European knowledge workers lose 44 working days a year to manual tasks. Learn what that costs, where the time goes, and what organizations like AddSecure, Fazer and ISS did about it.
Every knowledge worker in your organization loses roughly 7.6 hours a week to manual tasks that automation could handle. That is nearly a full working day, every week, for every person on your team. Over a year, it adds up to 44 working days — almost two full months of capacity — quietly consumed by data entry, report assembly, approval chasing and error correction.
For a 1,000-employee organization, the direct salary cost of that time alone comes to an estimated €10.7 million a year. The figure comes from State of Integration & AI 2026 report, the new European benchmark on automation maturity, commissioned by Frends and conducted independently by Sapio Research.
The cost of manual work in this case counts only salary. It does not count the decisions made late, the errors that slipped through, or the strategic work that never got started.
The numbers vary by country, as the report points out, but are large enough to get attention in a budget meeting, regardless of the market or company size. What it does not fully capture is the organizational cost underneath it.
Where time actually goes
Manual work is not evenly distributed, and it is not random. It concentrates around the same types of tasks in almost every organization.
Data entry and transfer is the single largest bottleneck, cited by 33% of respondents as their biggest operational time drain. It is also the clearest signal of an integration problem. When data cannot move automatically between systems, someone has to move it by hand. That person is almost always a knowledge worker who was hired to do something else.
Report generation with multiple data sources follows at 25%. When systems do not share information, producing a single report means pulling from several places, reconciling differences and hoping nothing changed before the deadline. Error checking and correction comes in at the same level, and it is directly linked: the more manual steps in a process, the more places for mistakes to accumulate.
These are consistent signals that the infrastructure underneath day-to-day work is not connected well enough to support the volume and speed modern operations require.
The cost beyond the salary line
The €10.7 million figure already understates the problem, because it only measures what manual work costs in salary terms. The organizational costs that do not appear in that calculation tend to be more consequential.
The State of Integration & AI data shows that 85% of organizations say manual work negatively affects speed of decision-making. Eighty-four percent say it harms data accuracy in reporting. Eighty-seven percent say it increases the risk of human error. These numbers are near-universal across the organizations surveyed.
When reporting is slow, decisions get made on stale information. When data accuracy suffers, confidence in that information erodes, and organizations add manual verification steps to compensate, which slows things down further. When error rates are high in high-stakes processes, the cost of fixing mistakes gets added on top of the cost of making them.
There is also a morale dimension. Eighty-one percent of respondents say manual work negatively affects employee morale. Asking skilled people to spend their days copying data between systems is not a neutral act.
Amanda Lindén, Head of Enterprise Platforms at GlobalConnect, captured the operational cascade well: "Every person that needs to take an action is probably delaying somebody else. If somebody else is delayed, they are likely context switching to other matters and the more work in progress means less work finished."
That pattern plays out across European enterprises at scale, in every sector and country in the survey.
Real-life impact: from 30 minutes to 2
AddSecure, a European provider of secure IoT connectivity and safety solutions, is a practical illustration of what this problem looks like when it is finally measured and addressed.
Before centralizing their integration approach, AddSecure had no overarching strategy for how systems connected. Integrations were built ad hoc, point-to-point, often by individuals who had then left the organization and take their knowledge with them. When the company began growing through acquisitions — 24 in total — the fragmentation became impossible to ignore.
"The problem was that there was no clear strategy for integrations. In cases where integrations even existed, they were very local and built in a way that was considered best practice in the past. It was inefficient, complex, and highly dependent on individuals," said Martin Hult, Director Digital Transformation at AddSecure.
The most concrete example of the cost was order handling. A single order required up to 30 minutes of manual processing across multiple systems: data entered, checked and re-entered by hand, with every step creating risk of error and delay. After automating the entire workflow end-to-end, from web order placement through to system provisioning, that same process now takes two minutes.
That 93% reduction in processing time is one figure. The less visible one is what the team does with the recovered capacity: instead of managing data movement, they focus on customer experience and building new capabilities. The work that was crowding out valuable contribution now happens automatically.
→ Read the full AddSecure story
Scaling beyond human limits
Single-process automation is one thing. Scaling automation across a complex, multi-system enterprise is another. The organizations that do it well share a common decision: they treat integration as architecture, not as a collection of one-off projects.
Fazer, the Nordic food company with operations in eight countries and net sales of €1.1 billion, reached a point where its integration landscape had grown too fragmented to support the pace the business needed. Point-to-point connections had accumulated over years. Adding a new supplier or business function meant navigating that complexity each time, slowing down the work that was supposed to drive growth.
The decision was to consolidate all integrations under a single platform and treat integration as a reusable, governed capability rather than a series of individual technical tasks. The practical result: new EDI partners, new system connections, new business functions — each addition became faster because the architecture was already in place to receive it. As Sami Tillgren, Director of ICT Architecture and Solution at Fazer, put it: "Frends enables us to reuse existing components across various processes, leading to significant resource savings."
Finland's leading facility services provider faced a different version of the same challenge. Before modernizing, ISS was running a fragmented mix of legacy BizTalk solutions, point-to-point integrations, custom RPA scripts and task schedulers, all with poor monitoring and error handling. When something broke, the team's default mode was reactive investigation and reverse engineering rather than proactive management.
After migrating to a unified integration platform, ISS moved from keeping systems running to building capabilities on top of them. Routine tasks were automated, freeing employees for work requiring judgment and context. Generative AI was introduced to handle invoice description writing — correcting errors, sequencing tasks, removing inconsistencies — with human oversight rather than human effort doing the repetitive work. The shift positioned ISS among the leading operations in the global ISS Group for automation and data capabilities.
Sami Öfverberg, Head of Automation Technologies at ISS, summed up the principle: "Technology should simplify complexity and empower people to do their best work to deliver better experiences for customers."
What recovered time would buy
When survey respondents were asked what they would do with time freed by successful automation, the answer was unambiguous: 40% said they would reinvest it in accelerating AI adoption. Data analysis and strategic decision-making came second at 38%. Strategic planning and innovation followed at 32%.
Cost reduction and headcount reduction ranked last.
That result clarifies what the manual work problem is actually costing. Organizations do not primarily want to do the same work with fewer people. They want their people doing different work, one that requires judgment, context and creativity rather than repetition.
The 44 days a year currently lost to manual tasks are 44 days not going into analysis, customer engagement or building new capabilities.
The expected revenue upside from addressing this reinforces the point: respondents estimate an average gain of 11.3% if all automatable manual tasks were handled through automation. Half believe the uplift could exceed 10%.
Toward the agentic horizon
The case for reducing manual work has always been clear in operational terms. What changes now is the cost of not acting.
AI agents, systems capable of taking autonomous action across enterprise workflows, are moving from pilots into practical planning. Ninety-seven percent of organizations in the State of Integration & AI survey are already factoring them into their governance thinking. The infrastructure question those systems raise is the same one that determines whether manual work automation succeeds in the first place: are the underlying systems connected well enough to support automated action at scale?
The organizations seeing the most consistent results are those that built the integration layer first. They report 43% faster project delivery, 35% better AI and automation ROI and 33% easier AI adoption compared to those that did not. The gains compound when the architecture is already in place, and stall when it is not.
See where your organization stands
The State of Integration & AI 2026 is the first independent European benchmark of AI adoption, integration maturity and automation readiness, conducted by Sapio Research across 611 IT decision-makers in six European countries and nine industries. It covers every dimension of this challenge: manual work costs by country and sector, integration maturity, AI project success rates and the governance readiness of organizations preparing for what comes next.
→ Download the full State of Integration & AI 2026 report
FAQ: The cost of manual work in the enterprise
How much does manual work cost European enterprises?
According to State of Integration & AI 2026, the mean annual cost of unnecessary manual work is €10.7 million for a 1,000-employee organization. That figure covers direct salary cost only, not the downstream impact on decisions, errors and missed opportunities.
What are manual work automation statistics for European enterprises?
Knowledge workers spend a mean of 7.6 hours per week on tasks that could be automated, or 44 full working days per year. Data entry and transfer is the top bottleneck at 33%, followed by report generation and error correction at 25% each. Germany reports the highest individual burden at 8.5 hours per week; the Netherlands the lowest at 6.6 hours.
What is the productivity loss from disconnected systems?
Beyond salary cost, disconnected systems slow decision-making (85% of organizations), reduce reporting accuracy (84%), increase human error risk (87%) and negatively affect employee morale (81%). These second-order costs do not appear in salary-based calculations, meaning the true productivity loss is higher than the headline figure.
What are the automation benefits for enterprises?
Organizations with an integration-first approach report 43% faster project delivery, 35% better AI and automation ROI and 33% easier AI adoption. Combined AI tools and workflow automation save a mean of 16 hours per employee per month. At the process level, outcomes like AddSecure's 93% reduction in order handling time show what targeted automation delivers in practice.
Why is integration the prerequisite for reducing manual work?
Most manual work exists because systems do not share data automatically, forcing people to act as connectors between them. Adding automation tools on top of a fragmented architecture tends to create new handoffs rather than eliminating existing ones. Building integration as a design decision before deploying automation on top of it is what determines whether productivity gains actually materialize.
How does manual work affect AI project success?
Among organizations that have piloted or deployed AI, integration challenges are the single largest barrier to further progress, tied with skills gaps and security concerns at 36%. AI that cannot reliably access current data across connected systems cannot deliver the outcomes it was built for. The manual labor problem and the AI productivity problem share the same root cause.