Knowledge management in the company: why most systems fail and what works instead

Knowledge management in a company means the practice of systematically capturing, structuring and making findable what an organisation knows, so that it does not stay tied to individual people. As a management discipline in its own right it is described in the international standard ISO 30401, adopted in German speaking markets as DIN ISO 30401. It is not about filing documents. It is about handling the things that actually make the difference: client histories, the reasons behind decisions, pricing logic, technical tricks, the unwritten how. In most companies that knowledge lives in people's heads, in old email threads and in chats nobody can find again. A 2018 survey by Kyocera and Statista shows how tightly it is bound to individuals: 38 percent of employees in Germany and Austria said much or even all knowledge would be lost if they left the company without a handover.

The topic is gaining urgency for a demographic reason. According to the Institut der deutschen Wirtschaft (IW Köln, IW-Kurzbericht no. 78, October 2024), around 19.5 million baby boomers in Germany will reach retirement age by 2036, while only about 12.5 million younger people follow behind them. Every one of those retirements takes experience that is rarely written down. The awareness is there, the execution is not: according to Deloitte's Global Human Capital Trends (2020), 75 percent of organisations consider creating and preserving knowledge important, but only 9 percent feel very ready to do it. You can roughly estimate what a single departure costs with our knowledge loss calculator; the visible costs are only the smaller part, as we break down in What turnover really costs.

The reason so many knowledge management projects run into the sand anyway is almost always the same. Classic systems, whether a wiki, Confluence or a SharePoint folder, require people to document on top of their actual work. Documentation becomes an unpaid second job, the first thing dropped when things get tight. Almost everyone knows the result: a knowledge base that is half out of date and that nobody searches, because it already seems clear the answer will not be in there. That information is hard to find is not just a feeling: according to Gartner (2023), 47 percent of digital workers struggle to find the information they need to do their jobs. The problem is not the software. The problem is the assumption that people will document voluntarily and consistently.

What has changed is where knowledge can be captured. Most of the relevant knowledge is not created while writing, it is created in conversation: in the client meeting, the handover, the project review, the performance conversation. When those conversations are transcribed, structured by topic and made searchable with their sources, the knowledge base emerges alongside normal work, as a by-product of it. That is where RECO starts: people ask in natural language and get grounded answers from their own company instead of generic AI noise. What to look for when choosing such a tool is summarised in The best AI tool for knowledge management.

For mid sized companies of 50 to 500 employees the entry point is smaller than it sounds. A workable approach usually looks like this:

1. Start with one type of knowledge, not the whole company. Handovers and recurring client meetings deliver the fastest visible benefit.
2. Capture knowledge where it is created instead of maintaining a separate wiki. What has to happen on top of the work does not happen.
3. Sort out data protection and co-determination early. In Germany the works council must be involved, in Austria a works agreement is required; a one page summary of purpose, access and retention resolves most concerns.
4. Make the result searchable and its origin traceable. An answer without a source does not get used.
5. Check after eight weeks whether the content is actually searched and found. If not, it is rarely the technology and usually the scope or the access.

Knowledge management in a company is therefore less a question of tooling than of whether documentation happens inside daily work or beside it. Make it a by-product and you retain experience before the next resignation or retirement carries it out the door. For a structured start, the freely available DIN SPEC 91443 offers a guideline tailored to SMEs. More on this in the topic hub Knowledge retention and turnover and under Knowledge transfer software for SMEs.