Competition & Award Submissions

Every competition and award has a brief, and buried in it - sometimes explicitly as a scoring rubric, sometimes implicitly in the questions asked - is how you will be judged. The single most common reason strong projects lose is not weak design; it is a submission that does not answer the brief: a required question skipped, a criterion the jury must score left unaddressed, a word limit blown so the last board is cut. Before you write a sentence, extract the rules, and Claude is genuinely useful as a comprehension aid here - a second reader who never skims.

Paste the full brief and ask it to lay the requirements bare:

Here is the full competition brief [paste]. Extract, as a checklist:
(1) every question or section I must respond to, in order; (2) the word
or character limit for each; (3) the stated judging criteria and any
weightings; (4) mandatory inclusions (drawings, declarations,
eligibility); (5) anything easy to miss. Quote the brief for each
point; do not add requirements it does not state.

That last instruction matters - you want what the brief actually says, not what Claude assumes competitions usually want. Then verify the checklist against the brief yourself, because you are the one who is disqualified if it is wrong, and Claude can misread or invent a requirement as easily as miss one. Pay special attention to the mechanical rules that quietly disqualify entries - the character limit that counts spaces, the board size and orientation, the file format, the anonymity requirement that bars your name or logo from the sheets. These are unglamorous and absolute, and a brilliant entry that breaks one is often not even read. What you now hold is the specification your entry must meet: the exact questions, the exact limits, the exact criteria. Every later step - structuring, drafting, trimming - is measured against it. A jury marks to the criteria; your job, before it is to be brilliant, is to be complete and compliant, and this checklist is how you guarantee it.

With the checklist in hand, structure the entry to mirror it exactly. The figure shows the discipline: each question the brief asks maps to a response block, in the brief's own order, sized to the brief's own limit. This sounds obvious and is constantly ignored - designers write the narrative they want to tell and hope it covers the questions, when a scored jury wants to find each answer where the rubric expects it. Make it easy for a tired juror to award you every mark: if they ask about sustainability in 100 words, give them a 100-word sustainability block they can find at a glance, not a sustainability theme woven poetically through eight paragraphs.

Claude helps you build and populate that structure from material you already have - your concept and rationale from the earlier lessons. A strong move is to give it your rationale and the question checklist and ask it to map one onto the other:

Here is my design rationale [paste] and the competition's required
questions with word limits [paste]. For each question, draft a
response using only what is in my rationale, within the limit. Flag
any question my material does not yet answer - do not invent content
to fill a gap.

The flag is the valuable part. Where Claude reports "your material does not address the community-benefit question", it has found a real hole in your entry, not a writing problem - and the fix is a design or thinking decision you make, then feed back, never an invention Claude papers over. Ask Claude for each block at, and slightly under, its limit, so you have words to trade between sections when the whole entry has to fit a fixed board. Populate every block from true material, keep each within its limit, and hold the brief's order. What emerges is an entry that answers the jury's scoring sheet line by line - the unglamorous discipline that wins far more competitions than a clever turn of phrase.

Word limits are where good entries go to die, and where Claude is at its most genuinely useful. Compression is a real skill - saying the same thing in 300 words instead of 520 without losing the argument - and it is slow, painful work by hand at midnight. Claude does a first pass in seconds, and the loop in the figure keeps it honest: you draft long, Claude trims to the limit while preserving meaning, you check that nothing true was dropped and nothing false was added, and it fits.

Brief the compression precisely, because how you ask changes what survives:

This is 520 words; the limit is 300. Cut it to 300 or fewer WITHOUT
losing any of these three points [list your must-keeps]. Prefer
cutting adjectives and throat-clearing over cutting substance. Keep my
voice. Return the word count.

Naming your must-keeps protects the argument, because left to itself Claude may trim the very sentence that wins - the one specific, unusual point that distinguishes your entry - because it reads as prunable detail. That is the first thing to check on the way back: is the idea still sharp, or has compression sanded it into something smooth and generic? The second check is the one that matters most in a judged document: did the trim introduce anything untrue? Compression sometimes over-claims - "reduces energy use" quietly becoming "eliminates energy use", a hedge dropped so a careful claim reads as an absolute. Read the trimmed version against the truth, not just against the word count. Claude is a superb sub-editor for length; it is not the author who is accountable for the claims, so the fit-to-limit is Claude's contribution and the fit-to-truth is yours.

Everything in this module has circled one discipline, and in a submission it becomes non-negotiable, because the document is public, judged, and often signed. A competition entry or award submission frequently carries declarations - of eligibility, of authorship, of the truth of what you claim - and juries and awarding bodies do check, sometimes forensically for shortlisted schemes. The failure mode is specific and seductive: Claude, asked to make an entry compelling, will happily supply an impressive-sounding performance figure, a superlative, or a claim of impact that your project cannot actually support, and it will read beautifully. Every such claim is a liability you are signing your name beside.

So run a dedicated honesty pass before anything is submitted, separate from the writing. Go through the entry and mark every factual claim - every number, every superlative, every statement about performance, cost, precedent or impact - and confirm each against what you actually did or measured. You can enlist Claude to find them - "list every factual or numerical claim in this submission so I can verify each" - but the verifying is yours, at the source, because Claude cannot tell you whether your building actually achieves what the sentence says. Strike or soften anything you cannot stand behind. An honest, specific, complete entry beats an inflated one, and it does not risk your reputation.

Then the final polish, and here your voice matters more than anywhere. Juries read dozens of entries, and the ones that blur together are the ones written in the smooth, ownerless prose that Claude produces by default - the architalk of the last lesson, now competing against a hundred others just like it. Read your entry aloud, cut the generic sentences, restore the specific and the particular, and make sure it sounds like a designer who made real choices on a real site. Claude got you complete, compliant, within limits and clear; the last pass - the truth-check and the voice - is what makes it both safe to submit and worth reading. That combination, far more than a poetic flourish, is what wins.