Google's Blog Search is NOT Blog Search, Opinions

8 comments
Story Text:

Since Google launched their Blog Search early this morning, i've had chance to play around a little bit, form opinions and take stock of what a few others are saying. The first thing is: It's NOT blog search.

It's Blog Search Jim, but not as we know it...

What it is, is feed search. On the surface there's not a greatly noticeable difference for many, but the difference is there, and it's quite profound once you understand why.

Many feed items are truncated, or otherwise edited - just snippets of the complete "on site" post. This of course means that you're not getting the full picture. You're getting a pretty good slice of that picture, but it's not complete, not by a long shot.

Historical and Linkage Data
Duncan pointed out at the blogherald that the link counts are way off base. Well, Search marketers wouldn't be surprised to learn that anyway right? Google have never given full linkage data and although the reasons may be different on blog search, the fact remains the same.

the good news for the competition is that the links count sucks. It’s delivered very, very quickly, but depth of numbers just isn’t their. The Blog Herald came in at 1,322 links where as on Feedster its over 3,000 now. Other blogs I’ve tested aren’t much better. I can only guess that perhaps its because its new and their might not be the history there, or because they are indexing RSS feeds as opposed to the blogs themselves that they are not getting a full picture.

Quite right. It does suck, but the links im seeing are very, very fresh. This is good, and although you may not be able to get that ego-propping high figure fix every day, I did like the way you can use regular search operators to produce some very fresh link data.

Try

Now, what i'd like to know is why can't i combine [link:] and [-site:]? That would be really useful.

They've also added a few extra operators:

  • inblogtitle:
  • inposttitle:
  • inpostauthor:
  • blogurl:

They're all blindingly obvious to you guys, but if anyone can explain what the hell [blogurl:] does i'd be most obliged :)

ADDED: Just reading Danny's observations and he says that the [site:] command does not work (which i'd noticed) but [blogurl:] seems to be doing the same job - there, mystery solved! Danny has a bunch of other stuff in his post worth reading aswell, so take a wander over there and have a look.

Spam

Unsurprisingly enough, the blog index is pretty well littered with scraper spam. Needless to say, my old mate phil is less than impressed and neither is Damien Mulley, calling it a "monumental letdown".

Opinions Vary, Depends What you know eh?

Opinions vary in the blogosphere's response to Google Blog Search. It's very much a case of those that know about search think it sucks and those that don't, love it. Not 100% accurate but it seems like a trend to me. There's nothing wrong with not knowing about Search, for some of those folks it will be a great tool - for those that do know about Search it will be a great tool, it's just that we notice the little problems and limitations, and ask why things don't work - pesky Search folk!

Some noted opinions:

In conclusion, there's a little "rah rahing" going on amongst the unwashed, but overall people are fairly positve about the future of this tool, but disappointed with the initial offereing. The biggest complaint is the filtering of one's own site from results when doing link checks, non-working operators (operator combo's) and lack of historical data (only goes back to march).

Well, for me, guess what?

I like it. I do think it's limited, I do think it's a little bit of a letdown, but I think we could see some interesting things as they improve, and im looking forward to it. So there.

Comments

subscribe to the SERPs

Did you notice that you can subscribe to the blog search SERPs via RSS/Atom feeds? That's pretty useful.

Sure, first thing i looked

Sure, first thing i looked for. So that makes GNews and GBlogSearch - what about those damn SERPS!

fast

gotta say that puppy is fast, It's picking things up in 30 minutes from post time

Faster

It's even faster if you run pingomatic. I'm not sure which of the pings did it, but I checked for a client's blog post using a couple of keywords, and it didn't come up on the first SERP, and they automatically ping about six services whenever they post. I then ran pingomatic, waited about 30 seconds, ran the search again and they were on top of the results.

I haven't noticed folks commenting on the whole IDEA...

...of blog search, though.

Perhaps it's all been hashed over earlier and I missed it, but as I note here, I'm rather concerned about what I perceive to be the balkanization of search results based upon (an artificially imposed) structure.

It's the content, stupid. The medium is not the message.

And hell, what's a blog anyway? And when will we see separate tabs for forums, wikis, etc? I mean, heck, is THREADWATCH a blog or a forum? You laugh, but are the boundaries that sharp? Is CNET a blog in whole or in part (hey, it's got comments, RSS feeds pretty much all the way through, etc.)?

Google et al took the easy way out... delivering results based upon supposedly neat categories, when what they really should have been doing was more quickly figuring out ways to gauge searcher intent. Personalization is perhaps one way of going about this, but I'd hope the smart minds at the 'plex would come up with deeper, more creative ways of delivering people the information they want without having to partition it into little boxes. When I'm searching for information on foo, I don't want to have to search the Web, News, Blogs... I want useful info in one place.

fast

been playing with it some more, the pick up time of posts is out of this world compared to the competition. As long as they can build up some of the older stuff and broaden the number of blogs quickly this is going to be the market leader very, very soon.

thinking

Seeing as how this is really a feed search and not a blog search it kind of draws a line in the sand. For people who don't monetize and publish full feeds it's a good thing. For people who publish snippets it's not so good. Do you make shorter posts or experiment with cloaking your feed.

Where does this leave technorati? Will they be able to stand on their own legs? We know Yahoo is working on something in house that leaked out, and I can't see MSN wanting it. If Ask was to snatch it up and marry it into bloglines it might be enough to get critical mass.

First impressions

You should ping pingomatic, and probably weblogs won't hurt (but it slows the process down and becomes critical with mass submissions). Time to index is between 10 minutes and max. 30 minutes. Feeds are grabbed by Googlebot-Mozilla.

Providing snippets or summaries, the feed title can boost the item's ranking, even when the search term isn't included in the item's title and description. Keywords in the feed description OTOH seem to be pretty useless in this case. Populating the item description with a hmmmm optimized summary can work better than full content delivery. Dynamic content sites where such optimized summaries are stored in a database for use in meta descriptions and alike should be able to get completely indexed in no time.

Cloaking promoted feeds to serve Googlebot more items makes sound sense, for example to get items published past March indexed. However, I'd put no more than 100 or so items per feed, although technically there is no upper limit. I've created monthly feeds on one site, category feeds on another ... a logical (preferential topical) partition of large sites into many feeds should work best. Categorizing the items is important, since pingomatic passes pings to tagging services too.

Starting to play with blogs now ...

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.