The senior professional reviews it. They immediately spot where the problem lies, adjust the approach, correct two assumptions and return a much stronger version.
The work gets done sooner. The risk decreases. The outcome improves.
And yet, something important may have been lost along the way.
Because part of what turns experience into learning happens precisely before you know the right answer.
A review published in Review of Educational Research analysed 53 studies and 166 comparisons examining what happens when people try to solve a problem before receiving an explanation. The results showed improvements in conceptual understanding and in the ability to apply what had been learnt to new situations (g = 0.36). In designs specifically based on productive failure, the effect was greater (g = 0.37–0.58).
Although these studies come mainly from learning contexts, the mechanism is particularly relevant in professional environments where developing judgement requires progressively confronting poorly structured problems.
Difficulty alone does not teach.
What teaches is having to formulate a hypothesis, test it, discover where it fails and then understand why another approach works better.
That process matters particularly in roles where problems rarely arrive perfectly defined.
In data, technology and consulting, methodologies, accelerators and accumulated knowledge allow people to move faster, avoid known mistakes and tackle many problems with greater autonomy. But developing expertise also requires learning how to act when that framework does not provide the answer on its own: when the data is incomplete, an unexpected technical problem arises, the client’s need is not fully defined, several solutions are possible, or an initial hypothesis does not address the real business problem.
That is where something develops that cannot be acquired through accumulated knowledge alone.
Learning to navigate complexity
At Cognodata, this is one of the challenges we consider when supporting team development: the more experienced a professional becomes, the faster they recognise paths they already know will not work.
Intervening early is tempting.
It is also efficient.
But if that intervention systematically replaces the learning process of the person developing their skills, complexity remains concentrated in the same people.
And a paradox emerges.
The organisation may be improving efficiency at task level while delaying the distribution of capability across the team.
The junior professional delivers, but still needs someone else to interpret the difficult cases.
The senior professional ensures quality, but continues to absorb the most complex part of the work.
Over time, that dynamic can reinforce itself: precisely because the junior professional still needs support, we continue reserving for more experienced professionals the situations in which less experienced colleagues could develop greater autonomy.
Research into cognitive apprenticeship, originally developed by Allan Collins, John Seely Brown and Susan Newman, uses the term scaffolding to describe the temporary support that allows someone to tackle tasks they have not yet mastered. Its logic includes something fundamental: support should be gradually withdrawn as the person’s capability increases.
That changes how we understand professional support.
Supporting does not always mean solving
Supporting someone can mean allowing space for a first attempt before correcting it.
Asking them to explain a hypothesis before offering another.
Making visible how an experienced professional reasons through a problem.
Reviewing the process as well as the outcome.
And progressively increasing exposure to real problems while keeping their potential consequences within clear boundaries.
Not all complexity creates learning. A task that is disproportionate to someone’s current capability, unclear objectives or an error with critical consequences creates a different kind of problem. Evidence on complexity and job demands shows that complexity can support learning or become a source of overload depending on the resources, support and conditions under which it is faced.
Today’s efficiency can become tomorrow’s capability debt
That is the invisible cost of solving too soon.
The work moves forward, but part of the learning remains unfinished. Less structured problems continue to reach the same people, expertise takes longer to spread across the team, and dependency persists for longer — even if it is barely visible while everything is working.
The efficiency gained in each delivery can eventually accumulate as capability debt.
That is why the relevant question may need to come before assigning the next task:
What part of this work can we allow the person to learn to solve for themselves, and what support do they need to do so?
Protecting the outcome remains important.
So does protecting something less visible: the experiences through which we build the capability we will need tomorrow.
Because developing talent requires more than transferring knowledge.
It requires leaving room to build it.
Frequently asked questions about learning and professional development
Why can solving a problem too soon limit learning?
Because part of developing judgement happens while trying to solve the problem: forming hypotheses, testing them, identifying mistakes and understanding why another approach works better. If someone else systematically provides the solution before that process takes place, part of the learning is lost.
What is scaffolding in professional learning?
Scaffolding is temporary support that allows someone to tackle tasks they cannot yet manage independently. The key is to reduce that support progressively as their capability increases.
How can a senior professional help without solving the problem for a junior colleague?
They can allow a first attempt, ask the person to explain their reasoning before intervening, make their own thought process visible and review not only the final outcome but also the path taken to reach it.
When should you not leave room for error?
When the potential consequences are critical, the task is significantly beyond the person’s current capability, or the necessary resources and support are not available. Learning requires exposure to complexity, but also clear boundaries around acceptable risk.
