On-site Search - Revisited (again, *zZz*)

[Mark_Elton]

Howdy Moz fans!

Okay so there's a mountain of information out there on the webernet about internal search results... but i'm finding some contradiction and a lot of pre-2014 stuff. Id like to hear some 2016 opinion and specifically around a couple of thoughts of my own, as well as some i've deduced from other sources. For clarity, I work on a large retail site with over 4 million products (product pages), and my predicament is thus - I want Google to be able to find and rank my product pages. Yes, I can link to a number of the best ones by creating well planned links via categorisation, silos, efficient menus etc (done), but can I utilise site search for this purpose?

1. It was my understanding that Google bots don't/can't/won't use a search function... how could it? It's like expeciting it to find your members only area, it can't login! How can it find and index the millions of combinations of search results without typing in "XXXXL underpants" and all the other search combinations? Do I really need to robots.txt my search query parameter? How/why/when would googlebot generate that query parameter?

2. Site Search is B.A.D - I read this everywhere I go, but is it really? I've read - "It eats up all your search quota", "search results have no content and are classed as spam", "results pages have no value"

I want to find a positive SEO output to having a search function on my website, not just try and stifle Mr Googlebot. What I am trying to learn here is what the options are, and what are their outcomes? So far I have -

_Robots.txt - _Remove the search pages from Google

_No Index - _Allow the crawl but don't index the search pages.

_No Follow - _I'm not sure this is even a valid idea, but I picked it up somewhere out there.

_Just leave it alone - _Some of your search results might get ranked and bring traffic in.

It appears that each and every option has it's positive and negative connotations. It'd be great to hear from this here community on their experiences in this practice.

[JordanLowry]

Hopefully that helps you some I know we ran into a similar situation for a client. Good luck!

[Mark_Elton]

Great idea! This has triggered a few other thoughts too... cheers Jordan.

[JordanLowry]

I would recommend using screaming frog to crawl only product level pages and export them to a csv or excel doc then copy and past your xml sitemap into an excel sheet. Then from there I would clean up the xml sitemap and sort it by product level pages and just compare the two side by side and see what is missing.

The other option would be to go into google webmaster tools or search console and look at Google Index -> index status and then click the advanced tab and just see what is indexed and what all is being blocked by the robots.txt.

[Mark_Elton]

@jordan & @matt,

I had done this, this was my initial go-to idea and implementation, and I completely agree this is a solution.

I guess I was hoping to answer the question "can Google even use site search?". as this would answer whether the parameter even needs excluding from robots.txt (I suspect they somehow do, as there wouldn't be this much noise about it otherwise).

That leaves the current situation - Does restricting google from searching my internal search results hinder it's ability to find and index my product pages? I'd argue it does, as since implementing this 6 months ago, the site index status has gone from 5.5m to 120k.

However, this could even be a good thing, as it lowers the Googlebot activity requirement, and should focus on the stronger pages... but the holy grail I am trying to achieve here is to get all my products indexed so I can get a few hits a month from each, i'm not trying to get the search results indexed.

[MattAntonino]

Agree with Jordan - block the parameter for search in robots.txt and forget it. It won't bring search traffic in, it shouldn't get crawled but if it does, it's always a negative.

[JordanLowry]

I cant speak for everyone but generally we like to robots.txt the search pages. I would imagine since you are working on a large retail site you would want to ensure your other pages get indexed properly so I would imagine blocking the search pages with a robots.txt would suffice. I would also look for some common reoccuring searches through the site search to possibly build content around as well.

I hope that helps some.