Rebuild the page from the question it answers rather than editing what is there. Additions accumulate at the bottom, which is where nobody reads.

How a page grows badly

Somebody notices the page does not mention a service, so a paragraph is added at the end.

Later a customer asks a question the page does not answer, so a section goes in below that.

A supplier suggests adding some detail, which becomes another block.

After four years the page is twice as long, contains everything it should, and is worse than it was, because all of the additions are below the point where anybody stops reading.

The signature of an accreted page

The fourth is the reliable tell. The first two paragraphs were written when the page was created and have never been revisited, because adding is easy and reopening the beginning feels like starting again.

Why adding feels safer

Worth understanding, since the instinct is strong and mostly wrong.

Adding cannot break anything. The existing page still says what it said, and the new material is extra.

Rewriting risks losing something that was working, particularly on a page that ranks, which produces a genuine fear of touching it.

And adding is quicker, since writing three paragraphs takes less time than deciding what a page should say.

All of that is true and none of it makes the resulting page good, which is the trade being made without anybody noticing.

Rebuild from the question

The method, and it is different from editing.

Write down the question the page exists to answer, in one sentence, as a customer would ask it.

Then list what somebody with that question needs to know, in the order they need it, without looking at the existing page.

Only then open the old page and take from it what fits the list.

That produces a structure driven by the reader rather than by the order in which material happened to be added.

It also reveals what is missing, which editing never does, because editing works from what is there rather than from what is needed.

Most of the old text survives. What changes is where it sits and what was cut.

A worked example

A service page had grown from six hundred words to nineteen hundred over five years.

The pricing information, which was the most valuable thing on it, had been added in year three and sat two thirds of the way down.

The opening two paragraphs described the business in general terms and had not been touched since the site was built.

Rewritten from the question, it opened with what the job involves and what it costs, followed by the process, then the exceptions.

It came out at about eleven hundred words, and roughly seven hundred of those were sentences that had already existed somewhere on the page.

Enquiries from it improved noticeably, and nothing had been added.

Cutting is the hard part

Since a rewrite that keeps everything is an edit with extra steps.

Cut anything said twice, keeping the better version.

Cut the general introduction, which almost every accreted page has and which almost never earns its position.

Cut material that belongs on a different page, and move it there rather than deleting it.

Cut anything a competitor could also have written, which is usually the oldest text on the page.

Expect a rewritten page to be shorter than the accreted one and longer than the original, which is what a well-structured version of the same material looks like.

Protecting what works

The legitimate fear, addressed practically.

Note what the page currently ranks for and how much traffic it gets, before touching it.

Keep the address unchanged, which is the thing that actually carries the ranking.

Keep the words the page is found for, since removing the phrase somebody searches is the one change that reliably costs something.

Then check the figures a month later against what you recorded.

In practice a rewrite that keeps the topic and improves the structure very rarely loses position, and the fear is larger than the risk.

Doing it on a schedule

Because accretion happens continuously and correction does not.

Once a year, take the three pages that matter most and rebuild each from its question.

Three pages is enough, since the pages that matter are few and the rest do not repay the effort.

Choose by traffic and by commercial importance rather than by which feels worst, since the page you dislike is frequently not the page doing the work.

That is roughly a day a year and it prevents the twenty-page version of this problem.

Read it aloud once

The check that catches accretion faster than any analysis.

Read the finished page aloud from the beginning, without stopping, as if explaining it to somebody.

The joins are audible in a way they are not visible: a sentence that repeats what you said two minutes ago, a section that arrives from nowhere, a paragraph you find yourself skipping.

Anywhere you would not say it that way to a customer standing in front of you is a place to change.

It takes about five minutes for a long page and it is the closest thing to reading your own work as a stranger.

The counter-case

Adding is sometimes exactly right.

A genuinely new section on a well-structured page, placed where it belongs rather than at the end, is an addition and not accretion.

Some pages are also reference material where order matters less, such as a specifications list, and appending to those is fine.

And a page performing well should not be rewritten for tidiness, since the risk is real and the return may be nothing.

Rebuild the pages that matter, place additions where they belong, and treat a page that has never had anything removed as a candidate.

The method

  1. Record what it ranks for and its traffic.
  2. Write the question in one sentence.
  3. List what a reader needs, in order.
  4. Then open the old page and take what fits.
  5. Cut duplication and the general opening.
  6. Keep the address unchanged.
  7. Check the figures a month later.

Step three is what makes this a rewrite rather than an edit, since working from the existing text can only rearrange what somebody already thought of.

Whether to update or start fresh is covered in updating an old post instead of writing a new one.


Frequently asked questions

How does a page grow badly?

By addition. Each new paragraph goes at the end, so after several years everything valuable is below the point where anybody stops reading.

What is the signature?

The best material near the bottom, sections that do not follow from each other, the same thing said twice, and an opening that is the oldest text on the page.

Why does adding feel safer?

It cannot break anything, the existing page still says what it said, and it is quicker than deciding what a page should say. All true, and none of it makes the page good.

What is the method?

Write the question the page answers, list what a reader needs in order, and only then open the old page and take what fits. That reveals what is missing, which editing never does.

Will a rewrite lose my ranking?

Rarely, if you keep the address, the topic, and the words the page is found for. Record the figures first and check a month later. The fear is larger than the risk.

How often should I do this?

Once a year, on the three pages that matter most, chosen by traffic and commercial importance rather than by which one feels worst.

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.

Page that has only ever grown?

Write the question it answers, then list what a reader needs before opening it. That order is the whole method.

Start a Conversation