Brain drain in companies: causes, costs and countermeasures for SMEs
A typical pattern: a long-serving project manager retires. The costly departure describes preparing a handover before the final working day. For SMEs, knowledge preservation needs to capture customer decisions before retirement turns routine questions into investigations. A few months later a long-standing customer asks why a component has for years been made differently from its data sheet. Nobody in the building knows anymore. That is what brain drain in a company usually looks like. There is no big bang, just a growing number of questions nobody can answer.
At RECO we think the turnover rate is the wrong metric for managing this risk. Many leadership teams take comfort in a low rate, and the argument has some truth to it. Few departures mean few replacements. What it misses is that brain drain grows with the amount of knowledge that depends on a single person, and hardly at all with the number of departures. A company with very low turnover can be hit harder than one with high turnover, if the few people who leave happen to be the one who does the costing and the one who owns the biggest customer. In family and traditional businesses, knowledge transfer should prioritise concentrated pricing and customer expertise rather than assume long tenure makes continuity secure.
Where the term comes from
Originally, brain drain described highly qualified people leaving a country. The term was coined in Britain in the early 1960s, when the Royal Society documented British scientists emigrating to the US. In Austria it still comes up in debates about doctors and researchers moving abroad.
Inside a company, brain drain means losing business-critical knowledge when people drop out: through resignation, retirement, parental leave, long illness or an internal move. The departure itself is rarely the problem. It gets expensive when the knowledge sat with that one person and nowhere else.
Why experience is what disappears
The philosopher Michael Polanyi wrote in 1966 that we know more than we can tell. Ikujiro Nonaka and Hirotaka Takeuchi built their SECI model on that idea in 1995. It distinguishes two kinds of knowledge.
- Explicit knowledge can be written down, such as process descriptions, price lists or technical drawings.
- Tacit knowledge lives in experience, such as which customer needs which tone, why a machine is set differently in cold weather, or who really makes the call at a supplier.
Brain drain hits tacit knowledge above all. Manuals and wikis record the outcome of a decision, rarely the reason for it. In his book Lost Knowledge, David DeLong argues that companies with an ageing workforce mainly lose judgement. The fact that something is done a certain way often survives. The knowledge of when it is fine to deviate does not.
Why SMEs in the DACH region are more exposed
Three trends reinforce each other.
- Demographics: according to Statistik Austria's population projection from November 2025, Austria's population aged 20 to 64 peaked in 2024 at about 5.57 million and will shrink by around 256,000 people by 2040. Over the same period the number of people aged 65 and over rises from 1.84 to 2.46 million.
- Low redundancy: in a large corporation almost every role has a backup. In a mid-sized company there is often exactly one person who truly understands the costing, the ERP customizing or the most important customer.
- Engagement: in the Gallup Engagement Index Germany 2025, 13 percent of employees have no emotional bond with their employer and 77 percent only a weak one. Among those with no bond, 39 percent are actively looking for a new job, compared with 3 percent of the highly engaged.
Then there is business succession. According to KfW Research, around 109,000 owners of small and medium-sized businesses in Germany plan to step back every year until the end of 2029. When the founder leaves, the whole web of relationships with customers and banks often leaves with her.
Measure knowledge concentration, not departures
What we recommend instead fits on one sheet of paper. Score every key role from 1 to 3 on three axes and multiply the scores.
- Criticality: 1 if losing the person is a nuisance; 2 if it costs weeks or individual orders; 3 if it puts revenue or the ability to deliver at risk.
- Likelihood of departure within a planning horizon you set yourself: 1 with no signals; 2 with signs of a job change or planned leave; 3 if the departure is certain, for example retirement or a resignation.
- Missing backup: 1 if a second person can fully take over; 2 for a partial backup; 3 if there is nobody.
We deliberately give no point threshold at which you must act. There is no solid basis for such a cut-off, and it would suggest a precision that three rough estimates do not have. The product is for ranking. Sort the roles from highest to lowest and work down the list. A 3 for missing backup always deserves a second look, even when the other scores are low.
What rises to the top is often telling. The highest-scoring role is frequently one that appears in no turnover statistic: someone who has not resigned, but will retire or go on leave, and is the only one who knows how something works.
What the loss costs
For the visible cost of replacing someone, SHRM gives 50 to 200 percent of an annual salary, depending on level. The invisible part is harder to pin down. In Panopto's 2018 Workplace Knowledge and Productivity Report, respondents said that 42 percent of the knowledge in their company is known only to the individual. According to the same survey, US knowledge workers lose 5.3 hours a week waiting for information from colleagues or recreating knowledge that already exists. That is an American survey, not an SME figure from the DACH region. As an order of magnitude it still shows why a single departure ties up so much time.
We have split the cost question into two articles, because they are two different calculations. How to work out the departure of a single key person, including every hidden item, is covered in What does it really cost when a key employee resigns? What turnover costs across a whole year and the entire workforce is in What every resignation truly costs your company. For a first estimate for your company, use the free Brain-Drain Calculator.
Four mistakes that make the risk bigger
- Confusing brain drain with turnover. A single departure in a key role can cost more than several departures in well-documented roles.
- Relying on the handover. A notice period is not enough to pass on years of experience, and handover notes usually stay on the surface.
- Treating knowledge management as a software project. A new wiki nobody fills solves nothing. Documentation has to happen in daily work, not as an extra task on a Friday afternoon.
- Thinking only about retirements. Parental leave, educational leave, long sick leave and internal moves leave the same gap, just for a limited time.
Countermeasures, sorted by effort
Draw the knowledge map
The scoring above needs no software, only the people who know the roles. Everything else builds on it, because it shows where to start and where you can wait.
Start early when a departure is foreseeable
For planned retirements, knowledge transfer should start as soon as the date is known, not at handover. The Fraunhofer IPK guide Wissen erfolgreich bewahren puts it plainly: the more preparation time there is, the more extensive the knowledge retention can be. It names a generous transition period, with the position staffed twice, as the safest route. Tandems of an experienced and a younger colleague, lessons-learned sessions after projects and interviews with the experts themselves have proven useful. Paul H. Schindler describes how such a tandem transfer works in Knowledge transfer before retirement.
Record the reasons
For important decisions, ask why, which alternatives were rejected and which exceptions apply. This is the knowledge that is missing most after a departure, and no process manual contains it.
Document where knowledge is exchanged
Most experience is passed on by talking, in project meetings, customer calls and employee reviews. If it is captured there, securing knowledge costs hardly any extra time. That is what we built AI-supported conversation documentation for.
Take retention seriously
The cheapest way to keep knowledge is for good people to stay. The Gallup gap between 3 and 39 percent job hunting shows how much engagement matters. Regular, well-prepared reviews and real development prospects lower the risk before it appears. And if someone leaves anyway, the same documented knowledge decides how quickly the successor can work. Our use case on hiring interviews and onboarding shows what that looks like.
What AI changes and what it doesn't
AI mainly lowers the effort. Language models can summarise conversations, sort them by topic and make them searchable. Speech recognition works well for standard German, but with strong dialect, jargon and background noise it still makes mistakes, as our comparison of speech-to-text models shows. Whatever never comes up in a conversation, even the best AI will not capture.
Legally, every recording needs a basis under the GDPR. In Austria, section 96a(1)(1) ArbVG requires works council consent for systems that process employees' personal data automatically beyond general personal details and professional qualifications. In Germany, section 87(1) no. 6 BetrVG gives the works council a say on technical systems intended to monitor employees' behaviour or performance. How to handle this in practice is covered in our article on documenting employee conversations. This section is not legal advice. You will find more in the topic hub AI in mid-sized companies.
Our suggestion for this week is a modest one. Score your most important roles with the scheme above, sort the list and take on the top role first. If you need a euro figure to go with it, work it out with your own inputs in the Brain-Drain Calculator.
Sources
- Polanyi, M. (1966): The Tacit Dimension. University of Chicago Press
- Nonaka, I. & Takeuchi, H. (1995): The Knowledge Creating Company. Oxford University Press
- DeLong, D. W. (2004): Lost Knowledge. Confronting the Threat of an Aging Workforce. Oxford University Press
- Royal Society (1963): Emigration of Scientists from the United Kingdom
- Gallup, Engagement Index Germany 2025
- SHRM Executive Network, The Myth of Replaceability
- Panopto (2018): Workplace Knowledge and Productivity Report
- Statistik Austria, press release on the population projection (12 Nov 2025)
- KfW Research, Succession Monitoring Mittelstand 2025 (Fokus no. 526, January 2026)
- Section 96a ArbVG (Labour Constitution Act, Austria)
- Section 87 BetrVG (Works Constitution Act, Germany)
- Fraunhofer IPK (2016): Wissen erfolgreich bewahren (guide to systematically supported transitions of specialists and managers)
Frequently asked questions
What is brain drain in a company?+
The loss of business-critical knowledge when employees drop out and that knowledge is not recorded anywhere else. It is different from know-how flowing to a competitor: that is about trade secrets and contract clauses, while brain drain is about knowledge your own company simply no longer has.
How does brain drain differ from turnover?+
The turnover rate puts departures in relation to the average headcount. For brain drain a second metric works better: the number of key roles without a backup. Report both side by side, so leadership also sees the risk that no resignation has triggered yet.
How do I calculate the cost of knowledge loss?+
Start by working through a single role with every cost item: replacement, colleagues' extra search time, rework and customer risk, for the whole period until the successor is genuinely up to speed. The inputs come from your own data: salaries from payroll, search time from a short survey of the team. Only then is it worth extrapolating to the whole year. The Brain-Drain Calculator at calculator.reco.works gives a quick estimate.
When should knowledge transfer before retirement start?+
As soon as the leaving date is set, not at handover. The Fraunhofer guide on knowledge retention names no fixed lead time, but notes that more preparation time allows a more extensive transfer and recommends a generous transition period with the position staffed twice. If partial retirement has been agreed, its start is a natural trigger, because the person will be around less from then on anyway. A notice period alone is practically never enough to pass on experience.
Can AI prevent brain drain?+
Not prevent, but soften considerably. AI can document conversations and make knowledge searchable, but it only captures what someone says out loud. Knowledge that never comes up in a meeting still needs targeted interviews with the experts. And it needs a GDPR-compliant framework, including works council involvement.
Related articles
In a 30 minute software demo we show you how RECO secures and transfers knowledge in your company. Request a demo now and we'll get back to you with a proposed time.
