Welcome to the Q&A Forum

We have a site that has extremely dynamic content.

Every day they publish around 15 news flashes, each of which is setup as a distinct page with around 500 words.    File structure is  bluewidget.com/news/long-news-article-name.  No timestamp in URL.

After a year, that's a lot of news flashes. The database was getting inefficient (it's managed by a ColdFusion CMS) so we started automatically physically deleting news flashes from the database, which sped things up.

The problem is that Google Webmaster Tools is detecting the freshly deleted pages and reporting large numbers of 404 pages.  There are so many 404s that it's hard to see the non-news 404s, and I understand it would be a negative quality indicator to Google having that many missing pages.

We were toying with setting up redirects, but the volume of redirects would be so large that it would slow the site down again to load a large htaccess file for each page.

Because there isn't a datestamp in the URL we couldn't create a mask in the htaccess file automatically redirecting all bluewidget.com/news/yymm* to bluewidget.com/news

These long tail pages do send traffic, but for speed we only want to keep the last month of news flashes at the most.

What would you do to avoid Google thinking its a poorly maintained site?

What is the most accurate/achievable way of tracking data about subscribers to your RSS feed through Google Analytics?

With standard WordPress sites, we place the RSS  link to Feedburner so we could track statistics.

However it wouldn't track the way that I use it.

I use Pulse on an Android Tablet to read my feeds offline on the bus each morning. At home, Pulse automatically downloads the latest feeds wirelessly overnight. So then I can read them without a connection.

The obvious downside for my reading experience is that I only get what is contained in the feeds. If the company only includes an excerpt, it's too annoying to read the teaser and be unable to connect and follow a link.  So I only subscribe to feeds that contain the full post. Yeah to seomoz, aimclear, SEL, adwordsblog. I  dont subscribe to bruceclays blog, much as i'd like to, because it doesn't contain the full feed. That's probably deliberate on their part, because I have to consciously visit their blog on my desktop at work, to see the whole post.

The other problem with say Pulse, is how it locates the feed. I typed in the URL, and Pulse subscribed me. I assume that Pulse simply looked for the domain.com/feed URL and added that, rather than look for feeds2.feedburner.com/domain.

I looked at Feedburner stats and they didn't go up for 2 days, so basically it didn't track me.

Would it be as simple as using the Google URL builder to add parameters to each post in the RSS feed?   Eg utm_source=feedreader, utm_medium=rss, utm_campaign=tracking.

But that still wouldn't track offline users.

I assume that most people are also not going to paste the Feedburner URL into their FeedReader, but would let the platform auto-detect the feed.

Any suggestions?

hi

We were looking at the Google Analytics for one  of our sites and noticed that there were NO pageviews from device=iphone and resolution=640x960 in the report.

Given that iphone4 is supposed to be 640x960, and would be the most popular device (at least in our offices and everyone I know), it seems wierd.

I sorted the Mobile Devices report by device and resolution to see what was available. The first 160 results were all device=not set.

Finally got to device=iPhone and there were three entries:

resolution 0x0 had 11 views

resolution 320x396 had 45 views

resolution 320x480 had 3,944 views.

Hopefully all iphone4 users havent been classified as not set.

Or is it possible that iphone4s claim to be 320x480 in browsers, as per http://www.alistapart.com/articles/a-pixel-identity-crisis/

Even worse, if I look at the Samsung Galaxy S II (myown phone), there are over 30 screen resolution combinations.

Does anyone have anything to shed on this? I asked about it on the google analytics twitter account last week but havent had a response.

Are there other analytics solutions that would distinguish between the iphones?

Warning - this is a link to a large image, with the not set stats at the top.

6Sjji

We do also own the image generating site, so we could add it to our central Webmaster Tools account.

It's used by a large number of other client sites.

My usual preference is to verify by uploading a file.

Do you know if we would have to reverify the imagesite for every other client site that uses it?

The main issue is that we want the images to be indexed, and for the resulting traffic to be sent to the client site - not to the image site.

Has anyone got experience in using image indexing this way with xml files, and proved that traffic does come through?

cheers

We often use images from a dynamic image server, to cater for image asset management, multiple sizes etc.  Similar to Adobe's Scene7, although that came out after we chose our product.

So widgets.com may have all its images with a URL like  myimageserver.com?src=widgets/catalogue/productname.jpg;width=100;height=100;angle=90,  instead of widgets.com/images/productname.jpg.

A potential client wants to capitalise on traffic from Google image search.

We could build an image sitemap, listing the images on each page.

What should we place in the image:loc field? The code: myimageserver.com?src=widgets/catalogue/productname.jpg;width=100;height=100;angle=90

Apparently we have to verify the hosting site in Webmaster Tools, but I can't see where.

I have a single Webmaster Tools account, where most of our sites are defined. Would I have to verify it for each site using the dynamic imaging program?

Love to hear from anyone with solutions for getting images indexed when they are hosted at third-party dynamic imaging programs.

hi Keri

The problem went away after 3 days without us doing anything much. But the client was pretty antsy during those 3 days.

I had posted on the Google forums, and one guy had suggested that there were problems with the dns  - the intodns.com tool highlighted some.  Possibly missing dns servers - there was unusual stuff in the intodns report.

We don't manage the nameservers for the siite with the problem, so there wasn't anything we could do.

The site was recrawled, possibly as a result of resubmitting the nameserver, and it was ok.

thanks

chris

We have several sites on the same server.

On the weekend we relocated some servers, changing IP address.

A client has since noticed something freaky with the meta tags.

1. They search for their companyname, and another site from the same server appears in position 1. It is completely unrelated, has never happened before, and the company name is not used in any incoming text links.

Eg search for  company1 on Google.  Company1.com.au appears at position 2, but at position1 is school1.com.au.   The words company1 don't appear anywhere on the site. I've analysed all incoming links with a gazillion tools, and can't find any link text of company1, linking to school1.

2. Even more freaky, searching for company1.com.au at Google. The results at Google in position 1 for the last three days has been:

Meta Title for school1 (but hovering/clicking actual goes to URL for company1)
Meta Description for school1
URL for company1.com.au

Clicking on the cached copy of result1, it shows a cached version of school1 taken on March 18.

Today is 29 March.

Logically we are trying to get Google to spider both sites again quickly. We've asked the clients to update their home pages. Resubmitted xml sitemaps.

Checked the HTTP status codes - both are happily returning 200s. Different cookies.

I found another instance on a forum:

http://webmasters.stackexchange.com/questions/10578/incorrect-meta-information-in-google

Any ideas?