Relative or Absolute???

[Mike.Bean]

https://moz.com/learn/seo/internal-link
https://moz.com/blog/relative-vs-absolute-urls-whiteboard-friday

I have read both articles above, however further reading the comments that followed, it appears that neither provided an answer. Some said Absolute and some were determined that Relative linking is the only way to go. There were heated comments, apologetic comments and mostly comments agreeing to disagree.

But in the end I think what I got out of it was there are two sides to developing a site.

Side A:             Web Developer Standpoint, Relative (easier)

Side B:             SEO Standpoint, Absolute (harder)

To me Side B puts money in the bank, so I will go with B.

Now since we’ve been dealt the recent https algo, there seem to more opinions that are thrown into the question. All I’m asking for is a straight answer so our employees can feel confident that our efforts are not in loss. I also want to ask about how images should be linked internally.

NOTE: Our sites are not all https but have SSL’s in place, and 301 redirects reflect http versions of pages.

So since it appears Absolute linking is the preferred method, I ask which is the best option or an option at all.

Internal Navigation:

Current internal link - /store/pc/Rollators-c379.htm

Absolute Options

http://www.company.com/store/pc/Rollators-c379.htm (what I see for most Absolute links)
www.company.com/store/pc/Rollators-c379.htm (would this work for both http and https?)
//www.company.com/store/pc/Rollators-c379.htm (would this work for both http and https?)

All three links above resolve to www.company.com/store/pc/Rollators-c379.htm with no http:// in front. If I type https://www.company.com/store/pc/Rollators-c379.htm in the url it resolves to https://www.company.com/store/pc/Rollators-c379.htm with the https:// in front.

Internal Images:

If I want my images to be indexed as well as pages should I be using Absolute links for them as well?

Current internal image link - /store/pc/images/rollators.jpg

Absolute Options

http://www.company.com/store/pc/images/rollators.jpg (what I see for most Absolute links)
www.company.com/store/pc/images/rollators.jpg (would this work for both http and https?)
//www.company.com/store/pc/images/rollators.jpg (would this work for both http and https?)

Just once I would like an answer that’s not:
It depends on the site or it doesn’t matter which method you use, they both work lol.

I would like one that says google likes this better.

Thanks for taking the time to help us understand and can’t wait for MOZ Con this September.

[LynnPatchett]

Hi Mike,

I think you might get a lot of different opinions again (just like in Ruth's WBF comments) but I think it is pretty safe to say that Google likes sites that are consistent and clear in their internal linking and that do not return a bunch of 404 pages due to bad linking structure or linking mistakes.

So with this in mind the first thing to do is decide what your final url structure is going to be. Is it https or http? is it with www on non www?

Once this is decided then really 95% of any potential troubles should be solvable with a few lines of code in your htaccess file which will 301 any queries to urls that are not in the final format you have decided on to that format. This means that if you have decided on https and non www then even if google (or any user) comes into the site on a http or www url then the redirect takes them to the https:// version and everything else that is crawled from then should be in the format you have decided on. So far so good.

There are two main issues where you might still have problems and this is likely where different opinions will appear as to which problem is more likely to happen and/or be more difficult to fix.

Problem 1: Bad implementation of internal relative linking
The comments on Ruth's post about spider traps due to missing the / at the beginning of relative urls (and lots of other weird relative url formatting mistakes) happens quite often and on larger sites can seriously damage your crawl budget by creating literally thousands of bad links to 404 pages. This is bad news! There are ways to address this either using a base href tag or simply by editing the relevant links. The moz crawl report will flag these kind of issues so you will find out pretty quickly if you have this kind of problem.

Problem 2: Having to manually edit absolute urls because you have decided on changing the final url structure of your site.
You mention you have a ssl certificate in place but are not using https at the moment. It is probably safe to assume you will be at some point in the future though I would assume? If so, how many man hours are you willing to dedicate to editing all your absolute http urls to https urls if you decide to do this change? Maybe your tech team can do this by editing urls directly in the database and this is not a big issue for you (although these kind of actions need special care to make sure mistakes do not creep in) - if so then this might not be a big issue for you and in that case feel free to use absolute urls and feel good about it!

The above is relevant for images as well as for pages. At the end of the day google wants to be able to crawl your site easily and to find real content/images and not 404 errors all over the place. If you ensure this it doesn't matter much if you are using absolute or relative urls (I guess I better say in my opinion) - you are returning the same content on the same urls in both cases so really why would google care?
if it is easy to put absolute urls in the main menu then do it. If it is a nightmare to edit 1000's of relative urls in product descriptions then just make sure they return valid urls and you should be fine.

A note on your url examples:
http://www.company.com/store/pc/Rollators-c379.htm (what I see for most Absolute links)
www.company.com/store/pc/Rollators-c379.htm (would this work for both http and https?)
//www.company.com/store/pc/Rollators-c379.htm (would this work for both http and https?)

Only the first of those is an absolute url, the second and third are relative urls of one sort or another and are likely to cause you lots of problems. Either use absolute urls with full https or http, or use relative urls with a trailing / (and make sure you do not have spider traps). Do not use those other examples, they will cause you grief! Hope it helps, I am sure others will chime in