What's the consensus on noarchive in 2008

[venti]

We have a fairly new website (about 5 months) and traffic is taking off. However, since it is designed to be very easy to crawl for search engine bots we are also facing issues of rogue bots (which gets worse everyday). We can handle the rogue bots directly but it seems some of these bots are ripping content via search engine cache.

I am curious as to how going the noarchive route would effect rankings in the big four search engines both short and long term? There are numerous posts on the noarchive option here on WebmasterWorld but most are pretty old and I am wondering if the landscape has changed.

It seems most of us feel it doesn't effect rankings but then I see posts from prominent members that state otherwise - ex: [webmasterworld.com...]

[edited by: tedster at 7:04 pm (utc) on May 13, 2008]
[edit reason] link fix [/edit]

[tedster]

First, note that you referenced a seven year old thread. From what I can see today, if a site is following Google's guidelines, then there's no negative effect on rankings. I often use noarchive on pages with time sensitive data, such as sale offers, and have never seen a problem.

However, if there are borderline spammy signs associated with the page, then a noarchive tag can be one more bad sign and possibly contribute to a ranking problem. Matt Cutts has mentioned this several times at conferences.

[edited by: tedster at 11:06 pm (utc) on May 13, 2008]

[Receptional Andy]

I agree with tedster, and to put it a slightly different way:

Only a small percentage of sites will use noarchive; so, it puts your site in a slightly different group, and one that is more likely to be closely-scrutinised.

If you have site that you think would enjoy closer examination, then go for it. Personally, i've not seen any negative effect on performance from using the tag across entire sites. Search engines are still storing all the data, so why would it matter?

[venti]

Interesting, so with the notion that they will probably keep a closer eye on you (which is fine).. Any guess as to if these people keeping an eye on you are the people that could help indicate a positive/negative trustrank (human evaluators)? Or are these the people that may only penalize you for shady practices (spam team)? Perhaps a combination of both?

[edited by: venti at 7:30 pm (utc) on May 13, 2008]

[pageoneresults]

I've made diligent use of the noarchive directive since it came out. More so these past 12-18 months. I think having that "cached" version opens up the door to some other issues that I may not want to deal with. I don't know as I've not fully investigated a few claims I've seen over the years.

I see no reason for the SEs to keep a cached copy of the page and make it available to users with one exception and that is when a site may be down.

Also, if you are using CSS and absolute positioning, the display of the cached page may not be a pretty site and, the pun is intended.

What percentage of Internet surfers actually use the Cached link?

[Receptional Andy]

Interesting, so with the notion that they will probably keep a closer eye on you (which is fine).. Any guess as to if these people keeping an eye on you are the people that could help indicate a positive/negative trustrank (human evaluators)? Or are these the people that may only penalize you for shady practices (spam team)? Perhaps a combination of both?

I'm not answering your question, but think about analysing massive datasets like these - the interesting results are from aggregated data, not individual cases. Noarchive is another filter for this data.

[venti]

pageonresults's reply brings up another question. For those of you that have implemented the tag without changes to your serps, was there a noticeable change in the click through rate of your serps (since the cached link would no longer show)?

[Receptional Andy]

What percentage of Internet surfers actually use the Cached link?

An interesting question. It's obviously going to vary based on audience and search query. I'll throw some numbers in:

150,000 B2C visitors: just under 100 clicked the cache. So for that sample, 0.066%. 100,000 B2B visitors came out at 0.25%.

I would assume the numbers rise the more downtime a site has. Anyone else have any data on cache viewing?

a noticeable change in the click through rate of your serps (since the cached link would no longer show)

I haven't seen any impact whatsoever on CTR. Of course, there's no way of knowing whether of the percentage that click the cache, how many would not click if the link wasn't available.

[pageoneresults]

Of those Cached Link clicks, what percentage were actually human? Ya see, I think that cache is being used for things that I haven't found out yet. That's why I use noarchive.

And yes, its a great alternative if your site is offline at the time. But, what if it were a Specials page and the cached version was showing last week's specials? There are definitely areas within a site that you want to noarchive. Anything with time sensitive information would be a prime candidate. This is one of those areas where you can micro-manage the process and take some control over what Google is doing.

[Receptional Andy]

what percentage were actually human

Ah, the human filtering was in place before the cache filtering. Perhaps a few clever bots in there.

I agree with you though - the cache link is a usability feature for the search engines themselves, not for your own site. And downtime - well, there shouldn't be any should there? ;)

[internetheaven]

It never occurred to me that search engines would consider noarchive a red flag. What reason is there for that? No-one has given a reason a far as I can see:

However, if there are borderline spammy signs associated with the page, then a noarchive tag can be one more bad sign and possibly contribute to a ranking problem.

Why is it a bad sign?

Matt Cutts has mentioned this several times at conferences.

Links to articles would really be appreciated here.

[pageoneresults]

Why is it a bad sign?

I don't think it is a bad sign at all. It may be a signal that if combined with other signals may raise a flag. From my understanding, in some less than satisfactory cloaking implementations, the noarchive may be used. I don't know, tis not my forte. I do believe Google are smart enough to know whether or not a poor cloaking implementation is at play.