| 2:14 am on May 20, 2011 (gmt 0)|
This is one of the strangest reports that we've ever seen. It sounds like something never before seen to me. My first thought is that this is aligned with another thread: Are we entering another Panda user-engagement calibration phase? [webmasterworld.com]
One thing for sure - when it comes to Google SEO "we're not in Kansas anymore." The game has changed to the point that many (and I do mean MANY) people are going to realize they've been left far behind. And the rest will be digging long and hard, and struggling to find understandable language to even talk about what they are seeing.
I plan to start with this Google patent:
Large scale machine learning systems and methods [patft.uspto.gov]
Granted, that patent is only loosely related to what jabird is reporting in this thread. But what he is reporting is only a side symptom of a huge sea change at Google. Hold onto your hats!
| 2:49 am on May 20, 2011 (gmt 0)|
Sounds like a glitch.
Unless you don't have much *unique* content, home > continent > country > place sounds ripe for boilerplate text that Google hates. Depending on how many continents > countries and places you have it may be hard to provide unique content.
| 7:47 am on May 20, 2011 (gmt 0)|
Tedster, I couldn't see any references to this issue when I googled it, so yes it is very very strange to me! I have been off ill for much of the last 3 years, so the site was managed by my brother. He re-launched it about 2 months ago, but the new site left out much of the content I'd built and used a different url structure. Panda hit us hard and the new site earned little revenue, so we went back to the old one.
walkman. Obviously, continents and countries are finite; we have about 800 places. We have a quick summary at the top of each page, then a table of widgets relevant to the place, then some detailed text. Towards the bottom there is a link to other widgets in the same country - perhaps that is a slight boiler plate problem? Then there is a list of news headlines about widgets, then at the bottom 2 sentences of standardised text along the lines of 'widgets place, details about place widgets, widgets country'. But this is minimal (approx 35 words). Left and top nav are standardised - also about 35 words.
I should add that we also have news.domain.com - these pages are 'pure' content without the tables or any real attempt to optimise. They are indexed normally.
| 4:54 pm on May 20, 2011 (gmt 0)|
Hi jabird, I agree with Tedster this does look very strange.
I think the reason for the format of these odd listings is that Google is seeing your link structure (site hierarchy) as:
|Home -> Continent -> Country -> Location |
and using it to replace your title with "Location - Widgets"
Widgets, because you link to the home page with the relevant word. Location, because you link to your location pages as the location names. All very sensible IMO. Missing the middle navigation levels is normal when Google does breadcrumbs.
The questions remain why Google is using these link texts for your pages when you have perfectly sensible titles already, why there's no description and why there's no cache link.
This is what I might expect to happen if Google didn't fetch those URLs any more. That would happen because the pages are very deep in the site, or a little deep in a site that doesn't have many links to it. Neither of these things are the case here. It could be that Google isn't liking your site at the moment and has reduced crawl depth to a bare minimum. In industries like yours it's easy to have a lot of pages that look quite repetitive to a robot, even when they're extremely useful to human users.
Or possibly there's a /robots.txt problem. I wish Google would show /robots.txt in its cache. It would be quite useful for debugging. Maybe they don't for some security reason.
When you run cache:www.example.com type searches, Google does serve cached pages for these URLs. The cached dates all seem to be a few days ago so the cache could be lagging behind the SERPs.
Hopefully it's a temporary Google issue or a /robots.txt glitch. As Tedster points out, the game has changed a great deal over the past few months and we're seeing some odd things happen.
| 5:15 pm on May 20, 2011 (gmt 0)|
|Earlier today, I noticed that most pages on my site have been showing up in Google with just two words and the url itself. |
Isn't there any description below the title? Is the traffic normal?
| 5:20 pm on May 20, 2011 (gmt 0)|
When you said "no content", did you mean no snippet (description)? You also say that cache has disappeared? Is it for all the pages?
| 5:24 pm on May 20, 2011 (gmt 0)|
|I wish Google would show /robots.txt in its cache |
Google does show cache for robots.txt for many sites.
| 5:35 pm on May 20, 2011 (gmt 0)|
|most pages on my site have been showing up in Google with just two words and the url itself |
Just that indyank. No snippet, no cache link. You know when Google replaces the title with ODP or anchor text on a 'url-only' listing? It looks like that.
Google returns cache for some /robots.txt but not as a general rule. I think it would be nice to, but there was some reason people don't like it.
| 5:48 pm on May 20, 2011 (gmt 0)|
|Google returns cache for some /robots.txt but not as a general rule. |
you are right but it is strange that they don't do it for all sites.Not sure what influences it to show cache or not.
Interestingly many sites have title for robots.txt as "robots.txt - sitename".
In a few cases, the title is some text from the robots.txt file and for one of my sites it is - "View the robots.txt file to see if - sitename"
Google does change titles in SERPS and this has been happening for some time now.There are many factors which influence the change in page titles.In my above example, I am not sure what made google assign the text "View the robots.txt file to see if" to the title.
It is not just anchor text but a few other things influence the titles created by google. Though many have reported title changes in google, we are not sure what all influence it other than anchor text.
| 6:30 pm on May 20, 2011 (gmt 0)|
One big influence for a title change is the user's actual query terms. That's why just looking at the list that a site: operator query returns is no longer definitive.
| 6:53 pm on May 20, 2011 (gmt 0)|
/robots.txt gets into Google via links (sometimes from server statistics pages) which is why Google sometimes uses anchortext for the title. But sometimes not, as with these listings from Google itself [google.com]. How Google picks which anchor to use is a mystery to me. As Tedster says it's query dependent but I still find the choices odd sometimes.
jabird's issue for this thread is a bit different though. It looks more like Google's building breadcrumb-style titles for his pages, from internal anchor hierarchies. Taken with the missing description and cache links I would say that the URLs just aren't being fetched (even though they have cache from a few days ago).
| 11:30 pm on May 22, 2011 (gmt 0)|
Thanks for the replies everyone and especially to ciml for his assistance.
It seems to have been a problem with robots.txt getting corrupted, now the pages seem to be indexing properly again :)