Chains form as sites change over years. Point every redirect directly at the final address rather than at another redirect, and check the four combinations of www and secure resolve in one hop.

How three hops happen

Nobody builds a redirect chain deliberately. Each link was sensible when it was added.

The site moved to a secure address, so insecure requests redirect to secure. The site standardised on the www version, so the bare domain redirects to www. A page was renamed, so the old address redirects to the new one. Later a whole section was restructured, so that new address redirects again.

Follow one old link now and you go through four hops before anything is served.

Every one of those decisions was correct in isolation, which is exactly why nobody notices the result.

What a chain costs

The fourth is the practical risk. A chain is a sequence of dependencies, and removing an intermediate rule that looks obsolete breaks every address that passed through it.

The four combinations

The most common chain exists on almost every site and is worth checking first.

Your site can be requested four ways: with and without www, secure and insecure.

All four should arrive at one final address, and ideally in a single hop each.

What frequently happens instead is that the insecure bare domain redirects to the insecure www version, which then redirects to the secure www version, which is two hops where one would do.

Type all four into a browser and watch what the address bar settles on, then check whether it took one step or several.

Fixing that is usually a configuration change taking minutes and it affects every visitor arriving from an old link.

Point at the destination, not the next hop

The rule that resolves this and prevents it recurring.

When a page moves for a second time, update the original redirect to point at the new final address rather than adding another rule after the existing one.

So an address that redirected to page B, where B has now become C, should be changed to point directly at C.

That keeps every old address one hop from wherever the content actually lives, however many times it has moved.

It requires knowing what the existing redirects are, which is why the list below matters more than the rule.

Keep a list

The record that makes this maintainable rather than archaeological.

A simple table: the old address, the current destination, the date, and why.

Without it, redirects accumulate in a configuration file nobody reads, and the person maintaining the site in three years cannot tell which are load-bearing.

The why column is the valuable one, since a rule labelled old service page merged into services in 2022 can be judged, and an unlabelled rule cannot.

Add a line whenever you add a rule, which is the discipline that keeps the list true.

A worked example

A business had rebuilt their site twice in six years, each time by a different supplier.

A link in an old directory listing pointed at a page address from the first version.

Following it produced four hops: insecure to secure, bare domain to www, the original page address to its second-version equivalent, and that to the current one.

It worked, which is why nobody had looked at it.

Collapsing the rules so every legacy address pointed directly at its current page took about an hour, and reduced the worst case to a single hop.

The measurable difference for a visitor was modest. The useful outcome was that the redirect rules became something a person could read, and two rules that pointed at pages which no longer existed were found in the process.

Watch for loops

The failure mode worth recognising, because it takes a page down completely.

A loop is where A redirects to B and B redirects back to A, and the browser gives up after a few attempts and shows an error about too many redirects.

It happens most often when two systems are both applying rules: a plugin and the server, or a security service in front of the site and the site itself.

The characteristic sign is that the page fails entirely rather than loading slowly, and that it frequently appears after a change nobody connected to redirects.

When it happens, look for two things fighting rather than one broken rule, and check whether something in front of the site was recently enabled.

Redirect to the equivalent page

A point about quality rather than quantity, and it matters more than the hop count.

Where an old page has a genuine equivalent, redirect to it. Where it does not, redirect to the most relevant section rather than the homepage.

Sending everything to the homepage is the common shortcut, and it wastes the arrival: somebody clicked a link about a specific thing and has to start again.

It also tends to be treated as a soft failure rather than a genuine move.

A few minutes deciding the destination for each old address is worth more than eliminating a hop.

The counter-case

Chains are a minor problem and should be kept in proportion.

Two hops costs a fraction of a second and loses very little, and a small business site with modest traffic will not notice the difference in any measurable outcome.

There is also a real risk in tidying redirect rules aggressively, since removing something that looks obsolete can break addresses still in use on printed material or in somebody else's links.

Never delete a redirect rule you cannot explain. Collapse chains by repointing them rather than by removing intermediate steps.

Fix the four-combination case, which is universal and cheap, and treat the rest as housekeeping rather than as an urgent problem.

The check

  1. Type all four combinations and watch the address bar.
  2. Confirm each arrives in one hop.
  3. Follow two or three old links you know exist.
  4. Repoint chains at the final destination.
  5. Never delete a rule you cannot explain.
  6. Send old pages to their equivalent, not the homepage.
  7. Keep a list with a reason column.

Step one takes two minutes, applies to every visitor, and is wrong on a surprising number of otherwise well-built sites.

Connecting several domains is covered in the second domain pointing at the same site.

Frequently asked questions

How do redirect chains form?

One reasonable decision at a time: insecure to secure, bare domain to www, a page renamed, a section restructured. Each was correct in isolation.

What do they cost?

A round trip per hop, a little lost at each step, crawl effort spent on hops, and fragility, since removing an intermediate rule breaks everything that passed through it.

What should I check first?

The four combinations of www and secure. All should arrive at one address, ideally in a single hop each. Type them in and watch the address bar.

How do I fix a chain?

Repoint the original redirect at the final destination rather than adding another rule after the existing one. Never delete an intermediate rule you cannot explain.

What is a redirect loop?

Where two addresses redirect to each other and the browser gives up. It usually means two systems are both applying rules, such as a plugin and the server.

Should old pages redirect to the homepage?

No. Send them to the genuine equivalent, or the most relevant section. Sending everything to the homepage wastes the arrival and tends to be treated as a soft failure.

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.

Never checked your redirects?

Type your domain four ways and watch the address bar. One of the four is usually taking two hops.

Start a Conversation