The attendance sheet is there: fourteen names, date, topic, duration ninety minutes. Four weeks later, during a walk-through, we ask three of the signatories what it was about. Two remember the session and not the content. One remembers nothing at all.
Three answers prove nothing yet. They are a warning sign, and one we receive so regularly that it has become the opening question of every consultation on the subject: what exactly is the training here for? A finding about those three individuals it is not.
Four words that do not mean the same thing
A distinction first, without which the rest stays blurred.
Instruction is an employer's duty. On the basis of the risk assessment it explains the specific hazards at the workplace, the protective measures and the conduct expected. Familiarisation acquaints someone with a particular building, item of work equipment or process. Training goes further and conveys professional knowledge and skills. And a drill rehearses the doing of it.
In everyday use the four words are treated as interchangeable, and part of the trouble comes from that. What has to happen before the first deployment as a matter of law gets muddled with what sensibly comes later.
Two purposes that can get in each other's way
Training is supposed to do two things. It is supposed to enable people to do something differently. And it is supposed to evidence that a duty has been discharged.
These are not opposites. Good instruction does both: it enables safe performance, and it documents traceably who was instructed on what. Evidence obligations are real, they appear in occupational safety law, in hygiene plans, in contracts and in certification requirements, and anyone failing to meet them has a problem regardless of the quality of their work.
The trouble arises when a session is optimised for the signature rather than for understanding. In practice that looks like this: the list of topics is made as complete as possible, because everything is supposed to have come up once. It all happens in one sitting, because that is organisationally simpler. And everyone gets the same, because differentiated documentation is more work.
Each of these properties can reduce the effect, if nothing offsets it.
Why the standard format fails
Five reasons that work together in practice.
The timing. Training takes place after the shift, because that is when everyone is there. Somebody who has worked physically for six hours and is then expected to sit for ninety minutes is not receptive. For night shift staff the situation is worse still.
The group size. With fourteen people in a room the format tips easily into a lecture, and then gaps in understanding stay invisible. Anyone who would have to admit in front of the whole group that they did not understand something asks less often, and silence gets read as comprehension. Fourteen people can perfectly well be trained, just not in one block and not by lecture: in stations, in small groups, by demonstrating and copying.
The language. In both sectors a substantial proportion of staff work with German as a second or third language. Technical terms such as alkaline, contact time, escalation level or usage unit are not self-explanatory even for native speakers. Use them in a lecture without showing them and you fail to reach part of the room. With work equipment this is not a question of style: the Industrial Safety Regulation (Betriebssicherheitsverordnung) requires information in a form and language comprehensible to the employees.
The abstraction. Standard training deals with floor coverings in general or escalation routes in general. The person listening, however, does not work in general but on the second floor of the east wing, where there is linoleum and where the door to the server room must never stand open. Nothing about that building appears in the training.
The volume. Ninety minutes across twelve topics gives seven and a half minutes per topic. What gets conveyed in that time is a mention.
Timing matters, but differently from how it is assumed
Putting everything on the first day is the most widespread mistake. Pushing everything back would be the more dangerous one.
Because part of the content has to be conveyed before work begins. Section 12 of the Occupational Safety and Health Act (Arbeitsschutzgesetz) requires adequate and appropriate instruction on recruitment and before taking up altered duties, related to the workplace and the task and given during working hours. For work equipment, the Industrial Safety Regulation requires instruction before first use and at least annually thereafter. Anyone deferring the operating instructions for a hazardous substance until after the first few shifts is economising in the wrong place.
The answer therefore lies not in pushing things back but in staging them across three steps.
Before the first deployment comes what is needed to start safely: immediate hazards, express prohibitions, protective equipment, emergency call and reporting route, safe handling of the products and equipment for the day, the sensitive surfaces or critical areas, access points and contacts. Kept brief, specific and related to the day's task.
After the first few shifts comes the deepening. Now the person has questions, because something has happened. They know which door sticks, which floor behaves oddly and which day of the week the waiting area gets difficult. When that point is reached depends on the building and the risk; after a few days it may be there, in a complex building it takes longer.
Recurring comes everything that changes or stays critical: new products, new equipment, findings from inspections, incidents and near misses, refreshing the procedures that are rarely needed and have to be known in an emergency.
What training can and cannot do
Here is the point at which expectations are pitched too high.
Training can convey knowledge. It can explain why an alkaline cleaner damages a linoleum floor, or in what order reports are made in an incident. That works, where format and timing are right.
Knowledge alone does not change behaviour. That takes practical rehearsal, feedback, supervision and above all the possibility of actually doing it within the workflow. Somebody who knows that the floor in the east wing needs a different product, and would need three minutes more in their round schedule that they do not have, will not change the product regardless.
That is the constraint that beats any training. Where the method taught is not possible in the time costed, it will not be applied, and no format changes that. At that point the training even becomes problematic, because it conceals a structural shortcoming and puts the responsibility on the staff for a problem that lies in the costing.
What works instead
Four properties we see in buildings where knowledge does land.
Short and repeated rather than long and one-off. Short units on a single topic, repeated at intervals and combined with active recall, embed knowledge more durably than a single concentrated session. That corresponds to what learning research shows about spaced practice, and it matches what we observe in buildings. There is no general formula for length and interval, though; both have to suit the content and the risk.
In the building rather than in a training room. In front of the floor covering being discussed. At the door in question. In the cleaning store where the products are kept. The physical context replaces a considerable part of the explanation.
On the matter rather than on the syllabus. The occasion for a short unit is a real incident, a new floor covering, a finding from an inspection. Build on that and you are explaining something somebody is currently interested in.
Shown rather than said. The dosing is demonstrated. The reporting route is rehearsed once, with the telephone in hand. That costs five minutes and replaces half an hour of lecture.
For the evidence obligations this means no sacrifice. Short units can be documented just as well as long ones; there are simply more entries. Whether several short units can substitute for a prescribed recurring instruction depends on the particular legal basis, however. Some require a specific occasion, a coherent body of content or a fixed interval. Check that for each duty rather than summing across the year as a matter of course.
What a building document contributes
Training that refers to nothing written evaporates along with the memory of it. Where a cleaning concept or building-specific operating instructions exist, the training is the explanation of them, and the document stays in the building afterwards.
That also changes the significance of staff turnover. Where the knowledge conveyed in the training sits only in the heads of those present, it leaves with them. Where it is set down in a document that stays in the building and that people are familiarised with, it remains.
The most effective part is the shortest: the list of what expressly must not be used. It can be read in two minutes and answers the most frequent everyday question.
How to establish whether it worked
The question is rarely asked, because the answer might be uncomfortable. It is nonetheless easy to answer.
Walk through the building a few weeks after a training session and ask something specific from the content. Not in the abstract but on the matter: which product do you use on this floor? Whom do you call if something happens here? And ask them to show you, because explaining and being able to do it are two different things.
Three rules go with this. Do not ask in front of the group, because an examination setting produces defensive answers. Treat uncertainty as a reason for immediate re-training rather than as a failing. And evaluate the findings anonymously as a pattern, not by individual.
A correct answer is a first indication and not yet evidence. The picture becomes robust when you additionally see that the method is applied day to day and that errors are declining. If not, it is worth asking why. In most cases it is the constraint and not the format.
What belongs in the tender and the contract
Three formulations make the difference between a training clause that achieves something and one that merely creates an obligation.
First: building-specific rather than general. The obligation should be to train on the surfaces, areas and particularities present in the building, not that training takes place annually.
Second: before the first deployment and recurring thereafter. Require only annual training and, with high turnover, you get staff working for eleven months without familiarisation.
Third: with evidence containing more than an attendance sheet. Topic, content, duration, participants, who delivered it, in what language and with what aids, whether it was practised, and which version of the underlying instructions it was based on. These details are the basis for being able to evidence anything at all in a dispute.
And one question to ask yourself before drafting such a clause: is the time costed sufficient for what is to be trained? If not, change the costing first.