What to Take With You and What to Leave

The default is to move everything

It is the path of least resistance and it is almost always wrong.

Moving everything requires no decisions, offends nobody, and feels safe. It also means the new site launches carrying every weak page, every duplicate, and every piece of content written for a business you stopped being.

You will have spent significant money on a new foundation and put the old problem back on top of it.

Why this is the only moment

Removing content from a live site is a fight.

Somebody wrote the page, somebody else believes a client reads it, and nobody wants to be the person who deleted something.

During a rebuild the question changes shape. It is no longer are we deleting this, it is are we rebuilding this, and rebuilding something requires a positive reason rather than merely the absence of an objection.

That reframing is worth a great deal and it lasts about six weeks.

Four piles, not two

• Take, as it is, because it works.
• Rewrite, because the topic earns a page and this version does not.
• Merge, folding several thin pages into one good one.
• Leave, and redirect the address somewhere sensible.

Two piles produce an argument about deletion. Four piles produce a plan, because most contested pages turn out to belong in rewrite or merge rather than at either extreme.

What earns a place

A page justifies rebuilding if it does at least one of three things.

It brings people in, which the traffic data will tell you.

It answers a question that costs you sales when unanswered, which your own inbox will tell you.

Or it is required, whether legally or because the business genuinely needs it published.

A page doing none of those is being carried, and a rebuild is the moment to stop carrying it.

What people move that they should not

News items about events that have happened.

Staff profiles for people who left, which are also a small privacy discourtesy.

Pages for services you would rather not be asked about.

Three near-identical pages built for locations, differing only in a place name, which were a tactic that no longer works well and reads badly.

And a blog archive of forty posts, thirty of which were written to fill a schedule and say nothing.

The blog archive question

This is the argument that takes longest, so it is worth settling directly.

The instinct is to keep it all, on the grounds that more pages is better. That was truer once than it is now.

Look at the archive the way you looked at the rest of the site: which posts bring people in, which answer a real question, and which were filler.

Keep the first two categories, and consider merging clusters of thin posts on one topic into a single substantial page.

A well-kept archive of fifteen useful articles is a stronger asset than eighty pages of which sixty-five have never been read.

A worked example

A professional firm rebuilding a ninety-one page site started by assuming all ninety-one would move.

The sort took two afternoons with the traffic data open.

Thirty-four went straight into take. Nineteen went to rewrite, mostly service pages that were accurate and thin. Twenty-two collapsed into six merged pages. Sixteen were left behind.

The new site launched with sixty-three pages instead of ninety-one.

The build itself came in lower, because the quote had been priced per template and per page.

More usefully, the nineteen rewrites got real attention, which would not have happened if the same budget had been spread across ninety-one pages nobody had time to look at properly.

Take the assets, not just the pages

Content is not only text, and the other material is easier to lose.

Original photographs at full resolution, which are frequently only on the old site in a compressed form and nowhere else.

Downloadable documents, specification sheets, and price lists, including ones linked from pages you are leaving behind.

Customer reviews and testimonials published on the site, with the names and dates attached.

And any form submissions or enquiry records held inside the old system, which will not migrate on their own and are frequently the only record of who contacted you.

Decide before the build, not during

The sort has to be finished before anybody designs anything, for a practical reason.

The number of pages and the number of distinct page types determine the shape and the price of the build.

A decision to drop a third of the site made in week six means redesigning templates that have already been built, and the saving evaporates into rework.

It is also the point where people are most willing to be decisive, because nothing has been made yet and no effort is being wasted.

Who decides, and how to keep it civil

The sort is not really a technical exercise. It is a series of small judgements about other people's work, and it goes wrong socially more often than analytically.

Two things keep it civil.

The first is deciding from data rather than taste. Twelve months of traffic beside each page turns an argument about whether a page is any good into a conversation about whether anybody reads it, which is a much easier discussion to have with the person who wrote it.

The second is offering rewrite as a real option rather than a polite way of saying no. A page moved to rewrite is a commitment to spend time on the topic, and people who wrote the original are usually the right ones to do it.

Where a page has no traffic and somebody insists it matters, ask what it is for and who reads it. Sometimes the answer is good and the page stays, which is the system working rather than failing.

The counter-case

Pruning can be overdone, and there are two ways it goes wrong.

The first is dropping pages that get little traffic but serve a specific purpose, such as a page that answers one expensive objection for the handful of buyers who reach it.

The second is cutting so deep that the site no longer covers what the business does. A site trimmed to eight pages may be elegant and unable to rank for anything, because there is nothing on it about most of what you sell.

The goal is a site where every page earns its place, not the smallest possible site. Those are different targets and the second one is easy to hit by accident.

The sort

1. List every page with twelve months of traffic beside it.
2. Mark each take, rewrite, merge, or leave.
3. Group the merge candidates by topic.
4. Map every leave to a redirect destination.
5. Collect the assets: photographs, documents, reviews, enquiry records.
6. Finish before the design starts.

Two afternoons, and it changes both what the rebuild costs and what it is worth.

Judging whether a page is still true is covered in content written for a business you no longer are.

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.