Redesigning without losing your position

Your content is not what is at risk

The question that arrives first is always whether the text and the photographs survive. That is the easy part. Content migrates and there is nothing more to say about it.

What is at risk are the addresses. Every page on your site has one, and over the years that address quietly accumulates things: it appears in search results, other sites link to it, people have it saved, it is printed on something.

All of that is attached to the address, not to the page. Change the addresses without connecting old to new and each one returns an error. The engine stops finding what it had indexed, the external links point at nothing, and years of accumulation go at once.

It is not caused by redesigning. It is caused by redesigning without an inventory.

The fifteen per cent myth

You will hear that every redirect eats a share of a page's strength, usually quoted at fifteen per cent. It gets repeated in enough places to sound settled.

It is false, and this is not our opinion. Google's documentation says it outright: "301s and other permanent redirects do not cause loss of PageRank."

That single fact changes how you plan. If redirecting cost fifteen per cent, it would be worth fighting to keep bad old addresses just because they had history. Since it costs nothing, the new structure can be decided on what serves the project. What you have to do is redirect properly.

The inventory goes before the design

This is the step everybody skips and the one that prevents most disasters. It happens before anything is designed. Not after launch.

It is a list of every address that exists today, and for each one, three facts.

Does it get visits, which Search Console will tell you along with the searches people arrive on. Does anything link to it from outside, because a page with external links is worth far more than one with more traffic and none. What is its equivalent on the new site, meaning the specific page that answers the same thing.

With that, the guessing stops. You know which pages cannot fail, which can be merged, and which can be allowed to die because nobody has visited them in three years.

One more thing, because the inventory rots: the structure of the new site will change during the build, and the map has to change with it. A redirect map written in month one and never revisited is how live pages end up pointing nowhere.

The mistake that empties a site in a week

Sending everything to the homepage.

It is tempting because it is one line of configuration and nothing returns an error afterwards, which makes it look like it worked.

What it actually says is that none of those pages has a real equivalent any more. Treated that way, the accumulated signals tend to evaporate rather than transfer, and the visitor who clicked a link about one specific thing lands on your front door with no idea why. They leave. That is measured too.

A redirect is a statement that this page has become that page. If that is not true, do not make it.

The related trap: chains. An old address that redirects to another redirect that redirects again wastes what it was carrying and slows everybody down. Point the original straight at the final destination.

What to check after launch, and for how long

Verification happens afterwards, on the live site, and it is short work.

Take the list of addresses that mattered from the inventory and open them. Every one should land on its equivalent in one step. Then watch Search Console for errors, which is where the ones you forgot show up first.

Expect a wobble. Positions move for a few weeks while the new structure is re-crawled, and it is normal. What is not normal is a fall that keeps falling after a month, and that is your signal to go back to the map rather than to wait it out.

And keep the redirects. There is a habit of cleaning them up after six months, which quietly reopens the same hole. For addresses with external links pointing at them, keep them indefinitely. You do not control who links to you or when they will next follow it.

What is genuinely worth changing anyway

None of the above is an argument for keeping a bad structure out of fear.

If your addresses were generated by a system nobody chose, if they carry parameters and numbers that mean nothing, or if the structure reflects a company reorganisation from six years ago, change them. Redirecting is free, in the specific sense stated above.

What is not worth it is changing addresses that are working, for tidiness. Every change is a small risk taken for a small return, and a hundred of them together is not a small risk.

When to worry, and when not to

If your site gets most of its visitors from search, this work is not optional and the inventory is the first invoice line.

If your visitors arrive from an email list, from a sales team sending links, or from people typing your name, the risk is much smaller than the anxiety around it. Do the inventory anyway, because it is cheap, but the redesign is not the threat you have been told it is.

And if somebody quotes you for a redesign without asking what your current addresses are, that quote does not include this work. It is worth asking whether that is because it is cheap or because it is not happening.