homepage Welcome to WebmasterWorld Guest from 184.73.87.85
register, free tools, login, search, subscribe, help, library, announcements, recent posts, open posts,
Pubcon Platinum Sponsor
Home / Forums Index / Google / Google News Archive
Forum Library, Charter, Moderator: open

Google News Archive Forum

This 38 message thread spans 2 pages: 38 ( [1] 2 > >     
Small Site filter?
martinibuster




msg:218577
 6:28 pm on Apr 9, 2004 (gmt 0)

In another thread someone made mention of Google's Mom & Pop Filter.

Hmm... That's a new one on me. I've optimized a couple Mom & Pop websites and they're doing well in the top five across multiple cities and counties.

Sorry, but in my experience there is no Mom and Pop Filter.

 

Jakpot




msg:218578
 6:44 pm on Apr 9, 2004 (gmt 0)

Great! Good news

bcolflesh




msg:218579
 6:50 pm on Apr 9, 2004 (gmt 0)

A search for mom and pop returns almost 2 million results, so I say nay ;)

ciml




msg:218580
 6:51 pm on Apr 9, 2004 (gmt 0)

The keyphrase penalties (think Florida/Austin) have hit 'mom & pop' sites much more than previous widescale penalties did (such as the crosslinking PR0 penalty).

Some of the types of pages/sites hit by this have not practiced agressive SEO, at least not compared to many of the sites hit by previous Google penalty campaigns. Some people refer to these as 'mom & pop' penalties; others refer to them as 'money word' penalties. When something applies as widely as these changes do, people tend to feel victimised.

In my opinion, the safest description is 'keyphrase penalty' as it applies to corporate sites, mom & pop sites, money sites, etc.

thumpcyc




msg:218581
 7:09 pm on Apr 9, 2004 (gmt 0)

I totally agree with ciml!

This is exactly what I see. For Money keywords, and combos, there is some sort of seperate filter, not pertaining only to Mom & Pop, not for just commercial, but informational sites also, it affected many revelant sites, and promoted directories.

On terms that people pay for(paid listings showing) I no longer place well in the serps on G.

I cannot reduce the words I use to describe what widgets the parts we sell are for. I have many widget parts that fit many widgets, and by describing the widget it fits, BOOM, out of the SERPS for money keywords(In G, not everywhere else). For keywords that are not considered Money Terms, and there are no paid listings, I do VERY well, and they account for the majority of traffic being referred from G these days.

At this point I still consider G broken, and hope they fix themselves soon. G is not showing the most revelant sites, and should not expect the web to change to make their SERPS better, this would be a case of the tail wagging the dog.

It's a bummer, but it is the way I see it.
Thumpcyc

egomaniac




msg:218582
 7:49 pm on Apr 9, 2004 (gmt 0)

If there's a mom and pop filter, it hasn't caught me.

I am at the top of my rankings for some very competitive keywords. Back during the Florida storm, I was pummeled way down for about a 1.5 mos. or so. But since then I have been back at #1,2 and 3 positions and holding strong.

Ozdachs




msg:218583
 7:50 pm on Apr 9, 2004 (gmt 0)

My experience is in synch with ciml and thumpcyc. My mom and pop city widget sites were at the top of Google SERPS for the services – and still are doing well at Yahoo and elsewhere. But, in Google, our pages talk too much about widgets so we are now not coming up. This is the case for several different cities and different widgets.

Instead, directory pages from Verizon are at the top at Google. Or, a movie review page is at the top apparently because one of the movie characters is a widget seller. Or, a jobs page is up top because one of the job openings is for a widget maker and another job on the page is available in the city.

These SERPs are not what I would want to find as a searcher for widgets in the city. Apparently, other searchers are not happy, either, because we’ve seen a significant increase in the Adword click throughs. Since we haven’t upped our bids or even tweeked our ads, I think we could be getting more clicks from frustrated city widget seekers.

So, I don’t believe in a mom and pop filter. I keep my faith in sure things like the Easter Bunny. But, I do think the algorithm for Google rankings is, ahh…. “unpredictable” at the moment.

jackson992




msg:218584
 7:55 pm on Apr 9, 2004 (gmt 0)

So where is a list of the filtered phrases:)

skipfactor




msg:218585
 8:27 pm on Apr 9, 2004 (gmt 0)

Mom & pop are still doing it; sometimes they just require more Viagra for 'money terms', and it takes a while to kick-in.

ILuvSrchEngines




msg:218586
 11:58 pm on Apr 9, 2004 (gmt 0)

>Instead, directory pages from Verizon are at the top at Google. Or, a movie review page is at the top apparently because one of the movie characters is a widget seller. Or, a jobs page is up top because one of the job openings is for a widget maker and another job on the page is available in the city.

They are at the top most likley because they are listed in the DMOZ. The mom-and-pop filter has a high reliance on DMOZ for the top results. If you are in the top results for a money phrase you will likely be nuked (sooner or later) by the G mom-and-pop filter because you are not in the DMOZ. Fairly simple logic really. Logic that no doubt gets some help by guys sending emails to Google complaining about mom-and-pop being listed over them. (accuse mom-and-pop of spamming or whatever to pressure G to change code and move huge directories to the top, adword pressure as well)

You will notice that Google has absolutely NO PROBLEM returning key money word phrases with adwords. ;-)

hutcheson




msg:218587
 12:49 am on Apr 10, 2004 (gmt 0)

I don't believe in the mom and pop filter. I think Sergei spends his lunch hour reading these forums and manually lowering the page rank of detractors.

I'd like to try a little experiment.

It's almost 8:00PM Central Standard Time. I'd like people to sticky-mail me URLs of these so-called "mom-and-pop" places that are getting filtered because they aren't in the ODP. I'll check around 9:00 and look at the first ten URLs sent. As a small incentive to participate in this experiment, if at least five of the sites are actually listable, I'll list at least one of them. If there aren't ten URLs sent within an hour, I'll extend the time. And I'll report gross results here (3 affiliate doorways, 2 mirror URLs, 5 listable, 2 listed). I won't mention URLs here.

This offer void if it is deemed to violate the forum guidelines: it is a bit unusual.

skipfactor




msg:218588
 1:10 am on Apr 10, 2004 (gmt 0)

>>I think Sergei spends his lunch hour reading these forums and manually lowering the page rank of detractors.

Sounds like the epitome of a Mom & Pop filter to me: I KNEW IT.

>>The mom-and-pop filter has a high reliance on DMOZ for the top results. If you are in the top results for a money phrase you will likely be nuked (sooner or later) by the G mom-and-pop filter because you are not in the DMOZ.

Horse-hockey: I have M&Ps kicking-tail with & without ODP listings, and I have M&Ps getting their butts kicked with & without ODP listings.

stevenb 1959




msg:218589
 1:28 am on Apr 10, 2004 (gmt 0)

I believe to survive mom and pop filters, websites have to work together, similar to pages of a director to gain relevancy in search engines , now and especially for the future. Thats why I am getting one link at a time and taking my time to do it right so I strengthen my associate linker as their strength strengthens me.
Anyone have the same ideas, sticky me as I am sure we can build something beneficial for both.

hutcheson




msg:218590
 2:09 am on Apr 10, 2004 (gmt 0)

Offer to check mom-and-pop-filtered sites is extended till midnight CDT, that's three hours from now.

Ozdachs




msg:218591
 2:41 am on Apr 10, 2004 (gmt 0)

They are at the top most likley because they are listed in the DMOZ. The mom-and-pop filter has a high reliance on DMOZ for the top results. If you are in the top results for a money phrase you will likely be nuked (sooner or later) by the G mom-and-pop filter because you are not in the DMOZ. Fairly simple logic really. Logic that no doubt gets some help by guys sending emails to Google complaining about mom-and-pop being listed over them. (accuse mom-and-pop of spamming or whatever to pressure G to change code and move huge directories to the top, adword pressure as well)

... except most of my Disappeared are in DMOZ and have been in DMOZ. I have no idea if the main pages of the directory are also in DMOZ but they are not in the directory for the “money phrases”.

And, my moms and pops are not spammers. And, they are not banned as spam sites. They can be found by some text or company name. They just don’t show up in the SERPs for the appropriate topic (“money”) keywords.

steveb




msg:218592
 6:05 am on Apr 10, 2004 (gmt 0)

To the question in the thread, there isn't a Mom and Pop filter. That's just one of the most goofy of the goofballisms floating around. The key question really is how much longer this era of Mom and Pop sites doing so well is going to last.

Mom and Pop (who know how to do sane seo) ride high in most sectors, due to quality + seo beating quantity + spam, but eventually mega quality sites are going to exert themselves far more than today, especially given the long url indexing the engines are managing now. These are the glory days of the Mom and pop. Enjoy them while they last.

hutcheson




msg:218593
 6:40 am on Apr 10, 2004 (gmt 0)

Thanks, all who sent URLs. I got URLs from six people, and thoroughly scratched my curiosity.

First, I got at least 6 definitely listable sites in as many distinct areas, almost all arguably true "mom and pop" businesses (two weren't, but the sites were not dissimilar from the M&P business sites.) And an interesting cross-section of M&P's they were.

Then the sites definitely hadn't disappeared (Google still included multiple pages from them) So, to borrow the tinfoil-hat language, it was a "penalty", not a "filter".

But -- and don't throw brickbats, people -- the Google top search results for the apparent keywords looked mostly very good to me -- a mixture of commercial, informational, and resource sites, all generally relevant in some way. And the sites were, NONE of them, what I'd call "necessarily top ten" sites. They were "small relevant sites" -- the kind that all too often gets pushed out of positions 100-200 by the Amazon pages. It was an accident that any of them would have ever been top 20. (But, on the other hand, they wouldn't have made Google look really bad if they had been in, say, positions 11-20.) So, in my view, Google results are as reasonable as, or more reasonable than, before the cataclysm.

Also, one site that had plummeted (excluded from the comments above) WAS in the ODP, just not in the directory that was considered most important.) Another site wasn't in the ODP, but was in the Yahoo! directory, which is a comparably valuable link.

This is an important counterexample. Also, I took a quick look around the OTHER unreviewed sites in each area. One of THEM sites had retained its top ten position in the Google results, even though it also was waiting for an ODP listing.

In keeping with the deal, I listed several of the sites, pushed others towards a listing, and also did some other editing (of sites not in the experiement) in the neighborhood of each involved category.

Conclusions:
(1) In some areas, neither M&P nor global corporations have a chance: it's all affiliate spam. Here, directories are a surfer's only hope, and doing a good job of incorporating directory data is critical. Google's doing better than anyone ever has before, which may not be good enough.
(2) In some moderately competitive areas, the relative relevance of 190 of the top 200 sites aren't really distinguishable. All the, say, commercial plumbing sites in Jersey City are as-near-as-no-never-mind equally relevant for "Jersey City Plumber". So if a change in the Google algorithm completely rearranges the top 500 sites, Google ISN'T GOING TO CARE! 400 of them are "good enough"; the surfer gets a local plumber and is happy. So Google is, I think, going to continue blithely doing things that completely rearrange the small niche business sites, if they see that doing so cleans up some major spam problem.
(3) I still think there is no specific "DMOZ filter." What I think much more likely, based on the fact that mathematicians are running the show, are things like:
-- tweaking of any of 100 or so onpage factors (probing for overoptimization, perhaps? I might add that one of my noncommercial sites seemed to be depressed for awhile, but has since rebounded higher than before, with no changes made to the site. By its nature it was all links and anchor text -- with most unusual word distributions. Was Google tweaking in that area?)
-- correlation between onpage and offpage relevance factors (defense against googlebombing, perhaps -- has anyone noticed whether that seems to be an issue these days?)
-- quasi-theming, based on the relevance of the pages containing the inbound links (or conversely, the relevance of the pages pointed to by the outbound links.)
-- deliberate random churn? (I've argued for a long time that there are both social and mathematical reasons why Google needs to make random changes in its ranking algorithm to improve results. Google doesn't listen to me, so far as I know, but it's always possible they have someone on staff as bright as I am.)
-- Something more esoteric? A real algebraist can do things with matrices that will make your head swim. I am VERY certain simplistic rules of thumb don't figure -- it's a big hairy formula with 100+ factors, and getting more complex. Five-state logical decision trees can explain the behavior you're seeing on any five of your personal sites; but they don't begin to explain the behavior of more realistic samples.

(4) If any of this makes sense, then I'd expect the variability of results to continue: particularly for small businesses that by nature don't attract lots of inbound links.

(5) Looking for good inbound links is going to stay important, whatever form pagerank/hilltop/whatever takes. LOCAL inbound links may be critical. The ODP Regional categories are your friend, if you'll treat them nicely. Things like the local Chamber of Commerce directory. etc., may end up being especially important. (But they won't save you from month-long dips into the deep end of the search result pool. One of these sites, remember, had a local DMOZ listing.) A VARIETY of kinds of links may turn out to be more important than the PR of any particular link.

(8) Big sites have an inherent advantage. Directories favor "large, stable" sites. Search engine spiders favor pages accessible through lots of different routes, which comes naturally in a large navigable site.

(9) (This is a bit of a stretch, but while I'm mumbling) Page rank hoarding just doesn't pay. Be an authority, and don't waste haggle over reciprocity. Yes, sometimes it leaks out, but Google is better at detecting artificial link patterns than you can imagine. (On the other hand, detecting similar pages like different-templated affiliate mirrors, will, I fear, remain beyond its reach.)

percentages




msg:218594
 6:54 am on Apr 10, 2004 (gmt 0)

>These are the glory days of the Mom and pop. Enjoy them while they last.

I think the glory days in Google were before March 2003, but you are right in as much as these days are not as bad as it is going to get.

"Mom and Pop" are on the wrong side of the Google bell curve now. The web is maturing slowly, the big players are circling the wagons and IMHO unfortunately "Mom and Pop" are going to be the losers.

But, some "Mom and Pops" grew in the glory days and are no longer of that ilk, they made hay while the sun was shining and can now feed on it. Yahoo still offers a last chance for the "Moms and Pops", I wonder how many will take it?

ciml




msg:218595
 11:33 am on Apr 10, 2004 (gmt 0)

Sorry, I think I should clarify a little. People refer to the mom & pop filter, the money filter, the filter that gives bonus to AdWords buyers, the filter that penalises AdWords buyers, etc.

Wide ranging penalties like these apply to all types of sites, the penalties target site/link building methods, not business sectors.

austtr




msg:218596
 1:49 pm on Apr 10, 2004 (gmt 0)

Just down the road from me is Australia's Gold Coast.. it's a wall-to-wall conglomeration of vacation accommodation properties. Think Florida coastline and you get the picture.

Nearly all properties have managers who have used their hard earned to buy the letting management rights. It typically costs upwards of a million to buy in and the annual gross on these properties is not chicken feed. Their websites are mostly designed by professionals at some considerable cost.

Yet by their nature they are managed by a husband and wife team ... ie our mom & pop sector.

You will not find a single one of these properties in a typical accommodation search on Google.... in competitive tourism and travel sectors the mom & pop operator may as well pump their advertising budgets into buying sandwich boards and standing beside the highway.

Their own informative sites have been reduced to an irrelevance. I hadn't thought of it as a filter, but rather a consequence of not being able to exist in the company of big sites able to manipulate Google's dependancy on links and anchor text.

So when debating the viability of mom & pop sites, it needs to be in the context of the business activity.

Midhurst




msg:218597
 2:30 pm on Apr 10, 2004 (gmt 0)

Hutcheson,
What is your definition of quasi-theming?

ciml,
You refer to 'Penalties target site/link building methods' We know Google doesn't like link exchanges between disparate businesses as it is open to seo abuse. Are there other penalties in link building to beware of besides the PRO site?
I don't believe for one moment Google penalises sites for unsolicited incoming links; only if they are reciprocated; but I accept I may be wrong. What do you think?
Regards
Midhurst

OldFatherTime




msg:218598
 2:42 pm on Apr 10, 2004 (gmt 0)

Well I've been either filtered or penalised (upon reflection, almost certainly filtered) The site is PR5 over 90% of its pages - has a fair amount of graphic content (owing to the nature of the product) but is well spidered by both text and image Gbots.

However, where the quality of my content had me always between P1 and P5 for my targeted phrases (and the P1 was consistently putting an Amazon product into P2.. :) ) I now find myself about P40. My Y! hits are way up and, over the past 2 weeks, have completely outstripped referrals from what used to be "Big-G". Frankly, I don't have the time to waste on trying to keep up with Google. I am a "Mom and Pop" size site that used to get a reasonable amount of traffic from Google - If Y! wants to replace Google as my primary provider of traffic, well, that's fine by me.

hutcheson




msg:218599
 2:51 pm on Apr 10, 2004 (gmt 0)

Caveat: this is all fuzzy thinking, you understand, no full analysis of exactly what is entailed in calculating the specified information and/or serving it. Most people are focused on page rank because they think they have a handle on relevance. I'm just trying to point out that relevance could have hidden depths.

There are three sets of factors that could be folded into relevance (1) the traditional on-page factors (2) the now-well-known link factors: anchor text, etc. (3) the not-unknown but very obscure offpage factors [linking to bad neighborhoods]. I'm suggesting that Google may be exploiting new ways of analyzing the latter two sets either by themselves or in conjunction with each other to produce relevancy factors relating to a "near neighborhood of a page", not just the page itself.

Examples of ideals that might be considered for exploitation:
(1) Deprecate page links from irrelevant pages: for very popular keywords, Google could save the FROM page's relevance, and factor, say, links to Hotel pages higher if they came from Hotel pages. This would tend to make random reciprocal links less relevant, and would tend to make links from directories more relevant -- both tendancies that have been observed.
(2) Deprecate page links TO irrelevant pages: same logic applied to outbound links. This would make directory pages much more relevant, and pages within big sites more relevant -- again, tendancies that have been observed.
(3) correlate link keywords to onpage keywords, and deprecate links containing words not particularly relevant to the page. This would, for instance, make googlebombing harder, and guestbook spamming less valuable.

I call this "quasi" theming because it gains some of the advantages of theming -- it tends to promote pages that link to similar pages, and demote FFA link pages -- while being based on calculations of the sort Google is already known to make.

This is not a specific proposal, just ideas about the sort of calculations that might be going on under the hood. My guess is that such factors (and checked correlations between them) would be easy to tweak, VERY hard to detect and analyze and therefore hard to SEO for. Some of them would be, I suspect, fairly easy to implement. They might often appear to be a filter on small sites vigorously SEO'ed with particular personal styles of SEL, or within a particular SEO idiom might appear highly dependant on links from "major" on-topic pages (i.e. directories).

The point? maybe PR isn't all that matters on links.

OldFatherTime




msg:218600
 3:03 pm on Apr 10, 2004 (gmt 0)

I dunno, my own observations lead me down the path of "Google doesn't care about the quality of your content anymore. We gave you a year or so to persuade the rest of the world to link to your content - if you failed in that task, well, tough - you're now at the bottom of the results."

Content, apparently, is no longer king. It would now appear that content is merely secondary since Google can't sell good content - it can only sell linked content.

BigDave




msg:218601
 4:34 pm on Apr 10, 2004 (gmt 0)

Of course Google does not care about the quality of *your content*, Google cares about the quality of *their content*.

What Google wants is enough sites within the top 3 pages to make the searcher happy. It has noting to do with how relevant all the other pages are, as long as some of the top results are relevant.

The time that google does care about your content is when you happen to be one of thei top results. Your content has to be good enough so that if you are at the top, Google will be satisfied that you are there.

steveb




msg:218602
 10:45 pm on Apr 10, 2004 (gmt 0)

With the thread title now changed, some of the comments above won't make much sense.

Small sites are at a disadvantage. There is no filter obviously, but the inability to generate significant anchor text to your own pages is a definite disadvantage.

elgrande




msg:218603
 1:09 am on Apr 11, 2004 (gmt 0)

There is no filter obviously

I would have to disagree with that statement.

If it's not a filter, it is a mechanism that behaves exactly like a filter.

hutcheson




msg:218604
 5:02 am on Apr 11, 2004 (gmt 0)

>I would have to disagree with that statement.

>If it's not a filter, it is a mechanism that behaves exactly like a filter.

Well, at least it's not a PAGE filter, nor does it behave like one. It may be a LINK filter, but I'm not even sure of that.

From the underside, it might behave like a penalty. But I don't believe it's a penalty either. I'd come much closer to believing it's a finer discriminant applied to links (in conjunction with either the source or target page).

How does Google measure quality? well, inbound links.

A small site doesn't have many inbound links, duh, it was only an accident or oversight (on Google's part) that they ever rank high. After all, with a thousand mom-and-pop ostrich breeders, they can't ALL place in the bottom 990 places, even though Google manages to place, um, 990 of them there -- rather arbitrarily picking which ones to place, of course -- how can you tell from a website who makes the biggest omelettes?

The problem isn't Google's treatment of small sites. The problem is Google's treatment of LARGE PLAGIARIZED AFFILIATE-DOORWAY sites (and I believe they know this.)

What's trashing the "Orlando Hotel" searches? There are ten thousand hotelnow.spam affiliates for every single hotel in the city, and each one of them has an Orlando page.

So: how do you detect them? It's not trivial. Their names have been so odious for years now, that they've been giving their affiliates CGI code to run on their own server, to avoid any visible indication that they are actually associated with hotelspam.now. There is very little text on those sites other than boilerplate Marketroiding-101 promotional gibberish, so checking for specific phrases doesn't work well. I'm not sure it's possible in general for a person to figure out whether the sites are affiliate or not.

Is this where Google's local search is going? Weed out the millions of hotel-rezzing spammers that have one page for every name in the zip code charts, even though there isn't a single hotel in the 'burg? By the simple expedient of using some mechanism other than text search to do geocoding, and doing on a SITE basis? If they can pull this off, they could remove a billion pages from their index, and still keep every single one of the useful pages they have now. (Yes, I really do think, based on actual calculated estimates from actual search results, that a quarter of the pages on Google may be that kind of spam.)

phantombookman




msg:218605
 8:01 am on Apr 11, 2004 (gmt 0)

I believe the problem is almost entirely due to being
'filtered' out for keywords/phrases.
The process is nor universal as non optimised and grandfathered sites (in my area anyway)seem largely unaffected.
The reason the mom&pop comes in I suspect is that those sites are often single subject/interest/hobby themes which brings the KWD problem

I have some niche sites which contain great content but have been dumped on certain phrases by Google.

I have managed to get some back in but the KWD has had to go below 3%!

On a specialist site this can be all but impossible especially on smaller sized pages.
Google have to sort this out because some excellent sites are being literally dumped.
Regards
Rod

Midhurst




msg:218606
 1:06 pm on Apr 11, 2004 (gmt 0)

Phantambookman
I don't follow your logic.
You say the hobby sites in your area which have not been optimized have not been affected by the Mom and Pop Filter.
What you haven't commented on is their use of keywords, their position in the Serps etc

In a sector in which I have a client site (two actually) the top player is an 'expert site' - simple, easy to load, few graphics, large HTML title, very big HTML header, larger than average text size, short pages, abundant use of the single keyword (a disease), masses of links to outside academic pages, research Gurus, etc all from within the page text (NOT as a separate links page).Quite alot of inward links.

I have two client sites in this category. One is a single disease site with a PR of 5, 30 or so incoming links to the index page and ranked #4 or5.
This is a direct competitor to the category leader which is also PR5.
What's the difference between the two? Very little; but the #1 has DMOZ listing and my site doesn't.

However, my other client site in this category is a product site with 80 pages of helpful info on 80 diseases. The site is #11 for the same keyword as in the two sites previously described.
The client is happy with that.

However, the inside pages of disease descriptions have in the last 4 weeks dropped well down the serps so overall visitors have dropped by well over 100%.
These inside pages have the individual disease title repeated in natural sentence construction throughout the text, in the header and in the title.And,the density is probably less than the category leader.
So I'm leaning towards the notion that it is the imbalance between inward links to the individual pages (none) and the use of the keyword.
In your case you opted for a reduction in keyword density (mo mention of having inward links to individual,pages); but one solution - in theory at any rate - is two gain inward links from on theme pages.
I'm left with the feeling that there is a complex set of factors in the algo, including Latent Semantic Indexing,which weighs the semantic relevance of the page in question as well as the page which is linking in.
If the outcome is a high 'mark' for the page, then the fact it has a high density of a particular keyword isn't carried against it.
This is a non techie evaluation.
I don't know enough about Inverse Documentation or Latent Sematic Indexing to say whether this is really what's happening with the algo.
Your comments would be appreciated/

This 38 message thread spans 2 pages: 38 ( [1] 2 > >
Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Google / Google News Archive
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About
© Webmaster World 1996-2014 all rights reserved