Redirect or not to redirect

[sesertin]

We are rebuilding a website and try to get rid of errors. The content remains exactly the same but we correct the code and make it load faster. The site has quite many backlinks and I can't decide whether to remove .html endings from the urls and 301 redirect to the new ones or leave them with the older ending. If I remove the endings how much of the link juice will be passed?

Anyone any idea?

[WilliamBay]

Well 301's typically have a tiny falloff of link juice (I've heard somewhere in the 5% range anecdotally). But i would think about what makes sense to your users over that. Does having that .html make the urls harder to remember for a user to type in? Is it .htm or .html?

I don't use .html because it is an extra thing to remember. And by not using it, I can have shorter urls that focus more on being simple and having keywords.

As far as 301ing, you could probably write something in your .htaccess file that automatically does that for the entire site.

But I would be inclined to get rid of them.

[EGOL]

I can't decide whether to remove .html endings from the urls and 301 redirect to the new ones or leave them with the older ending.

If you redirect you will lose some linkvalue and some anchor text value.

I have learned to tolerate "untidy little things" that are very expensive to change.  I don't want to put 5-10% (just tossing out a number) more effort into ranking my site just to get rid of .html endings.

[BTeubner]

Agree, keep the .html

[sesertin]

Thanks for your answer

[sesertin]

Thanks for your answer

[AlanMosley]

in a year of 2, will you then decide to make the change? only you know that, it is better to do it sooner than later if you think you may.

As for leaking link juice, the best guess is 15% going from the original algo google published, every hop lost 15% I dont think they would of changes this much as changing it can mean huge amounts if calulating, I believe they got it right then and it would still be right.

If hops did not lose pr, the caculations of PR would be endless.