Forum Moderators: open

Message Too Old, No Replies

To Spam or not to Spam

         

stratocaster

6:45 am on Mar 9, 2003 (gmt 0)

10+ Year Member



I just checked the results for some keywords that relate to my business. What did I see? keyword-keyword-keyword.com over and over with pages full of affiliate links and no content at all. It seems useless to try to create a brand when you have no chance of obtaining a good listing unless you have keywords in your domain. I have spent over 200k in the past year promoting my brand and im being beaten by kids creating spam affiliate sites.

Do you guys think that in the future google will return the best sites for particular keyword searches or just the best optimized sites to rank well? Should I dump my brand and go garbage?

jeremy goodrich

2:00 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Steve, that's a bit of a seperate issue...but, have you read this paper:
Anatomy of a large scale hypertextual search engine [www7.scu.edu.au] which describes in pretty good detail the 'original' Google, some stuff about their db architecture, and PageRank.

After reading that, it's pretty clear, the last thing Google's founders would have done was code in a 'bonus' for keyword.com :)

BigDave

2:00 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



stratocaster,

I am not an SEO, and I am only a mediochre webmaster. But I think you are really missing out if you dismiss everything that SEOs do as being spam.

Many of the SEO suggestions that I have picked up here have actually made my site BETTER for my users!

Quite often good navigation for the bot and the users is the same. Good Titles are good for the bots and good for the users. Limiting the bells and whistles to those few places necessary is good for both. Care taken in the copy will pay off for both.

And there is nothing wrong with paying someone to keep track of the links coming in to your site.

If you want an ethical SEO, it is really up to you to get one. If you hang a "get number one or you're fired, I don't care how you do it" on them, you will get a spammer. If you suggest the areas that you believe you could use some help, link structure, keywords, etc, then you will end up with a better website.

steveb

2:12 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Of course I read that, but that is ancient history.

It's not hard to examine results and the linking associated with them. It's plain as day when all sites are linked with the same keyword and the domain-keyword one can win with far fewer links.

One interesting thing I just noticed is a site that rocketed to #1 on the strength of doorway pages and guestbook entries did so with apparently *none* of the links being that two word keyword! However, each of the *doorway* pages where the links originate are titled that two word keyword!

In other words... for a search on blue widgets, the link text is "john smith" but the links originate from a dozen pages titled "blue widgets".

rfgdxm1

2:42 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



>After reading that, it's pretty clear, the last thing Google's founders would have done was code in a 'bonus' for keyword.com :)

And as steveb said, that is ancient history. The problem is that Google isn't going to tell what is in the secret sauce. Unless someone wants to do some careful experiments, we really don't know what the answer is. However, a keyword being in the domain name does tend to be highly correlated with the site being relevant to that. Thus it seems like a reasonable factor to add in the algo. The other practical problem is that even if it is all anchor text, if you are brand.com then it stands to reason some sites that link to you will do so with your brand name. Does everyone who links to microsoft.com use the anchor text "computer software"? Sure, if by some miracle you can control the anchor text that every site that links to you uses, brand.com may be as good as keyword1-keyword2.com. However, this isn't the way it tends to work in the real world.

jeremy goodrich

2:53 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



>However, this isn't the way it tends to work in the real world.

Dude, I manage more than 50 sites. I consult on hundreds more, routinely.

Do let me know the # of sites to study before I get to the 'real world' sounds more like an MTV show that search engine optimization. :)

deft_spyder

3:00 am on Mar 10, 2003 (gmt 0)

10+ Year Member



upon further review: del

jbauder

5:12 am on Mar 10, 2003 (gmt 0)

10+ Year Member



There seems to be a pretty solid camp of keyword in domain name (other than anchor text) has no impact at google ... what about at inktomi?

Do you have the same opinion there?

rfgdxm1

5:59 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



>Do let me know the # of sites to study before I get to the 'real world' sounds more like an MTV show that search engine optimization. :)

Well, there sure seems no shortage of people whose observation is that keyword in the domain name tends to do well on Google.

danny

6:22 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



It's not that keywords-in-domain directly influences Google results. What happens is that people who are optimising for "keyword1 keyword2 keyword3" often purchase the domain keyword1-keyword2-keyword3 -- because it influence other search engines, perhaps, or because it indirectly influences Google through "URL as anchor" links, or because they think it influences Google. So keyword1-keyword2-keyword3 domains do well on "keyword1 keyword2" searches...

Correlation is not causality.

vitaplease

7:02 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



It seems that everyone has its own observations on if kw's-in-domain helps or not with Google. Of course such an overblatantly obvious approach would never be part of any serious algorithm. Its the side effects of the algorithm that make for such observations/conclusions.

In real competitive areas the kw-rich-domains do not tend to survive high rankings.

It is in mildly competitive areas where one DMOZ PR6 listing and a few other directories with the url name as anchortext makes for outranking Joebrand.com.

Do not forget that those kw-rich domains only might rank higher for those keywords, the other 100 keyphrases important to that business, face the same natural competition as Joe-brand.com (and that is why many of the short-term kw-rich domain webmasters tend to register these other keywords as urls as well - all possibly leading to easy farming/linking/duplicate anti-spam targetting).

jbauder

7:14 am on Mar 10, 2003 (gmt 0)

10+ Year Member



vita

Search [Online Widgets] Top 20 results gotta to get to 17 before you see a listing without the keyword in the url ... not saying it isn't anchor text ...

<list of domains smipped>

[edited by: Woz at 7:36 am (utc) on Mar. 10, 2003]
[edit reason] no specifics please. [/edit]

jbauder

7:16 am on Mar 10, 2003 (gmt 0)

10+ Year Member



it is also interesting none of the top 20 included "widget" in the domain ... only widgets

[edited by: Woz at 7:37 am (utc) on Mar. 10, 2003]
[edit reason] no specifics please [/edit]

rfgdxm1

7:16 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



A lot may be that most webmasters that would put up a site at keyword1-keyword2-keyword3.com likely would well optimize the on page text for that. Most multi-keyword phrases aren't *that* competitive that doing well with good on page optimization, and some off page, is enough to do reasonably well. However, try to register travel.cx (which is open according to the whois I just checked) ain't gonna be enough to make top 10 for "travel" without some serious inbound links.

jbauder

7:23 am on Mar 10, 2003 (gmt 0)

10+ Year Member



rfg ... any idea how big google is in christmas island (.cx) ;-)

vitaplease

7:28 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



jbauder,

I agree there are certain sectors that just atract certain webmasters..
(my earlier posted example allinurl: cheap)

Would a search for "widget" be more competitive though?

If you take gambling, pharmaceutical and the likes aside, and then look at searches with 2 million plus results (competitive in the sense of number of results), the key-word-rich tend to fade away.

[edited by: Woz at 7:38 am (utc) on Mar. 10, 2003]

jbauder

7:34 am on Mar 10, 2003 (gmt 0)

10+ Year Member



actually the singular searches seem to match (more of not exactly though) what you say ... I'm with you on the whole widgets industry, I'll do some more checking ...

[edited by: ciml at 11:48 am (utc) on Mar. 10, 2003]
[edit reason] No specifics please. [/edit]

steveb

7:35 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



"overblatantly obvious approach would never be part of any serious algorithm"

Clearly a mistaken idea since page title is indisputably critical, and a far more blatantly obvious thing to manipulate than domain name.

It's impossible to imagine that there is not a presumption that a page is likely to be relevant to a topic if its url and/or title is the search term. The only question is if it matters a tiny bit or a lot.

vitaplease

8:29 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



..page title is indisputably critical..

I agree, most critical.

In my opinion, in a more ideal world, the anchortext would in many cases equal the title of the to be linked to page, as that most probably summarises the page's content.

In practice people often just copy and paste the url and do not always surround the anchortext with descriptive text.

So Joebrand.com with "Joebrand the blue widget specialist" as title , looses out against blue-widget-specialist.com.

Both could have all the normal/legitimate reasons to name their url as they did.

The only question is if it matters a tiny bit or a lot.

It matters in some search queries, for the reasons mentioned above.
Luckily once dominos is known for pizza things equal out.

deft_spyder

8:37 am on Mar 10, 2003 (gmt 0)

10+ Year Member



im almost convinced you must be kidding after that one. ill take 300 inbound to asdfgafgasfdg.com over kw1-kw2.com with 50 any day.

rfgdxm1

8:39 am on Mar 10, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



>Clearly a mistaken idea since page title is indisputably critical, and a far more blatantly obvious thing to manipulate than domain name.

Right. Why give a boost for keyword in title, but not keyword in URL?

This 50 message thread spans 2 pages: 50