance technicians for autonomous vehicles will increase.
• The energy and utilities industries are focussing on smart grids and automated power plant control.
• The construction industry is experimenting with autonomous construction machinery and 3D printing of building components.
• The IT sector itself uses automation (e.g. automatic code generation or cloud orchestration), whereby routine coding tasks in particular are being eliminated, but creative development jobs remain.
• Finally, the education sector has the lowest automation potential - teachers and educators will not be replaced by robots any time soon, but assistance systems (e.g. adaptive learning software, automated assessment of tests) can support teaching staff.

Each sector has its own specifics - and decision-makers would do well to keep a close eye on the relevant studies and pilot projects in their sector in order to assess where their company stands.

For managers - especially in IT, operations and HR - the question is: What does all this mean in concrete terms for our strategy and HR planning? How can we prepare for the era of automation and play an active role in shaping it? Below are some strategic recommendations and fields of action derived from the findings:

Companies should not leave automation to chance or individual initiatives, but should formulate a clear strategy.

This begins with the identification of use cases: Which processes offer the greatest potential for automation and at the same time a high level of benefit (cost savings, quality gains, speed)? A cross-departmental task force (ideally led by Operations with strong involvement from IT and the specialist departments) can analyse and prioritise processes.

This can be used to create a roadmap of which automation projects should be implemented in the short, medium and long term. Important aspects here are the profitability analysis (ROI of each measure) and feasibility checks (technical availability of solutions, maturity level).

Top management should actively support this roadmap and review it regularly. A strategic plan helps to avoid isolated solutions and ensures that automation is implemented in line with the company's objectives.

Implementing automation requires the right tools and, if necessary, partners. The IT department plays a key role because it has to create the technical prerequisites - from infrastructure for AI applications to the integration of robotics into existing systems.

Companies should evaluate which technologies are relevant for them (e.g. RPA software for administrative processes, AI platforms for data analysis, IoT sensor technology for production, etc.) and make targeted investments in these.

Not everything always has to be developed in-house - collaboration with technology partners can bring expertise and speed. Many companies cooperate with start-ups or specialised providers, for example, in order to implement pilot projects quickly.

It is important to maintain an open architecture in order to be able to flexibly integrate different solutions. In the case of physical automation projects (robots, machines), suppliers or system manufacturers are often also important partners.

Decision-makers should also take a realistic view of the cost-benefit analysis: Automation is not an end in itself, but must pay off in better performance. Pilot projects are therefore useful for testing on a small scale and learning from experience before rolling out on a large scale.

Perhaps the most decisive success factor is people. Without the acceptance and qualification of employees, no automation initiative will succeed in the long term.

Companies should invest in training programmes at an early stage to make their workforce fit to work with new technologies. This includes specialised training (e.g. operation of new software/robots, basics of data analytics for employees who were previously unfamiliar with the subject) as well as the promotion of digital skills and a willingness to learn. According to experts, anchoring a culture of lifelong learning in the company is essential to ensure that employees continuously adapt to changing requirements. This can be supported by e-learning platforms, internal knowledge communities or cooperation with training providers.

HR managers in particular should anticipate the need for retraining: If, for example, certain activities are to be discontinued in the future, consideration should be given at an early stage to which new roles these employees could take on and which qualifications are required for this.

In addition to skills, it is also about change management: changes understandably trigger uncertainty or resistance. Transparent communication about automation plans, highlighting the opportunities for everyone (e.g. relief from routine, opportunities for further development) and involving employees in the change (e.g. obtaining feedback, pilot teams with volunteers) are important to reduce fears.

Managers should act as promoters of change, but at the same time respond empathetically to concerns. Ultimately, technology alone is not enough - the workforce needs to be on board. Successful companies therefore not only promote technology, but also the talent within the organisation.

Automation and digitalisation often go hand in hand with new forms of organisation. Rigid departmental boundaries and hierarchies can be a hindrance when work processes are digitally reorganised.

Many companies are realising that they need to work in a more agile and cross-functional way in order to successfully implement automation projects. Cross-functional teams in which IT experts work together with technical experts and process owners to develop solutions are a tried and tested model. The concept of the "agile organisation" is also gaining in importance: i.e. networks of teams, flat hierarchies, iterative project procedures (e.g. Scrum) in order to be able to react quickly to changes.

Company decision-makers should check whether their organisation is still suitable to cope with the coming changes. Reorganisation may be necessary, such as the creation of a centre of excellence for automation that develops and manages best practices across departments. The division of tasks will also change: Some traditional tasks will be eliminated, but new ones will emerge, and these will have to be integrated into the organisation in a meaningful way. One keyword is "unbundling and rebundling of work " - i.e. breaking down work processes into individual tasks and reassembling them, possibly dividing them up into different roles (some of them are performed by machines, some by specialised employees).

Automation can also make it efficient to create centres of excellence - e.g. to have routine tasks carried out centrally by a small automated team instead of distributed in each department.

Overall, the organisational design should be regularly reviewed: it must be flexible enough to integrate automation and clear enough so that responsibilities do not become diffuse.

Finally, automation also affects managers and the HR function: managers need a good understanding of the new technologies in order to make the right decisions. HR must evolve from a purely administrative role to a strategic partner that utilises workforce analytics to manage workforce transformation with data support. Many companies are now factoring automation into their HR and succession planning in order to develop talent for future roles at an early stage.

The target groups named in the brief - IT, Operations, HR - should work hand in hand in the automation initiative. Each of these functions contributes an important perspective: IT managers know what is technologically feasible and how new tools fit into the IT portfolio; operations managers know the processes in depth and know where there is pain and potential; HR understands the workforce, skills profiles and the need for change.

Only together can a comprehensive automation concept succeed. In practice, regular consultations should take place and an automation steering committee should be set up, possibly with representatives from all three areas. This ensures that, for example, new RPA software (IT topic) is considered in terms of aspects such as process selection (Ops topic) and impact on job profiles (HR topic). This cross-departmental collaboration is not always easy for cultural reasons - IT and specialist departments often had to work separately in the past.

But in the age of digitalisation, the boundaries are becoming blurred: Technology is a business process, and business processes depend on technology; and both influence employees. A common understanding, supported by top management, helps to break down silos.

Companies that are already further along in this respect rely, for example, on interdisciplinary workshops or "automation labs" in which employees from different departments develop solutions together.

Act proactively instead of reactively. Companies that align their strategy, technology and employee development with automation at an early stage will master the transformation better than those that wait and see.

As McKinsey puts it, companies need to "retool" their structures, processes and talent in order to realise the full potential of new technologies. Although this requires change and sometimes far-reaching changes, it offers the opportunity to emerge stronger from the wave of automation.