One day after I gave my thoughts on Google’s Caffeine architecture and the Mayday Update, Google announces that the Caffeine architecture is now fully implemented. They give some nice explanations of why they did this massive change and also I think they gave some additional insight (for me at least) as to how they worked the old architecture. I always had a sense that they were doing things in layers with the old architecture, but their description confirms some of my thoughts – even if it isn’t an exact analogy.
As to what this means for web site designers, it’s hard to say. When Google says:
With Caffeine, we analyze the web in small portions and update our search index on a continuous basis, globally.
I’m not sure exactly what they are saying. It could be that these small portions are similar to the layers they updated with the old architecture, or it could mean something entirely different. The downside of not doing things in layers means that as they are continually updating their data, they are essentially adding new information to their existing database. I’m not sure how they are handling removal of old outdated information, and how they will handle errors. You wouldn’t want to pile new information onto erroneous data. But I’m sure at this junction my understanding is far from accurate with what Caffeine actually is. I think Google’s comments on what Caffeine accomplishes are probably more accurate than the analogy of how they accomplish it. Also, in the end, the observable search engine results will give more of an impression of what this really means for website designers.
I also can’t help but think that the Mayday update is majorly tied in to this new architecture. It seems silly to implement a dramatically new search engine algorithm (at least as far as many large websites go) while also implementing a new architecture. Well – there can be several reasons why Google would do this. I have to think that Caffeine allows them to implement some quality signals that the old architecture couldn’t. I’m still suspicious that they are integrating click-through data and some other data from either the toolbar or Chrome to better determine user habits.
I was reading the webmasterworld thread on this topic and I always find it interesting when people who have a site that suddenly does poorly with a new algorithm start claiming that Google is broken – or that the new algorithm is horrible. For example, someone said this and at least gave a specific example – “Black Duck Beauty In The Winds”. They thought the results were horrible because almost all of them had nothing to do with the Chinese band.
I know Google has been doing some knob turning the past few days, but I have to say that search looks better on Google than on Bing. On Google the top 2 results are now about the Chinese band and the third reesult is the webmasterworld post. On Bing it looked like only 1 of the top 10 was about the band – and that was the second result.
I think people need to put things in perspective when being critical of Google and first decide if there is a better search engine for the majority of their queries.
I also expect a lot of fine tuning from Google over the next few weeks / months. They know their bread and butter is organic search, as much as the past few months have looked pretty messed up.
Google has valued “minty freshness” for a long time. To me that just means they update their results quickly and have the ability to rank new information accurately to have it quickly available for consumption. In today’s “modern” age, speediness is what it’s all about.
As to my own personal results with the new changes – overall traffic hasn’t changed much for me. I have had a drop of traffic on this blog (which has no real backlinks), almost entirely related to the only category that got traffic losing most of its traffic. The traffic was only there for about 10 days after I posted the information and I was surprised it got as much traffic as it did (given that I have no off-site SEO). The recent drop could be due to the fact that Google turned some more knobs, or it could be because the information is no longer getting a freshness boost. Time will tell where those pages will wind up.
As to my first SEO post – which I know is a competitive category: I noticed dramatic changes in the search engine results the past few days but that could be from changes with Google or because of the dramatic increase in information about Google’s new architecture and buzz around the web. My simple page got 1 visit so far on a long tail search phrase. I’m kind of surprised I even got that. When I checked that phrase in Google, I didn’t see my site anywhere in the top 50 results. I’ll have to do some more digging in my logs and analytics to see what I can find. And I’ll see if the traffic changes over the next few days.
In the meantime, as this is still just a hobby for me, I’m going to keep doing what I’m doing and enjoy watching the landscape.
I haven’t checked the indexing time for posts to my forums lately, but this blog post is already indexed and did get a few visits. I was initially surprised to see it getting any visits as it’s a pretty competitive topic with many highly optimized sites discussing it. While it is only a few views, it looks like the traffic came from the UK and India. I know that over the past month some people have reported odd results as far as geo-location goes. It doesn’t matter too much to me where the traffic comes from on this topic, but it does seem a bit odd.
One other interesting finding was that one of the search phrases used to get here was a long tail search that actually included the phrase “long tail searches”. The rest of that particular search wasn’t part of my discussion, but is part of the discussion on plenty of other websites. Not sure how that poor person wound up here.
After the first day and 3-4 visits, traffic vanished from this page. I got one visit today – coming from google.be which I believe is in Belgium. It is interesting that the traffic I am getting is all from non-US searches. I’m guessing this has something to do with their testing of different languages and trying to better understand what the users’ intentions are. I figured the initial traffic was some sort of freshness boost, which even that wasn’t much – as I’d now expect from a site with no off-site SEO. Not sure how I managed to still get another visitor – although it looks like it was from the second page of search results on the Belgium version of Google. Maybe with time some people will have ideas on how this new Mayday algorithm is functioning.