Why Project Managers Get Pulled Toward the “Loudest Risks”

PMBOK is packed with practical wisdom for advancing project management. Yet when you actually try to apply it, you often find yourself wondering: “Where’s the best place to start?”

Risk assessment is one such area. While the parameters for evaluation are provided, the question of which order to use them in, and how to combine and organize them, remains unclear.

As one practical example, I’d like to introduce a “4-Step Risk Issue Identification” framework. Through a story set at a fictional medical device manufacturer, I’ll explain how to organize risks and where to concentrate limited resources.

Project sites are places where diverse risks constantly intersect. The tricky part is that the risks spoken about most loudly are not necessarily the most important ones. Reaching consensus on “where to start” alone requires considerable effort.

Imagine a transition from research and development to commercial manufacturing. The development team pursues “peak performance,” while the manufacturing team prioritizes “high productivity and reproducibility” — their working styles and evaluation criteria are fundamentally different. Add language barriers and differences in business culture, and the situation grows even more complex.

In such an environment, the biggest trap PMs tend to fall into is reflexively pouring resources into “risks that are making the most noise but actually have little impact.” Fighting fires around localized panic may bring temporary relief. But from a whole-project perspective, it isn’t always the best choice. To optimally allocate limited budgets, time, and human resources, you need a logical way to sort which risks to address.

In this article, I’ll introduce a “4-Step Risk Issue Identification” framework informed by PMBOK’s qualitative risk analysis. Through a case study set at a fictional medical device manufacturer, I’ll offer a practical explanation of how to organize intertwined risks and where to concentrate limited resources.

The 4-Step Risk Issue Identification Framework

The four steps are not independent tasks — they form a single continuous thought process. You cast a wide net in the first step, then narrow your focus in subsequent steps, ultimately arriving at a single point: “this is where we should concentrate.”

Step 1: Finding Blind Spots (Invisibility × Empathy) Comprehensively surface risks that may occur in the future.

Step 2: Assessing Value (Impact × Probability) Ask whether a risk is truly worth solving.

Step 3: Assessing Timing (Proximity × Urgency) Decide when to deploy resources.

Step 4: Identifying the Critical Point (Controllability × Cascading Impact) Anticipate side effects of countermeasures, and select an action your team can lead without harming other areas.

Case Study: Precision Corp.’s Commercial Manufacturing Transition Project

Precision Corp. is a mid-sized medical device manufacturer headquartered in Tokyo, focused on product development in the cardiovascular disease space, with approximately 500 employees. At the center of this project is an implantable device that detects and transmits signs of heart failure in real time — positioned as the company’s next-generation flagship product.

After several years of development, trial manufacturing in the development lab had cleared all performance tests with a yield rate of 98%, and the transition to commercial manufacturing was about to begin. Manufacturing would be outsourced to a U.S. contract manufacturing organization (CMO), Atlas Manufacturing, with contract negotiations already complete.

Key Players:

• Tanaka (PM / Precision Corp.): A PM with domestic product launch experience
• Nakamura (Development Lead / Precision Corp.): The architect of the 98% achievement; highly confident
• Ito (Procurement / Precision Corp.): Experienced in domestic procurement
• Rachel (Manufacturing / Atlas): A professional in mass production lines
• David (Quality Assurance / Atlas): Experienced, but not strong on forecasting
• Karen (External Consultant): A specialist in commercial manufacturing launches for medical devices

Step 1: Finding Blind Spots (Invisibility × Empathy)

Overview: The first step is to comprehensively surface “risks that may occur in the future.” At this stage, resist the urge to narrow things down — what matters is assembling a long list of risk items broadly. In doing so, pay conscious attention to risks that are difficult to notice in day-to-day operations (high invisibility) and about which stakeholders feel little sense of ownership (low empathy).

By holding this perspective, you can also surface “submarine risks” — those that lurk below the waterline and suddenly emerge one day. When necessary, it’s also effective to borrow the perspective of members outside the immediate team or external experts. Assembling a broad long list forms the foundation for the comprehensive evaluations that follow in Steps 2 and onward.

Story: The kickoff meeting for the commercial manufacturing transition project. A screen connecting Tokyo and Atlas was displayed in the conference room, with members from both Precision Corp. and Atlas gathered together. The mood was confident. A 98% yield rate, all performance tests cleared — an air of “with results like these, commercial manufacturing won’t be a problem” pervaded the entire team.

Tanaka, newly appointed as PM, arranged a risk identification session with consultant Karen. Nakamura, the development lead, spoke first.

Nakamura: “The technology transfer schedule is tight, but given the results from the development phase, the rollout to Atlas should proceed without issue. Our yield rate is 98%, and all performance tests have been cleared. Honestly, I’m not particularly worried.”

Karen: “Let me ask a few clarifying questions. That 98% yield rate — under what environmental conditions, by whom, and at what scale was it achieved?”

Nakamura: “It was a number from the development lab. Our team assembled small lots in a temperature- and humidity-controlled environment.”

Karen: “I see. So do you believe those same conditions can be replicated on Atlas’s commercial production line? Temperature and humidity fluctuations throughout the day, variability in operator skill levels, lot-to-lot variation in raw materials — these are constants in commercial manufacturing, aren’t they?”

Nakamura: “…We haven’t examined it that granularly, but I don’t think there will be a big difference. Atlas is also an experienced manufacturer.”

Tanaka listened in silence. There was no ill intent in Nakamura’s words. But it bothered him that the rationale behind “I don’t think there will be a big difference” was based on experience and trust rather than data.

Karen: “David, have you evaluated the reproducibility of this product on Atlas’s commercial production line?”

David: “…Honestly, we hadn’t evaluated reproducibility in the commercial production environment. We had been designing the line based on the development phase numbers.”

A silence followed. Karen continued, not in an accusatory tone, but gently.

Karen: “I’m not looking to assign blame. But the fact that something passed in development is not evidence that it will pass in commercial production. That 98% figure is a result produced under specific conditions. What happens to that number when those conditions change — that’s something nobody has confirmed yet.”

Nakamura: “…So you’re saying the 98% is not a guarantee in a commercial manufacturing environment?”

Karen: “Exactly. And I believe that’s the risk this project needs to confront first.”

Tanaka wrote a single phrase in his notebook: “Past success is silencing the questions.” The session that followed had a far more cautious tone than the opening atmosphere. Voices began to rise one after another: “That hasn’t been confirmed either,” “That assumption is also shaky,” and the risks surrounding the project were written out as a long list of approximately 30 items.

Analysis: The concern about reproducibility in the commercial manufacturing environment wasn’t discovered from scratch. Everyone involved knew somewhere in the back of their minds that development and commercial manufacturing conditions differ. Yet nobody spoke up — for two reasons.

Invisibility: What happens in a commercial manufacturing environment doesn’t appear as numbers until you actually run it. The “98%” right in front of them felt more real than risks yet unseen.

Low Empathy: Nakamura’s attention was already shifting to the next topic, and David was accustomed to a style of waiting for problems to surface. Each had unconsciously placed this risk outside their sense of personal ownership.

Step 2: Assessing Value (Impact × Probability)

Overview: This step asks whether a now-visible risk is truly worth solving. The two evaluation axes are: the magnitude of damage it would cause to the project or product quality (impact), and how likely it is to occur given the current situation (probability). The severity of a risk is objectively calculated by multiplying these two axes.

What you must be especially conscious of here is not overlooking “tail risks” — those with low probability but catastrophic consequences if they occur: regulatory changes, supply chain disruptions, reputational damage, and the like. For this reason, it’s important to establish a policy alongside the scoring: not just “retain risks ranked highest by score,” but also “retain all items with high impact, regardless of probability.”

Story: With 30 items on the long list in front of them, a risk evaluation workshop was held. The task was to assign impact and probability scores to each risk and narrow down the priorities. The discussion moved along smoothly. Technology transfer incompleteness, delays in validation planning, divergence in interpretations of quality standards — each risk received a score, and the list took shape.

Then the discussion stopped at one item: the risk that a specific electronic component central to the device could only be procured from a single supplier.

Ito: “There’s never been a supply interruption before, and our relationship with that supplier is solid. I think the probability is low. Looking at the scores, the priority isn’t high — I think we could lower it for this round.”

Several people nodded. “If the probability is low, lower the priority” — that logic was consistent with the scoring methodology. The room was leaning in that direction when Karen spoke.

Karen: “Wait a moment. Ito-san, what happens if that supplier’s supply is interrupted?”

Ito: “…We wouldn’t find an alternative supplier quickly. Without that component, manufacturing comes to a complete stop. It might take several months to resume.”

Karen: “If a several-month manufacturing stoppage occurred, what would that mean for the project?”

Ito: “…The launch schedule would collapse. The timing for market entry, the commitments to customers — all of it would fall apart.”

Karen: “So the impact would be maximum. Now let me ask: why was the score low? Was it because the impact is low, or because the probability is low?”

Ito was quiet for a moment. “…Because the probability is low.”

Karen: “Exactly. Maximum impact, low probability. These are two separate axes. A low probability does not reduce the damage if it occurs. And the fact that it hasn’t happened before is not a reason it won’t happen in the future.”

David added quietly: “A manufacturing stoppage for a medical device isn’t just a delivery problem. It also means the supply to patients is interrupted.”

Tanaka looked around at everyone’s faces, then spoke clearly.

Tanaka: “This risk stays on the list. Anything with maximum impact stays on the list regardless of probability. That’s our rule. Low probability and ‘safe to ignore’ are two different things.”

Ito raised no further objection. But his expression held a mix of acceptance and something he seemed to recognize — as he later confided to Tanaka, “I’d made similar calls several times before. I’d been telling myself it would be fine because the probability was low.”

The evaluation session narrowed 30 items down to 8 priority risk candidates. Alongside the high-scoring “incompleteness of process technology transfer,” the single-source procurement risk for specific electronic components was retained as a tail risk.

Analysis: The essence of this step is not “mechanically narrowing down by score” but rather “agreeing in advance on the criteria for which risks to retain.”

Scoring by multiplying impact and probability is an effective method for objectively visualizing risk priorities. However, relying solely on scores creates a structural blind spot: the dropout of tail risks.

Risks with low probability tend to receive low scores and are easily dismissed as “not a big deal.” But in a medical device project, supply chain disruptions or regulatory changes — once they occur — can lead to irreversible consequences: manufacturing shutdowns, reapplications, market withdrawals. Low probability is not a reason they won’t happen, and the fact that they haven’t happened before does not guarantee they won’t in the future.

Step 3: Assessing Timing (Proximity × Urgency)

Overview: Once you’ve narrowed down the risks worth addressing, the next question is when to deploy resources. What matters is treating the timing of actual harm (proximity) and the deadline for implementing effective countermeasures (urgency) as separate, independent axes.

This surfaces “hidden urgent risks” where the impact may be distant but design changes or budget approvals mean action is needed immediately — and enables drawing a timeline based on overall optimization. By treating these two axes independently, you can structurally prevent the judgment error of missing the window for action by being misled solely by whether the harm is near or far.

Story: The session on “when to act” began. With the 8-item priority list from Step 2 in front of them, the team worked through the response timing for each risk one by one.

The discussion moved steadily forward, but Tanaka paused at one item. Procurement of the specific electronic component central to the device — separate from the single-source risk retained in Step 2 as a tail risk, there was the problem of securing the initial inventory needed for the commercial manufacturing launch.

Tanaka: “Ito-san, what’s the current status of initial inventory for that component?”

Ito: “I was planning to place the order once the commercial manufacturing start was formally decided. It’s still a ways off, and I thought it would be better to act once the inventory plan was finalized, to avoid waste.”

Karen: “What’s the lead time from ordering to receiving that component?”

Ito: “…The manufacturer-stated lead time is 16 weeks. But the minimum order quantity is quite large, so I was thinking of moving after the inventory plan was confirmed together with the order quantity.”

Tanaka: “How many weeks are there from now until commercial manufacturing starts?”

Ito opened the schedule on his desk. As he checked the screen, his expression slowly clouded.

Ito: “…18 weeks.”

Tanaka: “So even if we order today, the parts won’t arrive until 2 weeks before manufacturing starts. If we wait for the inventory plan to be finalized, it’ll be even later.”

Ito: “…That’s correct.”

The room went quiet. Karen continued gently.

Karen: “When is the inventory plan expected to be finalized?”

Ito: “That’s the thing…we need the production planning team at Atlas to finalize it, we’ve made the request, but we haven’t received a response yet.”

Tanaka: “What’s preventing Atlas from producing a production plan?”

Rachel answered from the screen.

Rachel: “We were operating on a sequence where we’d build the plan after receiving the official demand forecast from Precision Corp. Without the demand forecast, we can’t establish the basis for production volume.”

Tanaka: “Who is responsible for producing that demand forecast?”

A silence followed. Ito named a person from the sales department, but that person was not a member of the commercial manufacturing transition project team. The role of producing the demand forecast had been left floating outside the project, and time had been quietly passing without anyone noticing.

Karen: “Ito-san, the reason you felt it was ‘still a ways off’ was because the inventory plan wasn’t finalized, right? But the reason the inventory plan isn’t finalized is that the demand forecast hasn’t arrived — and the reason the demand forecast hasn’t arrived is that no one has been assigned to produce it.”

Ito: “…That’s exactly what it comes down to.”

Tanaka looked around at everyone, then stated clearly.

Tanaka: “I’m making this today’s top priority. I’ll personally take responsibility for getting the demand forecast from the sales department. We’ll hand it to Atlas to finalize the inventory plan, and I’ll make the ordering decision by the end of this week.”

Rachel gave a small nod. Ito replied “Understood,” and his expression showed something like relief — mixed with the awkwardness of realizing that a problem he’d thought was “still a ways off” was actually one that would be too late to address if they didn’t act immediately, and that he hadn’t seen it himself.

Analysis: The essence of this step is treating “proximity” and “urgency” as separate, independent axes.

In many projects, the implicit assumption that “it’s still a ways off, so we don’t need to act now” is made without examination. But as this case illustrates, the timing at which actual harm materializes (proximity) and the deadline for implementing effective countermeasures (urgency) do not necessarily coincide.

On the surface, the component procurement issue looked like a simple matter of having inventory ready before manufacturing starts. But working backward from the 16-week procurement lead time revealed the structure: a decision had to be made to order immediately, or it would be too late. A risk that, seen through the lens of proximity alone, seemed to have “plenty of time,” was transformed by adding the axis of urgency into “the top priority requiring immediate action.”

Also worth noting is why the decision was delayed. The reason Ito couldn’t act wasn’t negligence or a technical barrier. A chain of conditions — “after the inventory plan is decided,” “after the demand forecast arrives” — meant no one could take the first step. Because it was unclear who should be making the decision, time simply passed quietly.

Step 4: Identifying the Critical Point (Controllability × Cascading Impact)

Overview: The final step is finding the “critical point” with the highest return on investment among the top-priority risks. The two evaluation axes are: whether the team can control it with their own authority and capabilities (controllability), and whether the countermeasure will cause adverse effects on other elements (cascading impact).

Countermeasures for risks can themselves create problems. Trying to solve one problem can place new burdens on quality, schedule, cost, or regulatory compliance — the “side effects of countermeasures.” Identifying these side effects in advance is the essence of this step. Countermeasures with high controllability and small cascading impact are the ones a team can invest in with confidence.

Story: The day after Step 3’s session ended and Tanaka began moving on the ordering issue, a report arrived from the inspection equipment manufacturer. It stated that “the current inspection algorithm may exceed the acceptable range for false positive rates at the processing speed of Atlas’s commercial production line.”

For an implantable cardiac monitoring device, the reliability of the inspection process is directly linked to product safety. It was not a problem that could be overlooked. Tanaka assembled the team that same day.

Rachel: “Shouldn’t the proper course of action be to ask the manufacturer to revise the algorithm? It would be a fundamental fix, and it’s more reliable than us making modifications ourselves.”

Karen: “Who can control when the revision is completed?”

Rachel paused for a moment.

Rachel: “…That would be up to the manufacturer. We can’t predict when they’ll finish.”

Karen: “Is there any flexibility in the commercial manufacturing start schedule?”

Rachel: “That’s difficult.”

Tanaka listened in silence. Requesting a revision from the manufacturer meant placing the resolution of the problem somewhere beyond the team’s reach. Completion timing and quality would both depend on the manufacturer’s roadmap. Given that the schedule couldn’t slip, they couldn’t bet on a countermeasure they couldn’t control.

Next to speak was David.

David: “If we lower the processing speed of the commercial production line, it should fall within the acceptable range with the current algorithm. We can implement it ourselves immediately, and we don’t need to wait for the manufacturer.”

Tanaka: “If we lower the processing speed, how does that affect production capacity?”

David: “…I’d need to calculate it, but the number of units per day would drop considerably.”

Tanaka: “And the schedule?”

David: “There could be an impact.”

Tanaka: “And cost?”

David: “Fixed costs would be spread over fewer units, so the unit cost would rise.”

Tanaka noted this down. Lowering the processing speed was something the team could implement immediately. But the trade-off was simultaneously placing a burden on three elements: production capacity, schedule, and cost. A countermeasure that solved one problem while creating three others.

Nakamura leaned forward slightly.

Nakamura: “Both requesting the revision and lowering the processing speed involve sacrificing something, right? In the first place — what was the basis for the concern the manufacturer raised?”

Karen: “Good question. Rachel, what did the report say?”

Rachel: “…It says the evaluation was based on ‘estimated values’ of Atlas’s commercial production line processing speed.”

Karen: “Estimated values.”

Tanaka stopped writing. The manufacturer had raised the concern based not on actual measured values, but on estimated values. That being the case, no one had yet confirmed whether that premise was even correct.

Tanaka: “Rachel, is it possible to actually measure the processing speed on Atlas’s line and give that data to the manufacturer?”

Rachel thought for a moment before answering.

Rachel: “We can do the measurement ourselves. We could have actual measured values within a week.”

Tanaka: “We share those measured values with the manufacturer and ask them to evaluate again. If the actual values are better than estimated, the concern itself may be resolved. If they’re worse, we can have a more precise discussion about revisions based on accurate numbers. Either way, we’re on more solid ground than we are now.”

Karen followed up.

Karen: “Let’s check the cascading impact of this countermeasure. Does production capacity drop?”

Tanaka: “No.”

Karen: “Schedule?”

Tanaka: “No impact.”

Karen: “Cost?”

Tanaka: “Unchanged.”

Karen: “Validation plan?”

Tanaka: “Having actual measured values will actually improve accuracy. The conditions of the commercial production environment will be redefined to reflect reality, which also becomes usable for evaluating other environment-dependent risks.”

Karen nodded.

Karen: “And controllability?”

Tanaka: “The measurement can be completed entirely by Atlas’s team. No need to wait for the manufacturer, no need to halt the line.”

Karen: “So why do you think this option didn’t come up first?”

A silence followed. The first to speak was Nakamura.

Nakamura: “When the report arrived from the manufacturer, we started thinking on the assumption that the concern was valid. Before questioning whether it was based on estimates or actual measurements, our minds went straight to how to respond.”

David: “It’s the same as Step 1. The structure is similar to how we didn’t question the 98% from the development lab.”

Tanaka wrote in his notebook: “We were rushing to respond before questioning the premise.”

Tanaka: “Please proceed with the measurement. Share the results as soon as they’re available.”

Rachel gave a quiet nod.

Analysis: The essence of this step is not asking “how do we solve this?” but asking “will this way of solving it break something else?”

Countermeasures for risks can themselves become risks. In this case, the two options initially raised — requesting a revision from the manufacturer and lowering the processing speed — both involved solving one problem while simultaneously creating another.

Requesting the revision meant surrendering control of the schedule to an outside party, leaving completion timing and quality dependent on their roadmap. Lowering the processing speed placed cascading burdens on production capacity, schedule, and cost. The more countermeasures were applied, the more other parts of the project were damaged.

The two axes of controllability and cascading impact provide coordinates for making this structure visible. A countermeasure with low controllability hands the key to resolution to an external party — completion timing and quality both depend on others, and the project’s autonomy is compromised. A countermeasure with large cascading impact risks generating more problems than it solves. The critical point lies at the intersection of these two axes: an action the team can lead, and that does not harm other elements.

The move Tanaka arrived at — obtaining actual measured values and asking the manufacturer to re-evaluate — was not merely a solution to the problem, but a re-examination of the premise of the problem itself. Recognizing the fact that the manufacturer’s concern was based on “estimated values,” the simple act of “taking actual measurements” emerged as the countermeasure with the highest controllability and smallest cascading impact.

Worth noting is why this option didn’t surface first. As David pointed out, the moment the report arrived from the manufacturer, the team had unconsciously accepted the premise that the concern was valid, and directed their thinking toward how to respond. They were rushing to respond before questioning the premise. This structure is identical to how the team in Step 1 didn’t question the 98% from the development lab.

Step 4 is the culmination of the 4-step framework, and simultaneously the step that interrogates the quality of countermeasures themselves. “Doing something” and “doing the right thing” are two different questions. Having the foresight to anticipate the side effects of countermeasures and the eye to question premises — these, I believe, are among the most sophisticated judgments required of a project manager.

Conclusion

What the four steps share is the attitude of “questioning, once, what you think you can see.”

The 98% yield rate, the low probability, the distant-seeming schedule, the concern report from an external party — in each case, the initial appearance became the premise for the decision. What Karen repeatedly asked was not difficult expertise, but a simple question: “What is your basis for that?”

Risk management is not about listing risks. It is a continuous series of judgments — persistently questioning what lies beneath what is visible, and deciding where to concentrate limited resources. PMBOK’s framework provides the coordinate axes, but they only function through the quality of the questions asked by the people who use them.

“Past success is silencing the questions.”

This phrase reflects how difficult it is to keep asking questions precisely when things are going well. And that, perhaps, is the most important disposition required of a project manager.