Google Overwrites Title and Description Tags

[alika]

One site I have been involved with is a database of research findings on a particular academic topic. The database is poorly constructed, and consist of the same title tag for all 5,000+ findings

Site: Research Finding
and no description tag

Needless to say, the site's traffic has been poor for the last few years (they didn't have a budget to rework the entire database).

So I am surprised that Google now totally overwrites title and description tags -- on a consistent basis. Google is not picking up random terms within the page, but picking specific fields in the database and uses them for the title tag and description tag. The 5,000 findings that this site has now uses the title tag

Site: [Title of the study where the finding came from]
Description Tag: One sentence entry blurb summarizing the finding

I am just amazed that they do this consistently for all the 5,000 findings of this database.

[aristotle]

Google has been taking thye liberty of changing some page titles in the SERPs for a while now. Unfortunately the new titles they generate for a couple of my pages aren't as good, in my opinion, as the original titles that I gave to the pages. I'd prefer that they didn't change them.

[Robert Charlton]

...the same title tag for all 5,000+ findings

Under such circumstances, you should be pleased that Google values your content enough to dig into it and try to return it with a relevant title. It's possible that inbound links have helped Google identify what's valuable, either on an individual page or a sitewide basis.

Generally global identical titles on a domain are the kiss of death. They don't help the user and they don't help Google.

I'd take pains to get the database fixed and to make those titles unique. Pay special attention to pages which you've noticed that Google is returning. I think you would generate a lot more traffic for your site if you fixed this.

[phranque]

Google is not picking up random terms within the page, but picking specific fields in the database and uses them for the title tag and description tag.

unlikely that they are using your database.
i would look for a structural markup element or identified fragment that consistently uses these fields from the database.
something like <h2> or <div id="title">...

[louieramos]

Google also sometimes pickup meta description or snippets from dmoz & yahoo directory (if the site is listed) instead of using the ones embedded in the website.

[alika]

Yes I agree. The site is pretty lucky that Google is giving it relevant titles and description tags. It has gotten a lot of .edu and .gov links plus fair amount of media coverage, so all these must have helped the site.

But they really need to fix the problem of wrong global title tags.

The site is poorly designed, and it doesn't use H2 tags or any markup elements. What Google always picked up for the title is always the <b>Source</b> and the description tag they pick up is the one in italics.

I'm just amazed that Google is able to get the title and description tags spot on from the content, and that they do this consistently for the 5000+ pages of the site.

[Robert Charlton]

What Google always picked up for the title is always the <b>Source</b> and the description tag they pick up is the one in italics.

Not sure exactly what "Source" is. Can you expand on this, and also tell us where on your page it is. Also, where is the description on your page. This would be interesting to know. It's interesting to know that Google is still paying attention to emphasis via <b></b> and <i></i> tags, and that they're apparently parsing onpage structure on a consistent basis to find title and description matches.

In general, Google likes to return titles and descriptions with vocabulary that fits the query text. I've also seen Google use inbound anchor text, alt text on graphic links, and first text on a page that matches the query.

In your case, chances are that the page is ranking because of the Source vocabulary and description text in italics, perhaps in combination with inbound links. I'm guessing this text is somewhat close to the top of the page.

[alika]

To clarify:

I am seeing this by doing a site:www.URL.com command in Google. It is not a search for specific terms and Google is returning to me titles and descriptions with vocabulary that fits the query text.

The 5,000+ findings are presented as a 2-column page where the left column is the navigation and search options.

The main content column has the information culled from the database, and listed below is how they are listed on the page:

<i>Finding Summary (1 sentence)</i>
<p>Actual Research Finding</p>
<b>Sample or Data Description</b>
<b>Source</b> - this section includes in order of appearance, with each broken down by <br> source of the research finding such as: title of the research study, authors, journal where study was published, publication date and pages
<b> Finding ID</b>

Google uses the title of the research study, which can be found under the Source as title tag, and the finding summary as the description tag.

They display the title tag as:

Name of Site: Title of the Research Study Source

And Google shows this consistently for all the findings.

[alika]

Bing treats it differently, and searching in Bing for site:www.URL.com shows the global title tag for each of the 5,000+ finding as

Name of Site: Research Finding
but uses the finding summary as the meta description

Another interesting thing that Google does with this site is that it gives a different title tag for the printer friendly version of the findings page (yes, both Google and Bing are showing the printer friendly versions).

The actual Title tags in the printer friendly version is just the exact title tag of the site's homepage, which Bing uses.

Google changes it all to:

NAME OF SITE Top 10 Research Findings: This finding looks at ...

The site categorizes and chooses the top 10 findings for each of the topic it covers. Google automatically labels each printer friendly version as Top 10 Research Findings -- even if that finding is never included in any Top 10.