Knowledge Management · KM
Knowledge Management
What is knowledge management?
- Capture and share scattered knowledge and know-how
- Turning personal knowledge into an organizational asset
- The key is converting tacit knowledge into explicit knowledge
Once knowledge management takes hold, the starting point of work changes. In a company where only one person understands the quoting logic, that person's vacation means delayed quotes. When knowledge accumulates in the organization, the question in the same situation shifts from 'Who do I ask?' to 'Where do I find it?' That is why the results of knowledge management show up not in the number of registered documents but in the share of work that ran without the person in charge. What actually gets measured is how long it takes a new hire to make decisions independently, and how much time handovers tie up when someone leaves or transfers.
In practice, it works better to treat repository-type and connection-type knowledge separately. Knowledge with fixed answers, such as manuals, checklists, and troubleshooting cards, should be documented. Knowledge that changes with the situation, such as a feel for negotiation or judgment in handling complaints, is handled more efficiently by connecting people through case review meetings, mentoring, and internal communities than by writing it down. This is also where the line with an LMS falls. If an LMS is a tool that delivers a set curriculum, KM is closer to circulating answers that have just emerged on the job. That is why KM documents need an expiration date. When equipment, policies, or clients change, last year's document becomes not the right answer but the wrong one, so companies that assign each document an update owner (accountability) and a review cycle (typically 6 months or 1 year) are the ones that last.
The most common misconception is that installing a system will make knowledge accumulate. Open a portal, set 'number of entries' as a team KPI, and hundreds of entries pile up within three months, but most are copies of meeting minutes, so searches don't return the answers people need. An employee who comes away empty-handed once goes back to asking a senior colleague on messenger, and the portal turns into a ghost town. A portal that only features success stories is also a warning sign. The knowledge an organization most needs to protect is failure records, such as why a quote was wrong or how a delivery deadline was missed, yet the moment those are tied to performance reviews, no one writes them. Make failure cases anonymous and separate them from evaluations; set that rule first and the documents become honest.
Jihye ChoiExpert in design thinking, new-hire onboarding, and train-the-trainer programs
John KimAuthor of B2B Marketing Design and CEO of Nanum ConsultingIt was formalized in Ikujiro Nonaka's theory of tacit and explicit knowledge (the SECI model), among others, and spread with the rise of the knowledge-based economy.
Let's look at a hypothetical case: Company A, an auto parts manufacturer with 180 employees and KRW 62 billion (≈ US$46,386,400) in annual revenue. Mr. Kim, who had run production engineering for over 20 years, was eight months from mandatory retirement, and whenever equipment stopped overnight, the shift leader would call him even at 2 a.m. Management started organizing knowledge out of a sense of crisis: 'If he leaves, the line stops.'
- When the production technology team lead pulled the equipment stoppage records for the past two years, 217 of the 412 cases had been resolved by Manager Kim in person, more than half. Those 217 cases became the first priority to document.
- Manager Kim flatly said, "Ask me to write a manual and I can't fill a single page." So the team switched from writing to interviews: twice a week for 40 minutes over 12 weeks, they asked questions on the shop floor and filmed the fixes on a smartphone.
- A junior training manager refined those records into 68 three-column cards ('symptom → suspected cause → action steps') and put a QR code on each machine so shift leaders could open the right card on their phones.
- As an operating rule, the team's metric became 'issues solved by staff using a card' rather than number of entries, and any card that went unopened for six months was either retired or rewritten at the monthly review meeting.
- Six months after rollout, average night-shift equipment response time dropped from 47 minutes to 19, and after-hours calls to Manager Kim fell from 14 a month to 2. The time it took a new maintenance technician to be put on the night shift alone also fell from 8 months to 4.