Finishing the course doesn't mean they have the skill

The training dashboard is green. 87% of the team finished the course. The certificate was issued. The slide for the committee is ready.

The next week, the same team can't put the service the business asked for into production. Nobody is lying. The number measured something else.

Finishing a course measures attendance. It doesn't measure whether a person can do the job.

What the completion rate actually measures

The platform that hosts the courses answers, with reasonable precision, these questions:

• who enrolled
• who reached the last module
• who passed the short quiz
• how many hours were logged

Those are operational data points: useful to know whether people saw the material. They say nothing about what each person can do afterwards.

That's training logistics. It serves regulatory compliance, mandatory onboarding, and audits. If the course exists to prove it was delivered, the completion rate is the right indicator.

The problem starts when that same number is used to answer another question: does the team have the capability the business needs?

At that point the completion rate is an indirect signal. And a weak one. A person can click Next, pass a multiple-choice quiz, and still be unable to write the query, review a code change, or explain a failure in a real environment. Another person can skip half the content and already know how to do the work, because the course didn't start from their level.

Treating completion as proof of capability is like measuring a gym by how many people went through the turnstile.

The corporate university label

In several Latin American countries, especially Colombia, "universidad corporativa" is the name companies give their internal training program. It's not a university that grants degrees. It's the company's school: the courses, paths, and certificates the organization defines and maintains.

Companies that don't use that label call it an internal academy, a company university, or simply the training program. The name changes; the structure is the same.

The point of this article doesn't change with the name. If what gets reported is how many people finished the courses, you're still measuring attendance, not whether the person can do the job. The name of the program doesn't turn a completion rate into evidence of capability.

What the CFO and the CTO ask

In the budget meeting, nobody asks for "learning hours". They ask some version of:

• Can we cover this role from inside, or do we have to hire?
• Is the data program working?
• Who is ready for the next level of responsibility?

None of those questions is answered with "87% completed". That percentage says people showed up. It doesn't say whether capability moved, whether it can be applied with the technologies the company actually uses, or whether the change holds four weeks later.

That's why the training team loses the conversation even when the dashboard is flawless. It brings activity. The business asks for evidence of capability.

Three numbers that look like evidence and aren't

The short quiz at the end of the module. A well-designed short quiz can verify immediate recall. It doesn't verify whether the person solves a new problem, with dirty data, in the company's repository. 90% correct on Friday doesn't predict Monday.

The hours. Time on the platform is exposure, not progression. Ten hours of Kubernetes video isn't the same as standing up a cluster and recovering it when it breaks.

The certificate. The certificate confirms that a completion criterion was met. In most corporate programs, that criterion is having gone through the content. It's not a job evaluation.

All three are easy to pull from the platform. That's why they survive. Not because they answer the question.

What actually counts

Evidence of capability is harder to report and more useful:

1. A starting point. A technical, practical assessment — by peers or by the manager — before building the program. Without a starting point there is no progress, only activity.
2. A destination tied to a role or to the company's technologies. Not "learn Python". "Be able to do X with the technologies we use, at the level the role requires." That's a development plan per role, not a course catalog.
3. Observed practice. Exercises, practice environments, code changes, simulated incidents. The person has to do the thing, not just watch it.
4. A second measurement. The same kind of evidence, weeks later. If the only data point is the day-one test, there's no progression.

The cycle is short: assess, practice, assess again. Content is an input to that cycle, not the product.

How to change the report in 90 days

You don't need to throw out the course platform. You need to stop using it as proof of capability.

Days 1–15. Pick one role and one capability (one team, not the whole company). Define what "ready" means in concrete work: a task, a deliverable, a criterion a technical lead would sign off on. Run an initial assessment. Write down the starting point, even if it's rough.

Days 16–60. A short development plan around that capability. Limited content. Weekly practice. Coaching so it doesn't die in week 3. The platform can host the content; it can't be the only record-keeping system.

Days 61–90. Repeat the assessment (or an equivalent practical test). Report three columns, not one:

The committee can see, on the same slide, that completion and capability are not the same number. That's the conversation the training team should be having.

If by day 90 the content was completed and the evidence didn't move, the problem isn't motivation. The program doesn't develop the capability it declared.

Checklist: activity vs evidence

Print this or use it as a guide for the next program. If you can't fill the right-hand column, you're not measuring capability yet.

What to do now

If the next committee asks for "the training number", bring both columns. The completion rate can stay on the slide. Just don't let it travel alone.

TalentOps is built around that cycle: a Career Path (a development plan per role and per the company's technologies), practical exercises, recurring assessments, and visibility of gaps. It doesn't replace a compliance platform. It answers the question the platform doesn't ask.

See TalentOps for companies