|Is History Repeating Itself at Google?|
| 1:10 pm on Jun 7, 2010 (gmt 0)|
One of the advantages of having been in SEO for a VERY long time, is you get the opportunity to observe when history repeats itself.
I was recently reminded of that truism when working with Google SERPs, NOT as an SEO, but as a user.
One recent weekend, I was playing with the innards of an old laptop, and decided to upgrade the CPU. But I needed to know which ones would work with that model laptop, and information on the manufacturer's web site was woefully inadequate.
I knew the answer was out there on the Internet. So I did what anyone would do - I turned to Google! I expected to find the answer in short order.
Instead, I spent literally HOURS trying to find what I was looking for! But it seemed no matter what search I tried, most of the results were what could best be described as "near misses".
For example, I explicitly specified "processor upgrade" as part of the search, but instead, Google gave me mostly memory upgrades, video upgrades, etc. Or it gave me the right results, but for a different make of laptop! I tried every possible search phrase I could think of. I tried +'s, I tried quotes, I tried quotes with +'s. But no matter WHAT I tried, Google kept giving me results that were not QUITE what I had actually asked for.
After wasting literally several hours with Google, I finally found the answers I was seeking - in Bing!
WTH!? In what alternative universe did BING suddenly become a much better search engine than Google?
Fuming at having wasted so much time on what should have been a quick search, I thought back at the results I had been getting. Even the Title tags on most of the results clearly indicated the pages were not good matches - how did Google's SERPs get so f'd up? And then I suddenly realized....
I had seen all this happen before.
Those of you who have been in SEO long enough, may recall many years ago (around 2003, I think) when Google made the bone-headed move of trying to "target" SEO'd sites. If a site was "too close" of a match to what was searched for, it was kicked down in the SERPs.
This would go down as one of Google's very few (and perhaps most stupid), public mistakes they had made up until then.
The problem was, white-hat SEO, and a site that was just naturally a good match to a search, looked much alike! SERP quality suffered dramatically, with a bunch of poorly targeted junk and spammy sites floating to the top (sound familiar?).
Things got so bad, people actually started switching from Google to Yahoo in droves. I was an early evangelist for Google, but even I was advising clients to use another search engine until Google got its act back together.
It was one huge nightmare for all involved - SEO's, businesses, users, even Google itself!
Google then reversed their big mistake, and things went back to (relatively) normal.
So, why the history lesson?
Looking at the failed search results I kept getting, I suddenly realized that the SERPs I was seeing NOW, look just like they did back then!
So, I put my SEO hat on, and started investigating.
And the more I looked, the more convinced I became that Google was degrading the organic SERP quality, specifically, kicking-out pages that were a great match to what was actually searched for, in favor of pages that were "somewhat" of a match.
Investigation also revealed an extra "twist" they now added.
For example, a new web page added fairly recently for a client, which we expected (based on circumstances) to come-up at or near the top, but which was taking a VERY long time to show-up in the SERPs. Turns-out it IS now showing-up - but all the way back on the FOURTH page - Ouch!
Looking at the sites coming-up above it showed this same pattern of "near-misses" - for this 3-word search, out of the Top 30 results, only 2 of those even had all 3 words in the title!
I also noticed an interesting twist - many of the top results contained a not-quite-synonym, instead of the ACTUAL word I had searched. What I mean by this is a word that was of the same general "theme" as the word I had entered, but which had a somewhat different meaning - thus "skewing" the results away from what I had actually searched for!
So then I tried an experiment - I reversed it, replacing most of the words in my search with near-miss-synonyms, ones that BETTER matched the sites currently showing-up, but that were a WORSE match for our own page.
All of those better-matching sites then disappeared...and instead of being 4 pages back, our page now showed-up #2 in the Top 10!
It was like when I asked for Apples, I get Oranges...but when I actually ASK for Oranges, Google hands me Apples instead!
And just like what happened last time this sort of thing occurred, this degrading of search quality is again producing some of the same sort of wacky results we saw last time around.
For example, for one highly competitive local commercial search, the #1 ranking is now...the Dallas Cowboys! (as much as football fans may appreciate that, my client was not very amused!) An airport now also holds a Top-10 spot. Neither has anything to do with that commercial search.
Or highly competitive commercial searches returning Wikipedia pages. Or someone's personal MySpace page. Or #1 listings being replaced by Yellow Page sites.
Looking at numerous SERPs continues this same pattern. Specific search words being arbitrarily replaced with words in the same "theme", but with different meanings. Sites with the actual search phrase in the title being replaced by partial, out-of-order, disjointed ones. REVERSE keyword proximity - search words that should go together being separated by great distances in the body text, producing some truly bizarre results (For example, the word "New" in a search containing "New York" being treated separately). Or words appearing in the anchor text of a nav bar link being a good enough "match" to get the originating page Top-10.
It's like waking up in the "Evil Kirk" universe - well-optimized, focused, highly relevant sites are getting penalized, while flaky, spammy-looking, poor-quality, not-quite-relevant sites (that even Alta Vista would have devalued) are being granted Top-10 status!
Last time, the public backlash over this nightmare was severe (In fact, if I recall correctly, Google actually had to delay their IPO, until they could clean-up the mess).
I remember the last time Google pulled this stunt, it hurt the entire industry. With the current state of the economy, our nation CANNOT AFFORD to have Google mangle the SERPs like that again.
| 4:49 pm on Jun 7, 2010 (gmt 0)|
How long did this situation go on before it was rectified?
| 5:37 pm on Jun 7, 2010 (gmt 0)|
@aspdesigner, are you describing the Florida Update of November 2003? It seems you may be, and that certainly was quite a shake-up in the search results. And the timing was a few months before Google's IPO in spring 2004.
| 5:39 pm on Jun 7, 2010 (gmt 0)|
We reached the same conclusion at different times. Google is really messed up right now and is not returning what the searcher wants but what they THINK we want and is really messing things up royally.
| 6:59 pm on Jun 7, 2010 (gmt 0)|
"Near misses" or "fuzzy serps" are a very common phenomenon with Google, and not a recent one. It typically occurs when you're searching for something very specific - Google will give you all kinds of stuff that looks a bit like it, but is not it.
Bing and Yahoo take things more literally. They are far better for specific searches. However, the Google way has its advantages eg. when you're not 100% certain what you're after.
For advanced searchers, a lot of the time you know pretty much *exactly* what you are looking for. In that case Google is not the best, as it will give you near misses ad nauseam.
(I have noticed this difference because I have made a point of using Bing and Yahoo consistently throughout the past year or so in order to be able to compare the real search experience. Most people just assume that this or that engine is "better" - which is a wrong assumption. The truth is that they're all pretty dang good, they're just very different, and you have to know which one is best for what purpose.)
| 7:31 pm on Jun 7, 2010 (gmt 0)|
|"Near misses" or "fuzzy serps" are a very common phenomenon with Google, and not a recent one. It typically occurs when you're searching for something very specific - Google will give you all kinds of stuff that looks a bit like it, but is not it. |
And I think it's getting even more so recently, and I find it very frustrating. It's as if Google has optimized their search results specially for clueless searchers who don't really know what they are looking for, can't spell, and have no idea how to search for something.
| 8:15 pm on Jun 7, 2010 (gmt 0)|
|It's as if Google has optimized their search results specially for clueless searchers who don't really know what they are looking for, can't spell, and have no idea how to search for something. |
You just described the majority of the average searching public. The ones clicking the ads making Google money.
Me thinks it's exactly what Google wanted so those who are holding out hopes of a rollback, you might as well go get in line at Walmart and wait for that rollback.
|brotherhood of LAN|
| 8:21 pm on Jun 7, 2010 (gmt 0)|
When you found your result on bing, did you re-search for it on Google, by using a sentence of text from the returned serp? Just for interest's sake, to know whether Google had it in their index in the first place.
| 8:23 pm on Jun 7, 2010 (gmt 0)|
@arizonadude I still have a hard time believing this was the intended result. If G is deliberately sabotaging their own results to the detriment of the economy so they can make more money via Adwords? I find that highly unlikely because if that type of plan was ever out in the public eye they'd end up with all sorts of anti-trust issues.
Since the rollout of Mayday about 5% of our Google traffic has slid towards Bing. Different industries are being hit by this change I think to different extents but I know we're not alone in this rather large industry as it's the talk of one of the larger tradeshows today that "something is wrong" with Google. Manufacturers aren't getting the clicks from resellers and resellers aren't getting the clicks from buyers.
| 8:48 pm on Jun 7, 2010 (gmt 0)|
|It's like waking up in the "Evil Kirk" universe |
hahahha .. I like that.
And yes I've noticed when doing searches .. not seo stuf.. but searches for what I need or information i need I get all kinds of pages and not the good ones... or I'm seeing ones that arent even in my language... got some the other day that were russian!
I did look at our search engine referrals and google is slipping and yahoo/bing is gaining. I'm talking 20%.
| 4:27 am on Jun 8, 2010 (gmt 0)|
On many searches it looks like that Google, in trying to get rid of spam, has dumped its traditional relevance and quality factors, and has - as often as not - thrown the baby out with the bathwater. But not always, and not consistently.
I'm seeing some pages/sites that I would have thought were over-optimized, at least onpage, still hanging in at #1, and other pages/sites ranking for purely semantic associations (ie, co-occurrence) for one or more of the keywords, with hardly any keyword matches at all. Lots of synonym rankings too, particularly in long tail.
The most change in results I'm familiar with seems to be occurring after the top 3 or 4 or 5 spots, but in some markets, particularly those where most of the inbound links are questionable, I'm seeing big changes in all spots, from #1 down.
Sometimes, but not always, I can sort of guess what algorithmic factors Google might have chosen to decide a particular page satisfied a search and was relevant, but I'm seeing no overall patterns either for relevance or quality... and again, no consistency throughout a search to suggest reasons for rankings. It's as though in different markets and different spots on a serp, different conditions trigger entirely different algos... and there might be one algo for the top half of the serp, and then different algos for each spot after that, with Google perhaps hoping to normalize these as it's tried to do with Universal.
In some cases, the results are completely mystifying... a page with only a few words of text on it and a crappy chart and no inbound links to speak of, will rank for a moderately contested phrase that had previously taken some work and establishment of trust to attain. Gut level page quality in some areas after the first several results gets extremely low, and in fact I've never seen such low quality pages returned by Google.
Obviously, they've erred too far in one direction and they're going to need to make some corrections... not untypical for this degree of change... but I don't think I've ever seen them so far off course.
| 6:22 am on Jun 8, 2010 (gmt 0)|
scottsonline -- I agree, the intent of this update couldn't simply be to return 'almost right' results.
That's way too vague a goal to define concisely as an algorithm
They intended to reduce rankings for a certain specific type of site. They test changes against some set of regression cases before rolling it out. However in this case, perhaps they hit too broad a swath and took out good sites along with the ones they were targeting,and didn't notice the impact on their regular regression cases.
I too was trying some comparison searches involving big brands, and even those seemed really screwed up... the brand home pages or category pages were coming up ahead of any specific results.
Tried the same searches on bing, which responded like the 'old google'. '
What I'd like is to have a way to go back in time and try the same search on pre-Mayday google
vs the new one to compare the results. That seems like a minimal kind of requirement, and one that would be easy for search companies to offer so webmasters at least have some way to diagnose problems.
Google clearly has this capacity already - they released a preview of Caffeine on a sandbox data center so we could test and prepare for it...
| 6:24 am on Jun 8, 2010 (gmt 0)|
Excellent summary Robert!
I've thought about chiming in on this thread, but you enunciated exactly what I've been seeing far better than I could hope to.
| 6:35 am on Jun 8, 2010 (gmt 0)|
bewenched - speaking of international results, I seem to be getting way more non-US searches proportionately now. Could it be the case that MayDay has only been unleashed on US users so far?
Has anyone tried comparing searches between say UK and US google to see if the mayday/random stuff only shows up in the US?
| 7:09 am on Jun 8, 2010 (gmt 0)|
"How long did this situation go on before it was rectified?"
@td22: Several months. But a good part of that was the delay before SEO's realized what had actually happened. At first, everybody figured it was just another algo change, and SEO's worked franticly, trying to make sense out of nonsense.
Eventually, people realized that the problem was not that their SEO needed fixing, but that the SERPs themselves were total crap, and we then started figuring-out what was wrong.
But back then, all of us still firmly believed Google's claimed "Don't Do Evil" motto. Once we started realizing what was happening, our initial reaction was shock & disbelief. It was like catching your wife in bed with your best friend - even though you caught them in the act, you STILL didn't want to believe it! Even with the evidence staring us in the face, it STILL took a while for us to finally accept what had happened. Some never did.
History appears to be repeating itself. The initial reaction to Mayday was the same as 2003. It was only after looking at the SERPs from a USER'S perspective, that we recognized what was happening.
@tedster: I recall the "unusual" conversation I had with a client, actually advising them to use Yahoo or MSN for searches, and looked back in my records to determine around when that had happened. But I believe you are correct.
"And the timing was a few months before Google's IPO in spring 2004."
Actually, as I recall they had PLANNED to do it at the start of 2004. In fact, some people later stated that perhaps Google timed this deliberately, in order to increase AdWords sales to inflate the company's books just prior to their IPO!
But the flack they got from this was severe. As I recall, even the national media got involved. I think even Google realized that an IPO would have been ill-advised at that point in time.
They ended-up delaying their IPO until after they had cleaned-up the mess, and the dust had settled. The IPO didn't happen until August. And even the financial disclosure paperwork wasn't filed until well over two months after they had issued the Brandy update to fix this.
"Google is really messed up right now and is not returning what the searcher wants but what they THINK we want and is really messing things up royally."
@scottsonline: Agreed. But I think it's actually WORSE than that. If Google was merely trying to give us what it THINKS we want, it would not explain the reversals - getting oranges when we want apples, and when we want oranges, now we get apples instead! It's like Google is playing a game of hide & seek!
@claus: Fuzzy results by themselves are not the problem. I don't mind a broader result set, as long as the relevant sites come-up first. The problem is that Google is pushing not-quite-relevant sites to the top, above sites that are actually a "spot-on" match to what was searched!
@brotherhood of LAN: It wasn't just a single result, it was a collection of relevant sites to my queries. But this was for a major brand laptop, I can't believe Google didn't have that same information in its index.
"And yes I've noticed when doing searches .. not seo stuf.. but searches for what I need or information i need I get all kinds of pages and not the good ones... or I'm seeing ones that arent even in my language... got some the other day that were russian!"
@Bewenched: We are seeing the same thing. Looking at it from a user's perspective is the best way to spot this. When ALL of the SERPs are messed-up, you know the problem is NOT because you need to fix the SEO on your own site!
"Since the rollout of Mayday about 5% of our Google traffic has slid towards Bing."
@scottsonline: Last time this happened, we were still getting lots of traffic, the problem was that much of it was NON-RELEVANT traffic, that wasn't generating much in the way of actual sales. Check your Google conversion rate.
"it's the talk of one of the larger tradeshows today that "something is wrong" with Google. Manufacturers aren't getting the clicks from resellers and resellers aren't getting the clicks from buyers."
Last time, feedback from actual users was what finally convinced some SEOs that the problem was actually with Google.
"if that type of plan was ever out in the public eye they'd end up with all sorts of anti-trust issues."
I was thinking the same thing.
Unlike last time, Google now holds a near-monopoly on search, They have effectively become the de facto "yellow pages" for the Internet. But with great power, also comes greater responsibility.
When Microsoft started complaining about anti-trust issues and Google, talking about Google being the "gatekeeper of the Internet", and potentially abusing that power for its own financial gain, I figured it was just the two biggest kids tussling in the schoolyard.
But what Google has to realize is, degrading organic search quality while increasing AdWords prominence on the SERPs gives Microsoft's anti-trust claims legs!
But my biggest concern right now is the impact on the economy. The SERPs now impact a much greater portion of the economy than they did in 2003. And not just online business - as Google has replaced the phone book, many local & service-oriented businesses now rely on organic searches as well!
"I agree, the intent of this update couldn't simply be to return 'almost right' results.
That's way too vague a goal to define concisely as an algorithm "
@mantucket:They were able to do it in 2003...what makes you think they could not do it 7 years later?
From a programmer's standpoint, it is actually pretty easy - you just turn some of the traditional ranking factors (like complete search phrase in Title tag) into penalties, and then just let the sites below those (which are not quite as relevant) rise to the top!
"with hardly any keyword matches at all"
@Robert: Exactly. In the SERPs we have seen, more relevant pages exist, but are now ranked lower - just like last time.
"Lots of synonym rankings too"
Same thing here, replacing one of the search terms with a synonym seems to avoid the "relevancy" penalty. The problem is, they are often not actually synonyms, but similar themed words with somewhat different meanings...so that the search results are not quite what the user actually searched for.
"In some cases, the results are completely mystifying... a page with only a few words of text on it and a crappy chart and no inbound links to speak of, will rank for a moderately contested phrase that had previously taken some work and establishment of trust to attain."
These "junk" sites floating to the surface were the same thing we observed back in 2003.
"Gut level page quality in some areas after the first several results gets extremely low, and in fact I've never seen such low quality pages returned by Google."
Again, the same thing we saw in 2003.
"I don't think I've ever seen them so far off course."
Only once, but with enough public prodding, we were finally able to get them to see the light.
Hopefully, we can get them back off the "evil" track again!
| 7:42 am on Jun 8, 2010 (gmt 0)|
aspdesigner - Your post in the Update thread prompts me to include here a paragraph that I cut out of my post above before posting, largely because I thought bringing MSN into it would complicate things. Our thoughts and vocabulary overlap enough that it's worth posting anyway...
|What the results bring to mind, in terms of quality, are the early results I remember from MSN. The big difference between the MSN results (some four years ago) and the Google results now is that the MSN results were obviously rote matches inspired by SEO 101 keyword matches. Google appears to be trying to rid itself of simplistic keyword matches entirely, and pages seem to rank as much for what they don't contain as what they do. |
I promised myself I'd get to bed early tonight, or I'd get into it more. I don't think it's possible to reject certain very basic traditional relevance signals and produce satisfactory overall results.
Apart from that... and thinking out loud... there is a question lurking in my head about how much Google is now committed to a continuum of results, as opposed to a variety of results.
| 8:14 am on Jun 8, 2010 (gmt 0)|
To focus some added attention to a little gem in your opening post aspdesigner...
|I explicitly specified "processor upgrade" as part of the search, but instead, Google gave me mostly memory upgrades |
In my corner of the web there are a lot of word combinations that mean the same thing yet are technically different. example: tri widget, tri-widget and triwidget are all the same and when I used to search for those individual terms I'd get the individual term in results as it was entered. Now when I search for variations, in my corner of the web, I get one version from google (tri widgets in this case).
example: I looked for blue triwidget containers and two of the top 10 results contain both triwidgit AND tri widget but ONLY the google choosen version of tri widget is highlighted.
An assumption, to make caffeine faster is it possible long tail searches are now less defined in serps? In other words did Google make the pool of long tail keywords much smaller by grouping terms it believes mean the same thing together? If so the net would be faster and the results less accurate as you describe. A side effect of doing this is each keyword would also have more variations in each group which means more competing sites.
From what I've gathered thus far anchor text, and the number of times it is used, in links to a specific page are being weighed much more heavily. (less quality = more quantity, it's my old math professors law). Sites like WebmasterWorld don't have a bazillion internal links pointing to each page but spam sites, and especially mashup sites, do.
It couldn't be this simple could it?
| 8:21 am on Jun 8, 2010 (gmt 0)|
Surely the whole point of a long tail search is that it is very specific and should produce very specific results - would they really degrade this ?
| 8:34 am on Jun 8, 2010 (gmt 0)|
Even a search engine with big muscles eventually becomes too bulky if they continually get bigger and more ripped. Speed is expensive on the back end, perhaps they did.
I went back and did a few more searches. It's subtle but it's there everywhere I look. Widget = Widget's, I search for widget and widget's is highlighted when before just the widget part of widget's was highlighted. To Google they are now the same. It would take a while to define groups of keywords using this method and longer still to count how many times you've used anchor text on internal links that fits the bill.
All I know for certain is that quality sites lost some traffic and spam sites (large mashup and repetitive multi sub-domain sites in particular) gained some traffic. A shift from quality towards speed/quantity would produce that type of result. Of course we're told to beef up our "authority" factor, we need to in order to combat those who got a boost via sheer volume. We're not about to be told to "create pages at a greater rate" now are we.
| 8:48 am on Jun 8, 2010 (gmt 0)|
|Widget = Widget's, I search for widget and widget's is highlighted when before just the widget part of widget's was highlighted. |
Sgt_Kickaxe - Interesting observations. Keep in mind, though, that the highlighting is done separately from the ranking, and the two may not necessarily relate.
I'm noticing, particularly on long tail searches, that many synonyms (with different stems) are highlighted in a way that previously would have happened only if you'd searched with the "~" operator. Does this mean that Google is treating synonyms differently than it used to?
I'm thinking it does, particularly because of the timing... and because those highlighted terms are often the only terms in the serps connected to the search, but I can't say for sure.
I also assume that exact matches of highlighted terms still trump approximate matches, as they did with misspellings and stems, but in this new situation I haven't made enough observations to have an opinion. More precisely than highlighting, what are you seeing?
And I did stay up later than I'd hoped. :(
| 9:47 am on Jun 8, 2010 (gmt 0)|
"A shift from quality towards speed/quantity would produce that type of result."
Kickaxe, the few sites I have are mostly pure HTML CSS sites, pages in size around 10-15 Kb, tested them with different tools and they all pass the test of spead. Yet the traffic was compeltly killed on these sites from Google side.
By delivering the near to close results I think Google is trying to apply an old philosophic logics principle that I have learned long time back in school: apple is a fruit, pear is a fruit, if I want an apple, i might actually want a pear...
Same now, it's like me looking to buy a Ferrari, and someone else tell me that I actually want a Renault (based on the principle that Ferrari is a sports car, and Renaul team won a few Formula 1 prizes, therefore Renault is a sports car, so I should consider buying a Renault). Long tail traffic is same, it's a niche, and Google should consider this is a particular market for particular people, people who know what they want, and that is not general products or info.
I will continue doing same job as before, see what happens after dust settles. Doing any extreme change now might hurt if Google goes back to prior results or anywhere close. If nothing changes, I will simply start considering registering many domains and have 1 page content sites if this is what it takes to rank again.
Either way, Google gives, Google takes!
| 11:41 am on Jun 8, 2010 (gmt 0)|
|Has anyone tried comparing searches between say UK and US google to see if the mayday/random stuff only shows up in the US? |
Its here in the UK, and not in a good way.
| 12:28 pm on Jun 8, 2010 (gmt 0)|
"It couldn't be this simple could it?"
Grouping could not explain it, as which search words were relevant or ignored were DIFFERENT for each result! Nor would it explain the other SERPs I have been seeing.
Google was simply trashing the relevant ones, and returning partial word matches as the top-ranked sites!
"Yet the traffic was compeltly killed on these sites from Google side."
@John_C: That is the same thing that happened last time - numerous relevant, white-hat sites were getting penalized. Those that were getting traffic, it was mainly non-relevant searches, so few actual sales.
@Robert: I have been looking-over some of the old stuff, and discovering even more matching factors than I had even remembered!
"Does this mean that Google is treating synonyms differently than it used to?"
I believe so. They appear to be forcing it on some searches that would otherwise return a narrow set of results (and thus be an easy target for SEO), in order to broaden the search. In many cases, the word is also not a real synonym, but just something in the same theme, resulting in changing the meaning of the search from what the user actually intended!
"I also assume that exact matches of highlighted terms still trump approximate matches"
Unfortunately, just the opposite! Exact matches are often penalized, while semi-related synonyms (or partial word matches) take priority, in either case resulting in SERPs that are often not quite what the user intended.
"I don't think it's possible to reject certain very basic traditional relevance signals and produce satisfactory overall results."
"in fact I've never seen such low quality pages returned by Google"
| 8:53 pm on Jun 9, 2010 (gmt 0)|
This type of "Relative Relevance" has increased considerably our own long-tail traffic, but has, at the same time, destroyed our AdSense CTR and eCPM.
If this proves true for most AdSense carrying sites, eventually affecting Google's ad performance in a similar manner it affects AdSense-carrying sites, this may prove the fundamental difference in Google's response between Florida and Mayday. The reason is AdSense consists today a more important factor in Google's perspective of the world, than it used to be back then at the time of the Florida mess. This different perspective provides currently a great incentive to speed up any corrections needed to Mayday in order to restore overall AdSense CTR and eCPM to optimal levels.
Will AdSense eventually save search relevance?
If so, how ironic . . . .
| 9:03 pm on Jun 9, 2010 (gmt 0)|
I got done reading ALL of Parts 1 - 3 of the Florida Update thread! (all 688 pages and 2,301 comments!) As well as part of another thread that actually made some strides determining what Florida was doing.
And I found out something amazing!
Not just the poor-quality SERPs, but ALL of the things we are seeing now...were in there!
ALL of it was also in Florida!
- The low relevancy.
- Masses of white-hat and natural sites getting knocked way down in the searches.
- The broad matching.
- Relevant players nowhere to be found.
- The top-ranking partial keyword matches.
- Pages simply mention keyword in body Top 10.
- The penalty for exact matches.
- Obvious spam and junk sites in the Top-10.
- Top-ranked foreign language.
- Pages simply LINKING to relevant sites coming-up.
(like affiliate sites)
- Top-ranked personal sites.
- Asking for oranges and getting apples.
- Alter Search to reduce title match - site moves from nowhere to Top-10.
- Reverse Keyword Proximity in top-ranked sites.
- Sports team ranking #1 for a commercial search.
- Even the synonym stemming Robert mentioned.
This seems WAY to close to Mayday to be mere coincidence.
This raises a VERY disturbing question...
Are they trying to resurrect a previously rejected algo approach?
Is Google trying a "Do-Over" of a favorite project - but without the pressure of media attention & an impending IPO?
Is "Mayday" in fact....
SON of FLORIDA?
| 4:33 am on Jun 10, 2010 (gmt 0)|
@heisje: There's already talk on another thread about switching from AdWords to Yahoo or Bing.
Back in 2003, a member here reported loosing two major AdWords clients during Florida, because they apparently checked their rankings in Google, and after what they saw decided they wanted nothing more to do with them -
|" Google blew this one as far as some of my clients are concerned....... |
I just lost two very good (large) adwords clients because they (like many of my clients) have been checking their Google, Yahoo and AOL rankings and decided (right or wrong) that Google is broken and that they no longer want to be involved with Google marketing.
Nothing I could do to convince them to wait a while -- they just said enough is enough and that Overture advertising with Yahoo and MSN was good enough.
I even asked if they would reconsider once Google got everything finished and was told definitively -- NO CHANCE, NOT INTERESTED.
This can't be good and my two clients are very likely not the only real businesses in the world whos is ticked at Google."
| 8:10 am on Jun 10, 2010 (gmt 0)|
This is not funny, but I have to laugh at the timing of finding this post. Just an hour ago I searched at Google for 'http' as a test to see results.
#1 & 2 ... of course Wikipedia (why not just put a link and get them out of every serp)
#3 ... W3.org How could 1 & 2 outrank these guys on that topic? (Oh yea, this is about money, not accuracy)
#4 is a link to a document about http dated 1999 (way to go Caffeine...)
#5 Yahoo's home page. (Google is suggesting I may have better results there?)
#6 Apple's home page. (see #5)
#7 Developer Resources for Java Technology (I guess thats what I really wanted)
#8 www.apache.org (close, but http is a protocol not a program)
#9 Our favorite again, Wikipedia
#10 httpd.apache.org (closer, but see #8)
also 'http wikipedia' was listed in the related serches. How? Why? (Oh yea, see #3 comment)
| 5:42 pm on Jun 10, 2010 (gmt 0)|
@Drag_Racer: That is TOO funny!
Right now, Wiki lost 1, but they added another W3 page. As it was actually RELEVANT, of course it goes to the bottom of the list!
Even funnier, is what it was...
an expired, working draft memo for something called "Basic HTTP", signed with the recognizable initials TimBL, and dated 1992!
I guess that's what Google means by FRESH! LOL!
BTW, the Yahoo and Apple entries also included extra links to their stock prices. Is Google trying to give us a hint? ;)