When we launched the "Employee Journey" project, our goal was not to cut training at any cost or reduce the volume of documentation. We wanted to understand something else: what the entire flow of safety information looks like through the eyes of the employee – from the moment of hire to authorization for independent work.
At one of Company N's facilities, we attempted to literally walk this path alongside a new hire. We observed what happens on their first workday, what safety inductions they receive, what they are taught, what materials they are given, what they study during on-the-job training, and what is required of them during the knowledge assessment.
Taken individually, each step seemed completely logical. Every document had a clear reason for being created, every briefing had its own purpose, and every specialist had their own area of responsibility. The problem only became visible when we assembled everything into a single sequence.
Initially, we didn't change anything.
It was essential for us to see the existing process as it really was: not according to a flowchart or standard operating procedures, but through the eyes of someone stepping onto the facility for the very first time.
We systematically deconstructed the journey from hiring to independent work: hiring → first workday → introductory orientation → workplace safety induction → training → on-the-job training → knowledge assessment → work authorization.
For each stage, we examined: what the employee is told; what materials they receive; what they are required to study; what gets repeated in subsequent stages; what is tested; and what information they actually need to use directly on the job.
In parallel, we observed how specific onboarding stages took place in reality, talked with employees and managers, and conducted interviews and a focus group.
In the end, our research dataset totaled 2,069 pages of documents and materials.
To be clear: this does not mean a single employee is expected to read all 2,069 pages sequentially. This represents the cumulative volume of the information environment we analyzed to understand which sources shape employee training.
According to our calculations, reading the entire dataset sequentially would take approximately 115 hours.
This figure alone does not prove anything. A large volume of information does not automatically mean the system is bad.
Therefore, we were interested in more than just volume. We began examining how information is distributed over time, where it repeats, at what point the employee actually needs it, and what remains retained after completing all mandatory stages.
One of the first things we observed was semantic duplication.
The exact same topics recurred across different stages of preparation—sometimes nearly identical, sometimes from a slightly different angle, and sometimes embedded within another module.
The reason is understandable. Each department is responsible for its own domain and tries to convey its related requirements as thoroughly as possible. Viewed in isolation, each information stream makes perfect sense.
However, the worker does not experience isolated streams. They receive everything all at once.
This is particularly evident during the first few days. A person is adjusting to a new facility, a new supervisor, upcoming job tasks, personal protective equipment, and internal company policies. Simultaneously, they must absorb safety requirements, memorize emergency routes and procedures, learn operational rules, and digest dozens of other unfamiliar details.
Thus, the worker's task at this stage is far more complex than simply "listening to information." They need to discern which parts apply directly to them, what is critical right now, what will be needed later, and how to react in a specific operational scenario.
Here, our findings aligned directly with research on information overload.
(A 2024 research review on the causes and consequences of information overload demonstrated that overload is not caused solely by the volume of information. It is influenced by the complexity of the material, the number of sources, task characteristics, and the organization of the information flow itself. In studies, information overload was associated with increased cognitive strain, reduced productivity, and impaired decision-making quality. The authors analyzed publications from major scientific databases, including Scopus, WebofScience and PubMed.)
This was a key insight for us. The issue may not always be that an employee is given too much information in absolute terms. Sometimes the problem is that too much information demands their attention simultaneously.
This is especially critical for someone who is not yet familiar with the production environment.
An experienced worker perceives much of the workplace as a familiar picture. They see equipment and immediately recognize where potential hazards lie. They know the sequence of operations and understand which sounds or movements warrant attention. A newcomer must process all of this piece by piece.
Here, we relied on cognitive load theory. Its core principle is quite straightforward: human working memory capacity is limited. When too much new information must be processed at once, some of it is inevitably lost or absorbed only superficially.
This effect is further corroborated by more recent studies on complex professional activities.
(In a 2025 systematic review focusing on cognitive load among emergency medical services professionals, 25 studies with a combined sample of 6,284 participants were analyzed. The authors demonstrated that cognitive load increases with task complexity, the number of simultaneous factors, and situational novelty. In several studies, elevated cognitive load was accompanied by degraded task performance and a higher error rate. While emergency medical teams and industrial manufacturing operate in different environments, the findings cannot be directly transferred. What matters to us is the proven mechanism of limited attention during complex, potentially hazardous activities.)
Following this, we began to view the problem statement differently.
Typically, when training an employee, it is easy to ask: "Have we told them everything they need to know?"
However, that question proved insufficient.
A second question emerged: "Is the person actually capable of absorbing all this and applying it later?"
And these two questions rarely yield the same answer.
This distinction became particularly evident during discussions with workers. We asked what safety information they recall first. They most frequently cited specific site safety rules, life-saving rules, and actions linked to real operational situations. Large volumes of formal regulatory requirements were recalled far less frequently.
These interviews should not be viewed as a large-scale statistical study. However, they provided a strong qualitative signal: information communicated to an individual is not the same as information they actually use.
Gradually, we began to distinguish between several fundamentally different outcomes:
This final transition proved to be the most interesting to us.
It aligns closely with modern research on safety training.
(A major 2026 meta-analysis combined findings from 157 independent safety training studies comprising 666 distinct measured effect sizes. Researchers observed a strong training effect on direct learning outcomes—knowledge acquisition and material comprehension. A positive but smaller effect was observed regarding the transfer of learned skills to actual work. The impact on long-term safety performance indicators was smaller still.)
For us, this conclusion was highly practical: training may be conducted well and an employee may pass the knowledge assessment, but that still does not guarantee the desired safety behavior on the shop floor.
At some point, we realized we were initiating the project with the wrong question.
Instead of the usual query: "What else do we need to tell the employee?" – we began asking:
What specific situation will they encounter on the job?
What could go wrong in this situation?
What hazards or cues must they notice?
What decision must they make?
What do they truly need to know in advance?
What must be practiced hands-on?
And what information is pointless to memorize long-term—where knowing how to find it quickly is sufficient?
After asking these questions, our core approach began to shift.
We stopped categorizing information solely by departmental ownership ("which department manages what"). Instead, we attempted to map it along the employee journey and align it with the exact scenarios where it will actually be required.
Certain information is essential on day one. Other topics are best discussed immediately prior to executing a specific task. Some concepts cannot be mastered without hands-on practice. And certain regulatory details are best revisited later, once the employee has gained operational context and understands the bigger picture.
A separate question is how long knowledge is retained post-training.
Test scores immediately following a safety briefing may look promising, but several weeks later, the picture can change significantly.
(A 2026 systematic review and meta-regression analysis on knowledge retention following safety training incorporated 83 peer-reviewed studies published through late 2024. In these studies, knowledge was evaluated at varying intervals—ranging from a few days to several months. The overall trend demonstrated a decline in knowledge retention over time. The authors emphasized the need for regular refresher training and evaluating not just immediate learning outcomes, but long-term retention stability.)
Therefore, we embedded an additional principle into our project: do not limit evaluation to immediate post-training testing.
As the next step, we selected a target job role and conducted a practical trial of this new framework. 16 employees participated in the pilot study. We compared the legacy training format against the new design, re-evaluating a subset of questions several weeks later to measure knowledge retention.
We will share those findings in our next article.
Perhaps the single most critical insight from the initial phase emerged even before analyzing the pilot numbers.
In occupational safety management, it is tempting to view information volume as an additional layer of protection. The logic seems sound: the more you tell an employee, the more they will know.
Our study proved that this correlation is far less direct.
Every new piece of training material competes for cognitive attention with what has already been absorbed. When the information stream becomes excessively dense, critical safety instructions risk getting lost alongside secondary details.
As a result, the fundamental question for me is no longer: "How much information have we delivered to the worker?"
It is now: "Will this information actually help the worker spot a hazard and make the right decision when it matters most?"
If you cannot answer that question affirmatively, perhaps it is best not to introduce yet another document, but rather to streamline the information the employee is already receiving.