Crawled page count in Search console

[Bob_van_Biezen]

Hi Guys,

I'm working on a project (premium-hookahs.nl) where I stumble upon a situation I can’t address. Attached is a screenshot of the crawled pages in Search Console.

History:

Doing to technical difficulties this webshop didn’t always no index filterpages resulting in thousands of duplicated pages. In reality this webshops has less than 1000 individual pages. At this point we took the following steps to result this:

1. Noindex filterpages.
2. Exclude those filterspages in Search Console and robots.txt.
3. Canonical the filterpages to the relevant categoriepages.

This however didn’t result in Google crawling less pages. Although the implementation wasn’t always sound (technical problems during updates) I’m sure this setup has been the same for the last two weeks. Personally I expected a drop of crawled pages but they are still sky high. Can’t imagine Google visits this site 40 times a day.

To complicate the situation:

We’re running an experiment to gain positions on around 250 long term searches. A few filters will be indexed (size, color, number of hoses and flavors) and three of them can be combined. This results in around 250 extra pages. Meta titles, descriptions, h1 and texts are unique as well.

Questions:

1. -          Excluding in robots.txt should result in Google not crawling those pages right?
2. -          Is this number of crawled pages normal for a website with around 1000 unique pages?
3. -          What am I missing?

BxlESTT

[donford]

Hello Bob,

Here is some food for thought. If you disallow a page in Robots.txt, google for example will not crawl that page. That does not however mean they will remove it from the index if it had previously been crawled. It simply treats it as inaccessible and moves on. It will take some time, months before Google finally says, we have no fresh crawls of page x, its time to remove it from the index.

On the other hand if you specifically allow Google to crawl those pages and show a no-index tag on it, Google now has a new directive it can act upon immediately.

So my evaluation of the situation would be to do 1 of 2 things.

1. Remove the disallow from robots and allow Google to crawl the pages again. However, this time use no-index, no-follow tags.

2. Remove the disallow from robots and allow Google to crawl the pages again, but use canonical tags to the main "filter" page to prevent further indexing the specific filter pages.

Which option is best depends on the amount of urls being indexed, a few thousand canonical would be my choice. A few hundred thousand, then no index would make more sense.

Whichever option, you will have to insure Google re-crawls, and then allow them time to re-index appropriately. Not a quick fix, but a fix none the less.

My thoughts and I hope it makes sense,

Don

[Bob_van_Biezen]

Hello Don,

Thanks for your advice. What would your advice be if the main goal would be the reduction of crawled pages per day? I think we got the right pages in the index and the old duplicates are mostly deindexed. At this point I’m mostly worried about Google spending it’s crawlbudget on the right pages. Somehow it still crawls 40.000 pages per day while we only got around 1000 pages that should be crawled. Looking at the current setup (with almost everything excluded though robots.txt) I can’t think of pages it does crawl to reach the 40k. And 40 times a day sounds like way to many crawled pages for a normal webshop.

Hope to hear from you!

[Bob_van_Biezen]

Hi Don,

Just wanted to add a quick note: your input made go through the indexation state of the website again which was worse than I through it was. I will take some steps to get this resolved, thanks!

Would love to hear your input about the number of crawled pages.

Best regards,

Bob

[donford]

Hi Bob,

You can "suggest" a crawl rate to Google by logging into your webmasters tools on Google and adjusting it there.

As for indexing pages.. I looked at your robots and site. It really looks like you need to employ some No Follow on some of your internal linking, specifically on the product page filters, that alone could reduce the total number of URLS that the crawlers even attempts to look at.

Additionally your sitemap http://premium-hookahs.nl/sitemap.xml shows a change frequency of daily, and probably should be broken out between Pages / Images so you end up using two sitemaps one for images and one for pages. You may also want to review what is in there. Using ScreamingFrog (free) the sitemap I made (link) only shows about 100 urls.

Hope it helps,

Don

[Bob_van_Biezen]

Hi Don,

You're right about the sitemap, noted it on the to do list!

Your point about nofollow is intersting. Isn't excluding in robots.txt giving the same result?

Before we went on with the robots.txt we didn't implant nofollow because we didn't want any linkjuice to pass away. Since we got robots.txt I assume this doesn’t matter anymore since Google won’t crawl those pages anyway.

Best regards,

Bob

[donford]

Hi Bob,

About the nofollow vs blocked. In the end I suppose you have the same results, but in practice it works a little differently. When you nofollow a link it tells the crawler as soon as it encounters the link not to request or follow that link path. When you block it via robots the crawler still attempts to access the url only to find it not accessible.

Imagine if I said go to the parking lot and collect all the loose change in all the unlocked cars. Now imagine how much easier that task would be if all the locked cars had a sign in the window that said "Locked", you could easily ignore the locked cars and go directly to the unlocked ones. Without the sign you would have to physically go check each car to see if it will open.

About link juice, if you have a link, juice will be passed regardless of the type of link. (You used to be able to use nofollow to preserve link juice but no longer). This is bit unfortunate for sites that use search filters because they are such a valuable tool for the users.

Don

[Bob_van_Biezen]

Hi Don,

Thanks for the clear explanation. I always though disallow in robots.txt would give a sort of map to Google (at the start of a site crawl) with the pages on the site that shouldn’t be crawled. So he therefore didn’t have to “check the locked cars”.

If I understand you correctly, google checks the robots.txt with every single page load?

That could definitely explain high number of crawled pages per day.

Thanks a lot!

[donford]

Ben,

I doubt that crawlers are going to access the robots.txt file for each request, but they still have to validate any url they find against the list of the blocked ones.

Glad to help,

Don