2xN

Craft

The First Version Is Always Too Clever

Our opening move is nearly always overcooked, and the real work is the second pass where we take out everything we put in to prove we could.

21 September 2026 6 min read Craft

The pattern repeats itself across every project. Someone writes the first draft, codes the first prototype, sketches the first layout. It arrives overloaded. Too many features, too much explanation, too many flourishes that announce 'look what I can do' rather than 'here is what you need'. We have learned to expect this. The first version is always too clever because it serves two masters: the brief and the maker's need to demonstrate capability.

Understanding why this happens matters more than lamenting it. When we start a piece of work, we carry uncertainty about whether we are good enough, whether we understand the problem fully, whether the client will trust our judgement. That uncertainty expresses itself as over-delivery. We include the elegant technical solution, the unexpected structural choice, the reference that shows we have read around the subject. None of this is cynical. It is simply what happens when competence seeks visible proof of itself.

Why proving and solving pull in different directions

The impulse to prove competence operates on different logic from the impulse to solve a problem cleanly. Proving competence wants to show range, handle edge cases, demonstrate mastery of tools and techniques. Solving a problem wants the smallest intervention that creates the desired change. These two drives cannot occupy the same space. One seeks to impress, the other seeks to disappear into usefulness. The first version nearly always privileges the former because we have not yet earned our own trust.

This explains why experienced practitioners still produce overcooked first versions. The uncertainty does not vanish with seniority. What changes is the recognition that this is a stage, not a failure. The writer who has published a hundred articles still writes a first draft stuffed with subordinate clauses and unnecessary adverbs. The developer who has shipped dozens of products still reaches for the complex architecture before the simple one. We know this about ourselves, so we plan for the second pass. When working on content writing projects, we schedule time specifically for reduction.

The mechanics of the second pass

The second pass is not editing in the conventional sense. It is not about fixing errors or smoothing sentences. It is a different activity entirely: reading what you made and asking what job it actually needs to do. This requires detachment. You must stop defending the choices you made and start interrogating whether they serve the outcome. Most of what felt essential during creation turns out to be scaffolding. It helped you arrive at the real idea, but it does not need to stay in the final thing.

Practically, this means reading each element and asking whether its absence would weaken the work. If removing a feature, a paragraph, a visual element does not create a gap, it should not be there. This is harder than it sounds because we conflate effort with value. The component that took three hours to build feels like it has earned its place. But time spent is not an argument for inclusion. The only question that matters is whether it makes the thing work better. When we approach website design, we often find that the strongest layouts emerge after removing half the elements from the first version.

Recognising what the clever parts were really for

The clever parts served a purpose. They allowed you to think through the problem, to explore the boundaries of what was possible, to feel your way towards the shape of the solution. They were thinking tools, not building materials. A research study from the Nielsen Norman Group found that users consistently prefer simpler interfaces, even when more complex versions offer additional functionality they claim to want. This is not because users are unsophisticated. It is because cognitive load is real, and every element adds to it.

When we review our first version with this understanding, we can appreciate what the clever parts gave us without keeping them. That elaborate metaphor in the opening paragraph helped us clarify our argument. It does not need to stay now that we know what we are saying. That extra navigation option let us map all the possible user paths. Most users will only need three of them. The statistical deep-dive demonstrated our research rigour. A single well-chosen figure will do the same job with less friction. According to guidance from gov.uk content design principles, clarity always beats comprehensiveness for user outcomes.

Building the expectation into the process

Once you accept that version one will be too clever, you can design your process around it. This changes everything. Instead of trying to produce a clean first draft, you give yourself permission to explore. You write the long version, build the complex prototype, sketch the ambitious layout. Then you schedule dedicated time for reduction. This is not revision time. It is not polish time. It is deletion time, and it needs to be treated as a distinct phase with its own rules and mindset.

We have found that separating these phases improves both. The creation phase becomes more generative because you are not simultaneously trying to judge and refine. The reduction phase becomes more effective because you can approach the work with fresh perspective. When running app development projects, we explicitly schedule 'simplification sprints' after the initial build. Teams know that the goal is not to add anything, only to take away until the app does exactly what it needs to and nothing else.

What stays after everything else goes

The material that survives the second pass has a particular quality. It is usually the parts you found easiest to write, the features that seemed obvious rather than ingenious, the structural choices that felt almost too simple to count as decisions. This is not coincidence. The obvious parts connect directly to the core problem without needing to prove anything. They were not written to impress or to demonstrate range. They were written because they were true, necessary, useful.

This is why experience makes you faster, not because you write better first drafts but because you recognise the good parts sooner. You still produce the overcooked version, but you can identify which fifteen percent is doing the real work while the other eighty-five percent is decorative anxiety. That recognition cuts days from the process. When optimising for SEO performance, we repeatedly find that the highest-ranking content is the simplest in structure, not because search algorithms favour simplicity, but because clear information architecture and straightforward language produce better user signals.

When clever is actually required

Some problems genuinely need complex solutions. A detailed Google Ads campaign structure might require sophisticated audience segmentation that looks overcomplicated but reflects real user behaviour patterns. A technical specification might need comprehensive detail because any gap creates implementation risk. The distinction is not between simple and complex. It is between necessary complexity and complexity that exists to signal competence.

Necessary complexity serves the user. It handles genuine edge cases, addresses real variations in need or context, provides the information required to make good decisions. Research from the W3C Web Accessibility Initiative shows that complexity becomes necessary when serving diverse user needs, but even then, progressive disclosure keeps the initial experience simple. Complexity as signal serves the maker's anxiety. It proves you thought of everything, anticipated every question, demonstrated every capability. The second pass removes the latter and keeps the former.

The first version will always be too clever. This is not a problem to fix but a pattern to expect and work with. Plan for it. Create the overcooked version, then reduce it by half. What remains will be twice as good as anything you could have produced in a single pass, not because you tried harder but because you allowed yourself to separate proving from solving. The real work is not the first version. The real work is recognising what to keep when you take everything else away.

Ready to talk about your project?

Start a conversation
Back to all posts