|"BigName MOD8765 Air Conditioner Bargains" |
After the update, it seems that each section of the title is being assessed THEN ranked.
HowDoesSiteRankFor - [brandname] = score
HowDoesSiteRankFor - [model-name] = score
HowDoesSiteRankFor - [product type] = score
HowDoesSiteRankFor - [Action words] = score
Something along those lines appears to make a lot of sense in light of Google's new "brand" emphasis.
HowDoesSiteRankFor - [factors] = score
be considered along those same lines?
As an example:
Signs of Getting Sick - Cold & Fever
The reason I ask is factors may not be related to brands but in terms of format it is similar. It is useful additional information.
I cannot believe what the world came to: it's 2010 and we are discussing if it's better to have a pipe or a dash in the title. And that's after repeated announcements that they can now read not only static but also JS generated HTML, they have data infrastructure that dwarfs that of any of the world's governments and they stored 10+ years worth of historical data.
I'm not trying to insult people who asked the question, trust me, I find myself pondering over more basic questions yet. However, if Google has not succeeded in bringing better search results with MayDay (most people agree they haven't) , they have definitely succeeded in another major task of theirs from the humble 1998 beginnings - confuse the heck out of webmasters worldwide.
By the way, I'm very happy to see that in August 2010 my traffic from Bing has for the first time in this century exceeded that of Google's! Maybe the trend will hold and we won't have to deal with this monopoly anymore?
Seems Google simply can not return relevant results and at the same time make more advertising revenue. I think they constantly fine tune the algo to a balance of revenue over relevance.
|I cannot believe what the world came to: it's 2010 and we are discussing if it's better to have a pipe or a dash in the title. And that's after repeated announcements that they can now read not only static but also JS generated HTML, they have data infrastructure that dwarfs that of any of the world's governments and they stored 10+ years worth of historical data. |
In the post above yours I wasn't asking about the pipe or dash (I know there is a dash in my example so I just wanted to mention), but yes, I did ask it earlier.
I'll remember that either one can be used.
|By the way, I'm very happy to see that in August 2010 my traffic from Bing has for the first time in this century exceeded that of Google's! Maybe the trend will hold and we won't have to deal with this monopoly anymore? |
Of course, if bing / Microsoft take over, that could be one of them "Out of the frying pan, into the fire" situations. i am sure that microsfot would like to squeeze every last penny out of web sites that they can.
Google's value of on page factors for ranking purposes seems to be so arbitrary and misguided that it is pointless to me to try to guess what they are going to favor this year, or next, or later, or not. I just write the page, write the title and put it out there and whatever it ranks for, is what it ranks for. I stopped giving a Doo-doo as to what Google thought because I never found them to be an objective judge.
My ability to find what I was searching for when using Google has decreased dramatically in the last year and I have to try all kinds of tricks and other search engines and advanced search options to find what I need.
The direction Google search has gone in the last year has been a disaster, but you can't tell a narcissistic company that. They're not going to change just because their users can't find what they are looking for. It's not about being a relevant search engine anymore.
For Bing, it's more about the relevance which is probably why they keep increasing in popularity.
This change in Google's treatment of page titles seems to have helped Wikipedia. Their pages still rank number 1 for close title matches, but now can also rank near the top even for loose matches. This is probably another example of the increased emphasis on authority and branding. Long-tail traffic that used to go to non-authority sites now goes to sites like Wikipedia.
Google's fundamental change has not been a move away from exact match in title, it's been away from match at all on words.
At least 33% of the searches I do now I can't find results that I want without adding "+" or "-" to prevent Google from adding thousands of unrelated results to my query. This phenomenon carries over to how pages rank too.
The actual words in the query, and therefore the words in the title, matter much less than what Google thinks the searcher is looking for (or should be looking for). The impact of this can be seen in longtail results, and other places.
It is still 100% an absolute tha it is better to have the exact match to the query in your title than not to, but as long as Google offers its own opinion on what the searcher should be searching for instead, the results will be poor and pages that should rank better will not rank as well.
In other words, Google has always said to build pages for searchers and not search engines. That is not true now. If a person searches for [red spinning pretty widgets] and you built a page titled "Red Spinning Pretty Widgets" that is about red spinning pretty widgets, you now have to reconsider because Google thinks that Maroon Circling Beautiful Widgetry is really what that searcher is looking for.
That's the key to most of Google's issues now. To rank well you have to build pages for what Google thinks the searcher really wants, not what the searcher has actually typed.
Unless something has changed recently Google treats a hyphen as a space and a pipe or other character as that character.
PS. a dash is longer than a hyphen.
|PS. a dash is longer than a hyphen. |
PPS - THAT kind of dash is not a legal character in a title, either - although it often does get accommodated today.
I'm seeing that Google is splitting up the the search term and treating the constituent parts being up played or down-played with weighting their inclusion in title tags.
It would appear that it looks at search volumes (or their own measure of demand) and then uprates that element and then downgrades user's other keywords which that they use to define (refine) the general topic.
I would also suggest that the scale of that reweihgting is also dependent on Google's own trust scores and over-all theming of the site.
I would anticipate this is an attempt to minimise the impact of cut-and-dice sites that are good at automating "original" content leveraging the historically successful tactic of kw stuffing title tags.
Will it work and what is the cost to non-spammy sites? Who knows?
I've noticed alot of the same changes as you guys are mentioning above, and I wonder if Google is deliberately giving somewhat vague results to terms that are targeted by Google Adwords advertisers. Essentially making people want to click on those ads because the appear more relevant than the results that are shown.
It would be very interesting to see what happens with results of a particular search phrase before and after someone places an ad exactly for it.
Google has really gone away from exact matching as everyone is mentioning. They are truly seeking quality, authoritative, value adding domains to show the most authoritative and best results. Gone are the days of direct SEO efforts for long term viability. This update has made us professionals focus more and more on quality as opposed to just relevance as I and everyone here can see in search results from Google. Bing, Yahoo, Ask are more direct.
Welcome to the forum, Joseph.
|They are truly seeking quality, authoritative, value adding domains to show the most authoritative and best results. |
This statement sounds good in theory but there is a lot of rubbish and useless content pages on "quality, authoritative, value adding domains" - I'm not talking about maliciously posted spam and intentional bending of rules, which also happens. At the very least there are pages with errors, duplicate pages, improperly formatted pages, lots of broken links etc. Moreover, the older the domain/site, the higher the authority (supposedly) but also the higher the number of problematic pages that accumulate with time.
What I'm trying to say is that the atomic element of the Web is not domain, it's an individual page. So, if you no longer care so much about exact matches on individual pages and rank by domain authority instead, you are bound to have lesser quality pages show up in your results. It wouldn't matter to a user if the rubbish page was found on CNN website or thisisthelongestexample.com website - it's still a rubbish page.
I can see how shifting focus to domains can help curb spam - you no longer have to evaluate for "spaminess" each page - save some CPU load. But this is really no way of improving the quality of results.
So, I think you cannot seek better search results and do away with keyword matching (including exact matches). You can seek a way to not get drowned in the sea of spam - also a legitimate concern of course - then you can elevate importance of trust at the expense of exactness (inventing words as I go) of matches. But is this what we are looking for in a Search engine?
|there is a lot of rubbish and useless content pages on "quality, authoritative, value adding domains" |
Absolutely true - Google has not yet achieved their intention. But knowing which way they are headed is useful and actionable information.
No this is not what we are looking for in a Search engine. That name is rapidly losing its original meaning.
"They are truly seeking quality, authoritative, value adding domains to show the most authoritative and best results."
Seeking and failing, perhaps. Two phenomenona have asserted themselves this year. The first is the return of blog/bbs/forum spam as a valuable tactic (additionally, gaining these links as quickly as possible is valued). The second is movement away from "quality of response to query". The more precise a query, the more precise results should be, but Google is hell and gone in the opposite direction. Give Google a four word query (+without +adding +plus +signs) and they will spit up an enormous amount of results that essentially have nothing to do with query because they are serving up synonyms and different spellings for all four words... even if when those four words are together they mean a complete different thing than when the words are viewed individually.
Instead of valuing genuine niche authority and relevant content on a query they are valuing mass of links and synonyms for individual words in a query. _Sometimes_ this results in finding a deep content page on a large authority domain... but very often it means finding a babbling page on domain with hundreds of thousands of rotten quality links.
|The more precise a query, the more precise results should be |
Do I ever hear a big "Oh Yeah!" from the crowd on that. Well stated.
Thinking out loud here....
|Titles such as.. "Keyword|Keywords|Purchase Keywords Here" |
seem to have lost their power.
I have noticed more and more structured sentence type titles doing well. eg - Compare the Best Low Rate keywords @ example.com
As I'm re-reading sections of the "Phrase-based searching in an information retrieval system" patent [appft1.uspto.gov], I'm thinking... and I'm perhaps summarizing badly (or thinking badly)... that the patent suggests that if the syntax of a the title doesn't include known "good" phrases, predictive of other good phrases, it's less likely that the title as a unit will survive intact in the index.
While single words might be considered phrases under phrase-based indexing, a longer title, to survive intact and show up as an exact search, might therefore need to be grammatical, or at least not too unusual. This feels like it's the opposite of how the situation used to be.
This is speculation... I haven't set up unusual word combinations in a title and put them up on the web to try this... but it may be that if something's too rare now but similar enough to something else in the index, it won't be found until more instances of the component phrases appear in Google's phrase tables.
It's clear to me that what's in the patent has already been applied in many ways... eg, with the phrase-based algo rejecting some rarely occurring phrases either at the Suggestion or at the query phase.
And if a phrase is recognized as predictive but incomplete, the patent suggests, the algo will suggest completion. As I read the patent, and it's admittedly tough going, it seems to me that phrase-based indexing might drop out some parts of titles, at least those as described above, if they are too unusual in their word combinations... and this may be at least part of why they're not being returned.
I'm trying to do a little thinking outside the square here as well - what's interesting is to crack the intent behind Google's change - it might tell us something if we look at their perspective - and that should be fairly transparent.
Was the use of Meta Title's considered too manipulative and widespread from their point of view ?
So what are they seeking to achieve by switching the emphasis? ( assuming it isn't screwed up in some way ) - I mean this is a fundamental shift in how Google asks webmasters to organise the page content in it's guideline isn't it ?
First, please let me pick a nit, Whitey. I don't mean anything personal by this - it's just that I also don't want new readers here to pick up any confusion.
The title element we're talking about is not a meta anything - it's the title ELEMENT. There actually can be a title meta tag, and the source code looks like this <meta name="title" content="Page Title Here">. Google laughs at that and it's probably not what you meant to talk about.
Now that I got that little pet peeve out of my system... thank you for indulging me...
The title element used to be one of the simpler ways to rank for a long tail phrase, no matter what level of quality was being offered. IMO the title element still matters... a LOT! But now it needs to have more reinforcement from other signals about the page and site before its power kicks in.
This looks to me like a complication from the Mayday algo change and the tweaks of that change that followed in June and July. The H1 tag got similar treatment a while ago. Google always seems to undercut any simplistic "magic button" approach to SEO whenever it starts to get out of hand. And I do understand that - I really do.
Unfortunately, as steveb posted, they don't always get the new version working all that well before they cut off the old version.
It makes sense to do this , i think , even though some will get get caught.
Perhaps there is an authority threshold in place to prevent site's falling off of the edge when it's fully switched to the new version -and maybe those sites with enough authority won't get dinged.
It makes me wonder what Google will do next to tighten the relevancy signals on a page
@ Steve B - spot on. This sums up EXACTLY my frustration with Google right now (as a searcher).
|I mean this is a fundamental shift in how Google asks webmasters to organise the page content in it's guideline isn't it ? |
This is what baffles me. What better signal as a webmaster (and *not* as an SEO) could I give as to the content of my page?
I did some very specific title tweaking a couple of days ago when this topic came up, to try to get a feel for how much or how little difference the title is making.
So far, the pages involved have been re-indexed with the new titles, but there has been exactly zero movement in the SERPs.
These were not radical changes, the pages are still relevant to the terms involved, but they are the sort of changes I would have expected to make a difference a few months ago (removing exact phrase matches for phrases that were ranking well, for example).
This is not proof of anything of course, but it certainly supports the idea that exact match in title is being given less weight than it once had.
It seems some people are drawing the wrong conclusion here. Exact match in title is still better than not having exact match in title. Writing some complete sentence is still not as good as writing a title as a title of a book or record or movie would be titled. That hasn't "changed".
What has changed is (amazingly) some words not in a query are now valued just as much as words that are in the query (!). And the importance of page title itself is less important overall. (It still is very important, but where it might be worth 114 points a few months ago it is worth 93 points now.)
So it's a bad idea to change your titles randomly, because tweaking a title can't help you with either of the above things. That is especially true in regards to the second factor.
In essence, Google's choices have been to move away from relevance... and the more precise a searcher searches, the further away from relevance Google gets. A webmaster can respond to this by making their pages more stupid (on a page about someone named Marie sometimes use "Mary" and "Mariah" instead), or just make pages that are not stupid and assume you'll still rank decently in 2010's Google-the-in-exact while also ranking just fine in the future when Google finally gets over its stubbornness and realize this change is a really and truly stupid one.
I'm not changing titles. Once in the SERPS, the CTR from exact match must be higher than for vaguely related snippets. Google still bold the searched for words so your listing will stand out.
I've seen it posited a few times here that instead of changing your titles what you can try is to get more links to the site as a whole with the different words you need to rank for.
Previously, scoring highly across a site for 'widgets' in anchor text meant you could then rank a page for 'small green widgets'. Now, scoring highly for 'small' and 'green' across the site may help.
This matches what I'm seeing on some sites and I'm going to try it out.
"The title element we're talking about is not a meta anything - it's the title ELEMENT. There actually can be a title meta tag, and the source code looks like this <meta name="title" content="Page Title Here">. Google laughs at that and it's probably not what you meant to talk about. "
Dont you mean the TITLE ***ATTRIBUTE*** rather than ***ELEMENT***
As in I could say the following ....
Any html element can have a title attribute
Just so Im clear.
This is why it's good to pick nits in this business, just to be sure everybody is on the same page or (ahem), element:
There are at least THREE title thingies - title element, title meta tag and title attribute. And to make things still more chaotic, some CMS label the H1 element the "title" in their interface.
Well - hmmm - im still not sure what to implement over and above what I have - IE the Meta Title.
Can you give a couple of syntax examples so I can nail it and put my keywords and things in there.
In the "no movements" thread - do you think the addition of this thingie will actually be worth a try? I would see it - if I can get the syntax - as another string to the bow and worth a shot.
| This 97 message thread spans 4 pages: < < 97 ( 1  3 4 ) > > |