The Capability Edge — September 2026

The Cost of Training Is Visible. The Cost of Insufficient Capability Often Is Not.

When organizations begin tightening budgets, professional development is frequently treated as a discretionary expense. The reasoning is understandable: training has a visible and immediate cost, while its financial return can be difficult to isolate. In medical device development, however, that framing is incomplete. The more useful economic question is not simply what it costs to train an engineer. It is what an organization spends correcting problems that stronger technical capability could have prevented in the first place.

Ben Rand recently examined a related issue in the Harvard Business Review article "Why Training Employees Pays Off Twice." The research discussed in the article found that employees who received targeted training subsequently completed approximately 10% more work. Importantly, the benefit was not confined to those employees. They also required less assistance from their managers, allowing managerial capacity to be redirected elsewhere. The result was a second-order productivity benefit that conventional calculations of training ROI could easily overlook.

That finding resonates with what I have observed throughout my career. Technical capability does not affect only the efficiency of an individual engineer. It influences the thoroughness of the requirements, the strength of design decisions, the appropriateness of design verification strategies, the quality of technical documentation, and ultimately, the amount of corrective work required downstream. An engineering decision made with incomplete knowledge may require only a few hours to make. Discovering that the decision was inadequate after design verification, design validation, design transfer, or regulatory review can require substantially more time and resources to correct.

Rework Is an Engineering Capability Problem

In The Business Case for Technical Training in Medical Device Development, I examined this issue through the economics of rework. Published estimates suggest that rework can account for approximately 40–50% of product-development costs, with a substantial portion originating in requirements and specification problems.

This is important because rework is not merely an accounting category. It represents engineering capacity that has already been purchased but must now be purchased again. Engineers revisit requirements, redesign components, repeat analyses, execute additional design verification, revise documentation, or respond to deficiencies that could potentially have been avoided through more sound decisions earlier in the development process.

The economics become clearer when we model the problem. Consider an illustrative 20-person technical organization with a fully loaded annual cost of $150,000 per employee. Using conservative assumptions, including a 35% rework share and a 20% reduction in that rework through improved technical capability, the resulting reduction in rework is approximately $210,000 annually. Against an illustrative $75,000 training investment, that represents approximately a four-month payback period and a first-year return of roughly 180% from rework reduction alone.

These figures are not intended to suggest that training mechanically produces a predetermined financial return. Organizations differ significantly in product complexity, engineering maturity, labor costs, development processes, and existing capability. The appropriate exercise is for an organization to substitute its own assumptions. The important point is that technical development can be evaluated as an engineering investment rather than simply an HR expense.

The Cost Extends Beyond Rework

For medical device manufacturers, there is another variable that deserves consideration: time.

Approximately two-thirds of first-cycle FDA 510(k) submissions receive an Additional Information request. A regulatory delay has consequences beyond the additional engineering hours required to prepare a response. When technical deficiencies extend a development or submission timeline, the organization may also postpone commercialization and the revenue associated with market entry.

When potential time-to-market benefits are incorporated into the illustrative model in my white paper, the economics become substantially larger: approximately $1.4 million in potential annual benefit against a $75,000 training investment. Again, the exact number is organization-specific. The important issue is identifying all of the mechanisms through which technical capability affects economic performance.

What I Heard at MEDevice Boston

I recently had an opportunity to test this reasoning against conversations at MEDevice Boston. One of my more interesting observations was how consistently the argument resonated, particularly among smaller medical device organizations and technical teams. These organizations often do not have the internal specialization available to a large manufacturer. Engineers are required to make consequential decisions across a broader range of responsibilities, while operating with limited time and resources.

Self-paced technical development was particularly relevant in those conversations. Traditional training can require travel, several days away from active projects, and significant scheduling coordination. Asynchronous professional development reduces some of those constraints while allowing engineers to develop capability around their existing responsibilities. More importantly, it preserves investment in engineering capability at precisely the time when organizations are under pressure to become more efficient.

That brings me back to the argument raised in Harvard Business Review. The return on training should not be measured only by asking whether the person who completed the training became more productive. In an engineering organization, we should also ask whether better capability reduces the downstream demand placed on colleagues, managers, quality teams, regulatory teams, and future project resources.

For medical device organizations evaluating budgets, I would therefore frame the question differently.

Rather than asking what technical training costs, determine what insufficient technical capability is already costing the organization.

That calculation may produce a very different answer.

Run Your Own Numbers

I built the model above as a spreadsheet. Eight inputs — headcount, loaded cost, rework share, cost of delay. Change them to your own figures and the payback recalculates. Assumptions and sources are documented on a second tab.

The edge isn't knowing the regulation. It's implementing it.

— Dr. John Salvato Founder & President, Design Quality Services