If the “Indexed URLs” in the sitemap within webmaster tools is more accurate, then why do we always get people in GWHG that show zero (or some unreal number) in that metric when clearly they do have many pages indexed? Is it more accurate, when updated, but slow to update? If that’s the case then accuracy isn’t truly accurate as they sure don’t update those stats too often.

Just wondering…

The lack of search is quite disturbing, the success of GMail & Google Reader owes a lot to their powerful search.

That said, I like what Google are obviously trying to do with this tool – to make a friendlier environment for n00b queries than Groups and increase communication with their users. I hope they develop the tool and it takes off.

I kind of agree on the indexability of the pages, but on the other hand a lot of that content is extremely short-lived — it’s used as a basis to get a discussion going and, as with this post, the actual content (answer) uses a different medium (which might get indexed or might be an answer given live).

One thing I’d like to see is a better way of linking to individual pages / questions though. The long anchor is confusing and not all CMS convert it into links properly (so I have to use something like the cool cli.gs service). Maybe that’s a part of the short life-span of a question queue as well?

Anyway, if you want to bring back some feedback to the people behind Google Moderator, check out their set at http://cli.gs/6UG72y .

John Honeck, I’m not sure about that. I haven’t seen the issue you describe. Maybe someone from the team will come by and give us more info.

John Mueller, I see your point, but my guess is that lots of people will use this for things other than short-lived talks with live answers. Looking at that scenario as a product manager, I’d say that a better solution would be to enable an event to set up a series that requires attendees to log in in some way (perhaps provide a conference code). After all, for events with live answers, being able to access the series at all isn’t all that helpful for those who aren’t attending, and in fact, the event organizers may not want those who aren’t attending to access any content from the event.

On the other hand, there are likely lots of other use cases were valuable content is being offered up and could be well-served by Google’s mission of organizing the world’s information and making it universally accessible.

andy, agreed — internal search would be useful as well. Even with the use case described above (live events), attendees may want to search to see if their question was already asked.

That “indexed URLs” number is a tricky one . It’s “of the URLs listed in the Sitemap file, how many are indexed exactly as they are listed in the Sitemap file.” The count will generally be lower than the total number of URLs indexed because there are almost always URLs that we discover which aren’t in the Sitemap file and there are some URLs that we discover which might be canonical versions of a Sitemaps URL (for example, you might list /folder/default.asp and we might choose to index /folder/ instead). These things make it hard to use that metric as a way of keeping track of the real number of URLs we have indexed.

It is, however, great at giving you information regarding your Sitemap file: If the count is much lower than what you have indexed, chances are you could improve your Sitemap file. Also, you might choose to split up the URLs on your site and create Sitemap files for logical sections of your site — in this case, you’d get feedback about how well each section is indexed, even if the URLs are not in separate folders.

John

Ah see, that’s exactly why I think it’s a *great” metric to keep track of how many of the URLs you want indexed actually are indexed. As you say, it only works if you list every URL you want indexed. If you want 10 URLs indexed and Google does indeed have 10 URLs indexed, are those the right 10? (OK, maybe 10 is a bad example because you could just look.)

As for indexing different versions of URLs, I’d phrase it a little differently. If someone puts a URL in the Sitemap and Google indexes a different version of that page, then generally Google isn’t indexing the canonical version. It’s indexing the non-canonical version that it *thinks* is canonical. This is something that site owners may want to know. Ideally, all the non-canonical versions 301 redirect to the canonical, which would reduce this problem. Also, my understanding is that the URL listed in the Sitemap is a signal (of course, one of many) in determining what the canonical version is.

In any case, the ideal situation is one in which the Sitemap file contains a complete list and Google indexes those versions and not others. This count should help site owners get to that ideal.

In theory, you’re right, Vanessa . However, in practice, I can tell you that most Sitemap files are far from perfect and therefore it’s not quite that easy or useful, at least for the sites with imperfect Sitemap files. If your Sitemap file is perfect (and hey, it’s not impossible, it just takes some thought — thinking about things that you should have thought about a long time ago anyway), then these numbers can help you, as you mentioned.

The other thing John Honeck mentioned is that sometimes the data is off (or just missing). I know the team is working on that, and it can be frustrating (and scary, say when you see “0″ as the number of indexed URLs).

Anyway, I’m glad to see Sitemaps and Webmaster Tools hasn’t lost it’s grip on you completely .

JohnMu,

Heh. Yep. I’ve certainly seen my share of imperfect Sitemap files. But I have talked to a lot of people who want a good metric for indexing and are willing to put in the time to make sure that their Sitemaps are accurate, in part to use this measurement.

Hopefully the bugs you mention about inaccurate counts will be resolved soon.

In Google webmaster tools mostly little difference between the total no. of urls in the sitemap and indexed urls by the google. In rare cases they both are same. Why the total no. urls in the sitemap aren’t indexing?