Our copy isn't written yet — should we start the redesign anyway?

Almost nobody has finished copy when they start. That is normal and it is not the problem. The problem is that copy is usually everyone’s job, which makes it nobody’s. The marketing lead assumes the founder will write the homepage because it is the founder’s story. The founder assumes marketing will draft something to react to. Nine weeks later the design is finished, the build is finished, and seven pages still say Lorem ipsum.

In our projects the delay is never spread evenly either. It lands on the same pages every time: the homepage, the pricing or services page, and the two or three pages that require someone to make a decision about what the company actually claims. Blog posts and legal pages arrive fine. The pages that need a judgement call sit still.

So the question to answer before you book anything is not “is the copy ready”. It is “who is writing the homepage, and which afternoons has that person cleared to do it”.

Content-complete: the copy exists, it is approved, it is in a document. This is the rarest state by a wide margin. Design can start tomorrow and layout decisions will hold.

Content-shaped: you know what each page has to argue and roughly how long it runs, but the sentences are not written. This is the normal state, and it is a perfectly good place to start from. Say so out loud to whoever is worried about it.

Content-absent: nobody has decided what the site says or why the company is worth choosing. Design here produces a handsome container for an argument you have not made. The site will launch, it will not work, and the redesign will get the blame. This is the only one of the three where starting is a mistake.

Do this yourself, today, in a plain document. No agency, no workshop, no template. It tells you more than any discovery call.

If step 3 defeats you, you are content-absent, and the honest next move is writing rather than designing. If steps 1 and 2 go fine but the sentences in step 3 are rough, you are content-shaped and you can start.

1. List every page the new site will have. If the list runs past fifteen pages, you have a content problem wearing a design costume — most of those pages exist because nobody wanted to delete them, and paying to redesign them will not make them worth reading.
2. For each page, write one sentence: what should a reader believe after reading this page? If you cannot write that sentence, design cannot help that page yet.
3. For the homepage and your two most important pages, write the headline and the first paragraph. Real words, not a description of the words. This is the cheapest test in existence of whether you know what you sell.
4. Count the pages that need a photograph or a case study that does not exist yet. Photography and case studies are content too, and they are the slowest kind — a case study needs a client's permission, and permission takes weeks.

Design needs the shape of the content, not the words. Shape means: how many words, how many sections, is there a list, is there an image, is there a number worth making large. Give us the shape early and the sentences can arrive late.

Four things we ask for before drawing anything: the page list and which three pages matter most; the one-sentence purpose per page; real headlines and first paragraphs for those top three pages; and an honest answer on photography — do you have real pictures, or are we using stock?

The headline point is the one people resist and the one that costs money to skip. A nine-word headline and a two-line headline are different layouts. Swapping one for the other after the design is approved is real paid rework, usually a day or two per page template.

These can follow behind safely: body copy on secondary pages, as long as the rough length is known; blog and help-centre articles; legal pages, careers pages, anything with no layout risk. Nobody has ever delayed a launch because the privacy policy was late.

Grey Latin is even. Every line breaks at a comfortable length, no word is awkward, there are no proper nouns. Real copy is lumpy. It has an eleven-word product name that wraps onto three lines, a client logo that is a wordmark in a colour that fights the background, a job title that is longer than the person’s name, and a sentence the legal team insisted on that runs to forty words.

A design approved against placeholder text is a design approved for a website that will never exist. The first day of real content is the day the cracks appear, and by then the design is signed off and the budget is spent.

What to insist on: before you sign off on the design, ask to see the three key pages with real words in them. Not all of them. Three. If those three hold up with your actual lumpy sentences in place, the rest of the site will hold up too.

Three situations where we have told people to spend their money elsewhere, and would again.

You cannot write your homepage headline because the business is mid-repositioning. Wait. A redesign fixes a story in place, and you are about to change the story. Come back when you know what you are.

Your problem is that the writing is unclear, not that the site is ugly. A good copywriter, working on the site you already have, will outperform a full redesign for a fraction of the cost — often a few thousand pounds against a five-figure project. Do that first. The redesign afterwards is faster and cheaper anyway, because the hard thinking is done.

You have no case studies, no photography, and no capacity to produce either. A better design makes that emptiness more visible, not less — good layouts have space in them, and space needs something to hold. Spend the first month of the budget on content production instead. Book a photographer, ring two clients for quotes.

This is for the person who will be blamed when the launch slips. Five mechanics, all boring, all effective.

The cut-off is the one that saves projects. It converts a stalled site into a launched site with a few weak pages, which is a much better position to be in.

• One named owner per page, written down, with a date. A person, not a department. "Marketing" writes nothing.
• Write into the layout, not into a blank document. People write two to three times faster when they can see the box the words go in and how many words fit it. We hand over draft pages for exactly this reason.
• Book the writing as meetings in a calendar — two-hour blocks with a title. Copy does not get written in the gaps between other work.
• Agree the approval chain before the first draft. Copy dies in review, not in drafting. Three reviewers works. Seven is a stall dressed up as diligence.
• Set a cut-off date: any page without approved copy by then launches with the old copy and gets updated later. Launching an imperfect page beats holding the entire site hostage to it.

A finished, unlaunched site is expensive in ways that do not appear on an invoice. It keeps occupying your team — status calls, half-answered questions, a design nobody can quite remember agreeing to. It burns the goodwill that made the project possible. In our experience, enthusiasm has a shelf life of about eight weeks past the point where the design is done.

Copy arriving at three times the expected length is straightforward paid rework: the section has to be redesigned, the component adjusted, the page retested. It is not catastrophic, but it is real, and it is the most avoidable line item in a redesign.

And a site launched with placeholder-shaped content — the vague sentences written at 11pm to unblock a deploy — gets quietly rewritten within about six months. The writing did not get skipped. It got postponed, and it got done worse.

The decision to make this week: pick one person, give them the homepage, and put two hours in their calendar before Friday. Everything else in this article follows from whether that happens.