How does Google handle dynamic FAQs?

[joshz]

Where is the data being pulled? From a database? Do you have an example of what you're describing?

[NathanielB]

If you want to run a search in your FAQ, why not have the FAQ page like you mentioned at first and then have a search feature at the top so people can search for specific questions if they like?

Kind of how a forum would work or how the knowledge base works with WHMCS? For example http://demo.whmcs.com/knowledgebase.php so you would have all your Q&A on the page as well as a search feature.

The only difference you would have from the WHMCS knowledge base is that your Q&A's would open on the same page when people click the title (if that's how you still want to run it?)

[TcxAnthony]

THIS IS A NORMAL WEB DEVELOPMENT METHOD USED TO ADD INTERACTIVITY AND STYLE TO A PAGE.

You. Have. Nothing. To. Worry. About.

I agree entirely with Joshz on this one. In my business (Online Automotive Parts Store) there are FAQ's all over the place. These FAQ's serve two major purposes:To rank higher in organic search engine traffic, and to help our customers. If your customer is thinking about buying a certain item and they look for an installation to see if they can do it themselves, it's listed on the page with the item. Then they buy the item and you help your customer twice while they help you rank.

[Dr.Marie]

I think what the OP is saying is something a little different than a FAQ. He used that as an example, but I think what he's saying is this:

A user comes to the site and is greeted with a form where they can enter a question. So, they type in the form, "How do I clean green widgets?" The site then pulls from the database any information that could answer that question and creates a page containing that information.

*If* I am understanding it right, then I see two issues:

1. Google is not going to index that page unless you have linked to it from somewhere.
2. If you have a lot of pages pulling in the same data and they are all indexed then you're setting yourself up for a Panda issue with duplicate content.

[joshz]

I think what the OP is saying is something a little different than a FAQ. He used that as an example, but I think what he's saying is this:

A user comes to the site and is greeted with a form where they can enter a question. So, they type in the form, "How do I clean green widgets?" The site then pulls from the database any information that could answer that question and creates a page containing that information.

*If* I am understanding it right, then I see two issues:

1. Google is not going to index that page unless you have linked to it from somewhere.
2. If you have a lot of pages pulling in the same data and they are all indexed then you're setting yourself up for a Panda issue with duplicate content.

Not entirely true. Google can and will scan JS, AJAX, etc and it will find the linked source files via the HTML/PHP code.

In every file there has to be a reference to said source JS/AJ files which Google scans and follows.

Assuming the information is on the site and not being pulled form a 3rd party there will be no problems. None. Zero. Zip. It won't be considered duplicate content, either.

The same concept applies to sites on a CMS tool that uses non-seo friendly URLs and SEO friendly URLs. It's not considered duplicate content. Maybe before 2010, but Google knows AJAX and JS are meant to SPEED up a site and PROVIDE BETTER USER EXPERIENCE, thus, since 2009/2010 they have adjusted their spiders accordingly.

I build sites like these on a daily basis and almost all sides use some form of JavaScript.

http://googlewebmastercentral.blogsp...crawlable.html

[Dr.Marie]

Interesting. I'm wondering if we're talking about the same thing regarding duplicate content though. It's hard to explain the potential issue that I see.

Let's say there is a database that contains the following:
paragraph 1
paragraph 2
paragraph 3
paragraph 4
paragraph 5

Someone comes along and asks a question and the site creates a page by pulling the appropriate data from the database. The page is essentially,

paragraph 1
paragraph 3
paragraph 4

Then, a second query is done and a page is created that looks like this:

paragraph 1
paragraph 3
paragraph 5

Those two pages would look very similar to google and could be considered duplicates.

Or perhaps I am totally misunderstanding what the OP is trying to do!

[joshz]

Interesting. I'm wondering if we're talking about the same thing regarding duplicate content though. It's hard to explain the potential issue that I see.

Let's say there is a database that contains the following:
paragraph 1
paragraph 2
paragraph 3
paragraph 4
paragraph 5

Someone comes along and asks a question and the site creates a page by pulling the appropriate data from the database. The page is essentially,

paragraph 1
paragraph 3
paragraph 4

Then, a second query is done and a page is created that looks like this:

paragraph 1
paragraph 3
paragraph 5

Those two pages would look very similar to google and could be considered duplicates.

Or perhaps I am totally misunderstanding what the OP is trying to do!

Incorrect. In your example those pages would be GENERATED by the WEBSITE per the users QUERY.

That's a DYNAMIC PAGE that was CREATED BY A USER SEARCH. What happens then? The page goes away when the user goes back or closes the window. That set page of results would NOT BE INDEXED BY GOOGLE as it's a temporary page generated by the website to bring up results.

Similar to Google. Does Google index every SERP page? What you're describing is the same programming parameters as a site 'Search' feature.

So, being said, the 'archived' or preloaded FAQ's would be indexed; however, the SEARCH RESULTS PAGE (acting from OP's example [Question ______________]) would never be crawled or indexed as their dynamically set per the users sessions. Assuming we're pulling the data from the said websites database.

[SteveA]

I appreciate all the comments coming in. We've drifted a bit from the original post but hopefully we're throwing off useful information that will make the thread interesting to other people.

Let me try first to clear up the "Just Ask" function. The database behind it will have answers to perhaps 100 basic questions relative to the site's subject. In this database the answers will be written using tokens, so that for example Token_get = [buy | get | purchase | obtain | pay for]. The question will be pre-processed so that it is expressed in tokens ("buy" => Token_get) and then the database will be searched for the answer that contains the most words/tokens from the question. The page created (PHP is at work here) will have the original question (no tokens) at the top as a title with the database answer (tokens removed) beneath it.

I doubt that Google will index anything on this dynamically generated page since the page won't be up when the googlebot comes calling. So there doesn't seem to be a duplicate page problem.

My problem is how to make Google appreciate all the information here so it will get excited and want to send people to me. (I WILL give the user a good experience.)

[Dr.Marie]

That set page of results would NOT BE INDEXED BY GOOGLE as it's a temporary page generated by the website to bring up results.

That's my point. If this was the type of page the OP was talking about it wouldn't be indexed unless there was something linking to it.

I think we need to have the OP come back and clarify exactly what he is trying to do.

[joshz]

I appreciate all the comments coming in. We've drifted a bit from the original post but hopefully we're throwing off useful information that will make the thread interesting to other people.

Let me try first to clear up the "Just Ask" function. The database behind it will have answers to perhaps 100 basic questions relative to the site's subject. In this database the answers will be written using tokens, so that for example Token_get = [buy | get | purchase | obtain | pay for]. The question will be pre-processed so that it is expressed in tokens ("buy" => Token_get) and then the database will be searched for the answer that contains the most words/tokens from the question. The page created (PHP is at work here) will have the original question (no tokens) at the top as a title with the database answer (tokens removed) beneath it.

I doubt that Google will index anything on this dynamically generated page since the page won't be up when the googlebot comes calling. So there doesn't seem to be a duplicate page problem.

My problem is how to make Google appreciate all the information here so it will send people to me.

Exactly. So if you're working with a batch of answers that will be rendered by user strings you have an easy solution. Simply make an FAQ archive that links the the questions. For example:

Ask your question here __________________
Or browse through our archive of known questions

Something like that. So you have the ability to 'ask', or search to create custom dynamic pages for the user or they could browse and thus, allowing Google to index the answers as well.

It would be extra work but you would get what you wanted.

Edit: Thinking about it. Why not make a call to serve up the questions from the DB on the archive page, as a static list? That way there's little work for you to do when 'importing' data and then your users can still search said dB with the ability to 'ask'. Those results from their 'answer' then would be attached to their session token and thus, voiding upon exit, eliminated any risk of duplicate content...

That's my point. If this was the type of page the OP was talking about it wouldn't be indexed unless there was something linking to it.

I think we need to have the OP come back and clarify exactly what he is trying to do.

Yes, you took one statement out of context. Read the text before and after that - it was a direct response to the duplicated content issue you brought up, not the index issue. I spent 4 years in a Computer Science degree. If you don't know development and coding then this really wouldn't make sense to you anyways.

[SteveA]

If you're working with a batch of answers that will be rendered by user strings you have an easy solution. Simply make a FAQ archive that links the questions.

Yes, that would get most of the information indexed, but it would only be one version of each question since you wouldn't put the "close relatives" all in a FAQ. I was hoping to index exact matches to the queries put into Google, so Google gets excited ("Wow! This guy answers the exact question my user entered.) But perhaps I'm going astray here and should just let Google index one version of the question and trust that they will be smart enough to send "How can I buy..." How can I get..." How can I purchase..." to me.

[joshz]

Yes, that would get most of the information indexed, but it would only be one version of each question since you wouldn't put the "close relatives" all in a FAQ. I was hoping to index exact matches to the queries put into Google, so Google gets excited ("Wow! This guy answers the exact question my user entered.) But perhaps I'm going astray here and should just let Google index one version of the question and trust that they will be smart enough to send "How can I buy..." How can I get..." How can I purchase..." to me.

That's just a better excuse to write beefy, content rich answers to offset the question's though ;-)

[Test-ok]

Since your on a roll here Josh.
I'm a little confused...Question?

Steve said:

It seems that I'm going to have to write out all of the questions and answers on a page somewhere on the site just for Google's benefit, even if no visitor ever uses the page.

I said:

IMO I would be careful, I'm not positive but I think Google doesn't like things made just for the search engine for ranking..Google could consider it cloaking.

and you say

If you're not positive then don't form an opinion that might cause the OP to panic or change his methods.

and your saying:
THIS IS A NORMAL WEB DEVELOPMENT METHOD USED TO ADD INTERACTIVITY AND STYLE TO A PAGE. And to go with that method?


Your suggesting it's normal practice to create a page and type out all the questions and answers for Google? What interaction and style will this page have with: question...answer...question...answer...question.. .answer...
all the way down the page? I would assume the page would be huge.

You think this is acceptable for what he want's to do?

I can understand Google can and will scan JS, AJAX, etc. and it will find the linked source files via the HTML/PHP code.

That's not what he said he was going to do...he wants to create a page of question and answer links so he can get google to link to em.
Well thats how I understood his comment.
Please un-confuse me.

[SteveA]

Test-ok

Let me interject something here before Josh comes back.

Yep, I don't need a FAQ for my visitors and the only reason I'm putting one in is to give Google something to find and index. I'll probably put the only link to it down in a footer with a small font so my visitors don't normally see it.

But I don't know how else to get Google to see and appreciate my content. I really do have good content and will give visitors a good user experience - the things Google looks for. I just don't know how to make Google see that. I'm certainly open to any suggestions.

Steve

[DMN Webmaster]

.......

Ask your question here __________________
Or browse through our archive of known questions

I am going to expand on Joshz's idea above,

.......
Edit: Thinking about it. Why not make a call to serve up the questions from the DB on the archive page, as a static list? That way there's little work for you to do when 'importing' data and then your users can still search said dB with the ability to 'ask'. Those results from their 'answer' then would be attached to their session token and thus, voiding upon exit, eliminated any risk of duplicate content... ;

With the above suggestion that Joshz has made, suppose you ...

Using Ajax, make the drop down that contains the static list respond as per Google Instant, ie display guestions as you type out the Question.

(edit I changed "you type out the answer" to "you type out the Question" above, Thanks SteveA for pointing this out)

Now in your database it would have to take into account which question is relevent to which answer. That could be done simply by having a memo field with record id's for the questions in the database table that contains the a relevent answer. To get the record id in that, simple have a button or clickable "did this answer your question" of some kind.

If you didn't to go to that trouble, just search the answers database table and look for wildcard matchs on word or phrase.

This would solve your problems for the search engines, (the static list as per Joshz) and actually provide functionality for the user. The answer would or course appear in a dynamically generated div on the page. This would address the duplicated content issue I would think.

edit
as an after thought, you could display a list of the most recently asked questions and their answers, This would / could be made dynamic, simply use "order by asc" based on the time stamp in the database.

example
select top 5 questions from tbl_questions where time > (date()-10) order by asc

(date() - 10 is to hopefully insure you get a result for you query)

store the answers from this query into an array
for i = 0 to 4
select answers from tbl_answers where question like array(i)
next

Now just display the question / answer pairs in your dynamic div

JMO