• Register
• Log in

SEO Theory - SEO Theory and Analysis Blog

Algorithm analysis, Web community relationship analysis, SEO practices and techniques, industry news, etc.

• Home
• About SEO Theory
• SEO Glossary

How SEOs should and should not use Webmaster Central

Posted by Michael Martinez on March 17, 2008 in Search Engine Optimization

I rarely comment on specific SEO tools as it’s pretty well known that I consider nearly all of them to be worthless, or at least worth less than the knowledge of how to structure informative queries is worth.

Google’s Webmaster Central Tools have received a lot of attention and as I have studied blog and forum evaluations over the past few days I have realized that a lot of people are now using Webmaster Central as an SEO tool. That’s a bad thing because, frankly, Webmaster Central’s data sucks.

Now, I understand a lot of thought and hard word have gone into Webmaster Central. I don’t want to take anything away from the engineers who brought the service to the Web. I’d rather have it than not have it, but there are serious deficiencies in Webmaster Central’s reporting tools that make it almost completely worthless for search engine optimization.

Let’s a take a look at a sample of the issues:

1. The link data never matches what the Web Search interface shows
2. The indexing data never matches what the Web Search interface shows
3. The verification procedures are inconsistent and don’t always work
4. There is no way to use the reporting tools to determine which of your pages are in the Supplemental Results Index

There are some positive things I can say about the service, too. For example, I’ve recently noticed that the XML sitemaps parsing seems to be more forgiving of injections over which Web site operators have no control. I’m also glad for the capability to designate geographic targets.

And though the user interface could use some work that’s really not bound up with how you should be using Webmaster Central for search engine optimization. I’ve found only limited usefulness in the service but I would recommend to anyone that they do use it for:

1. Resolving URL canonicalization (although this can be problematic for sub-domains if the main domain is not set up to resolve sub-domains with “www.” inserted in front of them)
2. Submitting XML sitemaps
3. Checking what Google sees in your robots.txt (sometimes it just helps to look at it through a different interface)
4. Setting geographic targets (if applicable)

If you try to use Webmaster Central for anything else related to search engine optimization you’re wasting your time. Here is why:

Link data - It’s not just that the link data is only updated on an infrequent basis. Google admitted up front that it would NOT report all available link data, and despite some claims of improvement in link data coverage I can pretty much find random examples of links (passing anchor text in Google’s search resuts) that are not reported by the tool.

Furthermore, Google admitted that it would not distinguish between Suppelemental Results Index links and Main Web Index links in the reports so even if they were showing us all their link data, it wouldn’t matter. We’d have no way of knowing which links are being counted. So why bother showing us the link data at all? It’s a waste of Google’s resources and SEOs’ time.

But the timeliness (or lack thereof) of the link data is a real deal-killer. It does an optimizer absolutely no good to see linking data that is anywhere from 2 weeks to 1 month old. We need to know if those links are real. As the tool works right now, the data is reporting bogus results (and remember that Yahoo!’s Site Explorer reports bogus results, so there are no good link reporting tools).

Indexing data - You have to dig to find any real information about indexing in Google’s Webmaster Central. It’s there, what little there is, but it’s almost useless. The worst part of this aspect of the service is that it offers a perfectly useless table filled with links to Google queries. Well, most of the SEO industry may not know those queries are broken, but I do. I’d like to know the REAL story about what is indexed.

For example, have you compared the results of your site: query to an inurl: query? Depending on your domain name you may have to modify the inurl: query so that it only shows pages from your site (hint: DON’T modify it with the site: query operator). You’ll find that most domains probably don’t get the same results. Don’t assume that simple explanations will clear up the ambiguities. I’ve looked for the simple answers and they don’t fit the data.

If you look at your sitemaps report, you’ll see that Google tells you a certain number of URLs from the sitemaps are indexed. This is about as honest as the data gets. You cannot use the linking report to see what’s indexed because that is pretty much bogus data (see above).

Verification process - I love verifying Web sites, especially since I can now place an XML sitemap for one verified site on another. Of course, I don’t like having to do that and maybe Google can update its handling of XML sitemaps enough so that we don’t have to cross-map our sites.

But if you’re verifying a sub-domain on a server whose Web admin has not set it up so that www.sub.example.com resolves to sub.example.com Google chokes when you try to set your canonical URLs for the sub-domain. Yes, you can talk to the Webmaster and say, “Hey, you need to accomodate Google on this little issue” but you might hear a tirade of profanity over how stupid Google is.

Look, I don’t run the search engine, I just manipulate its results.

In any event, not every sub-domain can be canonicalized through Webmaster Central and that’s a serious drawback to the service because not every person out there buys whole domains and gets leased servers. Hosting services run their shops their own ways and it would be nice if Google were a little more flexible in this area.

In fact, because of the error messages I have encountered when verifying sub-domains, I know that Google is using at least two different authentication processes. You can verify the site so that you can see the reports but still not be able to verify the site so that you can set the canonicalization preferences. That’s so uncool.

Supplemental Results Index - This would be such a non-issue if Google would only:

1. Index every word on a Supplemental Results page
2. Allow Supplemental Results pages to pass anchor text the same way Main Web Index pages do (or stop allowing ALL pages to pass anchor text)
3. Allow Supplemental Results Index pages to appear FIRST in the search results (ahead of Main Web Index pages) if the Supplemental pages really are more relevant than the Main Web Index pages

However, Google wants to make life difficult for searchers and providers alike so we have no reliable means of determining which of our pages are in the Supplemental Results Index. If all you need for your Supplemental pages is more links, what’s the problem with telling people which pages are Supplemental? The whole “the tags were confusing people in the search results” argument is completely irrelevant to what we see in Google Webmaster Tools.

It’s time for Google to show us a REAL indexing report (allowing for the fact there are different data centers — let us just pick one as our “authoritative data center” for crying out loud) that tells us:

1. Which pages are in the Main Web Index
2. Which pages are in the Supplemental Index
3. When each page was last crawled
4. When each page was last cached
5. Which pages have links passing anchor text to them
6. Which pages are passing link anchor text to other pages

That would be more helpful than some random sampling of outdated linking data that doesn’t reflect where the links currently are. That would also be more helpful than jumbling up all the Main Web and Supplemental Index pages together, too.

But I won’t hold my breath since, as most SEOs (who share their opinions openly) are pretty much clueless about the usefulness of Webmaster Central, it’s highly unlikely that Google will turn it into a really useful resource any time soon.

Oh, and I’m sick and tired of seeing people promote Google’s date search function. You know, where you supposedly use a time frame like the last 30 days to see which of your pages have been indexed? PUH-leeze! That function don’t work, has never worked, and probably never will work.

If you’re basing your analysis of Google’s knowledge of your site and links on the rantings of SEO pundits who tell you to use Google’s date search, Yahoo! Site Explorer, or Webmaster Central’s link data, you’re in a heap’o'trouble. (And don’t even get me started on using Blog Search for linking data — holy mackeral!)