Knowing When a Page Is Finished

The two failure modes

Publishing too early, which produces a thin page nobody wanted.

And never publishing, which produces nothing at all and is the more common failure among people who care about writing.

The second is harder to notice because it looks like diligence, and because an unpublished draft feels closer to done than a published page that could be better.

It is not. An adequate page that exists outperforms an excellent page that does not, on every measure, permanently.

The standard

One question, and it is lower than most people set for themselves.

Would somebody who arrived with this question leave satisfied.

Not impressed, not delighted, and not having read the definitive treatment of the subject. Satisfied, meaning they got the answer and could act on it.

Most drafts reach that point considerably earlier than their authors think, usually around the second revision.

Everything after that is improvement rather than completion, and improvement is better done after publishing when there is evidence about what to improve.

Signs you have passed it

• Changes are to wording, not substance.
• You are moving sentences back and forth.
• Each pass changes less than the last.
• You are adding caveats nobody asked for.
• You have started researching a new section.

The last is the one that extends a draft indefinitely. A page that keeps growing a new section each time you open it has stopped being revised and started being expanded.

Publish, then improve with evidence

The argument for a lower bar, and it is practical rather than a concession.

Once a page is live you learn what queries bring people to it, which sections they read, where they leave, and what they ask afterwards.

That is far better information than any amount of guessing before publication.

A page revised three times before publishing and never again is worse than one published at the second draft and improved twice on evidence.

Set a review date at the point of publishing, three months out, so improving is scheduled rather than dependent on remembering.

That converts the anxiety about publishing something imperfect into a plan, which is usually what the anxiety wanted.

A worked example

A business owner had four drafts he had been working on for eleven months.

Each was on its fifth or sixth revision, and none was published, because each still had a section he was not happy with.

He published all four as they were, with the unsatisfactory sections shortened rather than fixed.

Within two months the search reporting showed that two of the four were being found for questions he had not anticipated, and that one of the sections he had agonised over was on a subject nobody was arriving for.

He improved the two pages that mattered, in the directions the evidence indicated, and left the others.

Eleven months of revision had been spent on guesses, and two months of publication produced better direction than any of it.

The checklist that ends it

Since a definition of finished needs to be checkable rather than a feeling.

The question is answered in the opening.

The specifics are there: figures, durations, conditions.

There is something only you could have written.

A reader knows what to do next.

Nothing on it is wrong.

When those five hold, it is finished. Wording, structure, and elegance are not on the list, because a reader does not notice them and you do.

Time-box it instead

A practical alternative for anybody who cannot apply a standard to their own work.

Decide before starting how long the page gets: half a day, or three sessions, or whatever is proportionate.

Publish whatever exists at the end of that, provided nothing on it is wrong.

That removes the judgement entirely, which for some people is the only thing that works.

It also calibrates over time, since you learn what a page of a given quality costs and can budget honestly.

The pages produced this way are frequently indistinguishable from the ones agonised over, which is itself a useful finding.

Somebody else can end it

The simplest mechanism where the problem is persistent.

Show the draft to somebody who is not you and ask one question: is there anything here you would need to know that is missing.

Not what could be better, which invites suggestions indefinitely, but what is missing, which has an end.

Where the answer is nothing, publish it, and treat that as the decision rather than reopening it.

A person outside the work is also better placed to judge whether it is understandable, which is the thing the author can no longer see.

Unfinished is different from wrong

A distinction worth holding, since the two get treated the same and only one is urgent.

A page missing a section, or covering something briefly that deserves more, is unfinished and can be published.

A page containing an incorrect figure, an out-of-date rule, or advice that could cost somebody money is wrong and cannot.

Separating those two lets you publish the first without anxiety and take the time the second deserves.

The check before publishing is only ever for the second kind, and it is a much shorter check than the general sense that something is not ready.

The counter-case

Some pages deserve the extra passes.

Anything with legal, safety, or financial consequences should be checked more carefully, and getting those wrong costs more than a delay.

A page that will be the main entry point to a site, or one that took real research, reasonably gets more attention than a routine piece.

And publishing genuinely poor work quickly is not a strategy, since a thin page is a thin page whenever it appears.

Apply the five checks, publish, set a review date, and improve on evidence rather than on imagination.

How to finish

1. Ask whether a reader would leave satisfied.
2. Check the five items, not the wording.
3. Notice when passes stop changing substance.
4. Time-box it if judgement fails.
5. Ask one person what is missing.
6. Publish.
7. Set a review date three months out.

Step seven is what makes step six possible, since most reluctance to publish is really a fear that this is the last chance to get it right.

Improving it later is covered in rewriting rather than adding.

Frequently asked questions

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.