Welcome to the Q&A Forum

While it used to be possible to keyword stuff content to game the search engines, that is less true today. I'm a believer in LSI (latent semantic indexing) whereby the SEs understand the meaning of a page even if they keyword isn't exactly present. SEs today know about synonyms and can extract meaning from text, allowing for creative and interesting use of language. Without tedious keyword repetition the content you create will be of more interest and read better to a search user –increasing  the user experience.

I take the viewpoint that you should write for humans first, and then review the content to make sure the keyword is present at least a little bit. I would not go so far as to say you should have 1 occurrence of the keyword per 150 words, or any other hard and fast density metric. But that's just me.

Also, I think the external factors (links with appropriate anchor text, and variations of that anchor text) are more important than on-page factors like keyword density. Just my opinion, although I do rank well for many phrases and I ignore keyword density (other than making sure the keyword is present in some minimal fashion on the page that is optimized for it).

It only updates when Linkscape is updated, about once every 4-6 weeks. The schedule is here: http://apiwiki.seomoz.org/w/page/25141119/Linkscape-Schedule

Last update was Apr 5, so next will probably be mid-May. I believe SEOmoz is working on making the updates more frequent as many of us would love that...

Don't do it. Yes, they will be spammy, low-quality, and pass extremely little (probably zero) page rank. The directory pages are probably not indexed and if they are they are probably devalued. And if there are 65,000 of them then they're probably all owned by the same guy, in some templated database style, with the same C-class IP address. Google is wise to this and doesn't index or doesn't value those kinds of links.

I think that's because it uses the LinkScape database, which is only updated about once every 4 weeks. The next update is suppose to be tomorrow so you might see a change by tomorrow evening (i'm not sure what time of day the update goes live). There is a calendar of updates here: http://apiwiki.seomoz.org/w/page/25141119/Linkscape-Schedule

url length per se, no, but URI structure, yes. You should not have your site designed 10 layers deep, for example. There's a discussion of it here: http://www.webmasterworld.com/google/3579234.htm  and I like this example they give:

/readme.php?source=direct&page=34&articletype=widgetingforests&loc=uk

should rank better than:

/readme-source-direct-page-34-articletype-widgetingforests-loc-uk.html

or

/readme/source/direct/page-34/articletype/widgetingforests/loc-uk/

correct, a higher number is more competitive

Anything that is human reviewed, has editorial standards, and enforces longer word length. The opposite, which would be sites that accept anything, including posts that are less than 300 words, would not be good.

In the 'good' camp I would include

www.ezinearticles.com

www.buzzle.com (600 word min)

www.selfgrowth.com (500 word min)

would be interested in hearing from others on which ones they use that are human-edited or otherwise have more strict than post-anything-you-want submission guidelines.

Given all the recent talk about over optimization, when was the last time SEOMoz updated the on-page report card tool?

Rand wrote an excellent piece on Perfect On-Page Optimization (which is great, and thanks) in summer 2009. Is that still best practice 3 years later (and post-Penguin/Panda)? If not, has the SEOMoz on-page report card tool been updated to reflect current thinking for on-page best-practices?

I know the higher level concept is "write for humans, not for bots" but if you can do both (and not create an unreadable seo-frankenpage) then why not? Does getting an "A" grade wreak of over optimization now? Should I use the key phrase at the start of the title, h1, and strong (or bold) elements on a page? Should have an image with file name and alt text equal to (or containing at the start) the key phrase?

Thanks for asking. A few ideas:

1. Review of link-building services or networks. Rate each on cost, size of network, page rank of network, spammyness, etc. Any of them worth the trouble, or is the whole group to be avoided as grey/black hat spam?

2. How to get .edu or .gov links.

3. Find a small site that ranks higher than the wikipedia page for the generic keyword category the small site is about, and then offer an analysis of how/why the small site is able to rank higher than such an authoritarian site for that keyword/category.

4. List of top 10 article databases offering do-follow links and their requirements, if any (min word count, links per article, is the directory human reviewed, etc).

5. How to cost effectively have video testimonials made (by actors not working at your company), and then top 5 sites to post them on once you have them.

I know your pain. Been there. My experience was not good. Tried 2 of them. Even though they promised relevant, gramatically correct English postings, they did not deliver. I believe they were acting as middle-men to low-cost labor offshore who did not write solid English nor know anything about the niche my product was in. Plus, they did some questionable things with links that if done on large scale would probably get my site banned. Fortunately I was able to stop it before it got out of control.

The only way I would use an outsourced service again is to pre-approve every post/article in advance (at least for the first 50 or until I got very comfortable with their work). And that's tedious enough (and time consuming enough) that I'm not sure what value they're adding at that point. I could hire my own low-cost labor and then do the same approval process and then post them (cutting out the middle man).

It's really hard when site owners give testimonials like "my SEO firm is great" because the site owners are often unaware that their outsourced SEO firm is using black/grey hat tactics (eg JC Penny). If you're going to hire someone and pay them hundreds/month then take a long time to interview them. Ask for customer references. Ask for data on improvements they've made for their customers (what were the rankings before they started, and how did they improve over the months?). They should have multiple examples of monthly improvement for multiple clients, and then explain to you which white-hat techniques they used to achieve those results. If they refuse to share client references (where you can get on the phone and speak to some of their clients) or data then move along and look elsewhere; there are a million scamsters out there...

No, nav and footer links do not suck juice that you could otherwise redirect somewhere else via 'nofollow'.

Yes, juice is split evenly between all outgoing links on a page. So if you reduce the number of links (by deleting them, not just making them 'nofollow') then you can increase the juice passed to the remaining links on the page. This is why it's rarely a good idea to have a page with tons of links on it; each one is going to only pass a little bit of juice.

Having said that, nav and footer links can be very helpful for the humans who visit your site, so you don't want to strip them all away and make your site hard to use. But If you have dozens of barely-useful links jammed into your footer then you should probably nuke those.

Design your site for humans. And trust that Google will do something reasonable with the link juice flowing around your site.

Google doesn't know if it's a landing page or a regular page. It just sees a 'page'. Pages have on-page optimization (for a single keyword), and off-page optimization (other pages (internal and external) that point to them with different anchor text, relevancy, and authority). The page with the highest overall score is the one Google will rank highest. If you want your landing page to rank highest then do more SEO for it than for your non-landing-page. Of course, sometimes the best landing pages are really just regular pages that are relevant and well-optimized for the phrase in question...

Yes, a higher QS will lower the amount you have to bid to get the same position on the ad list. And, yes, on-page SEO related to the PPC ad will increase your QS. See this: http://adwords.google.com/support/aw/bin/answer.py?hl=en&answer=87411 and this http://adwords.google.com/support/aw/bin/answer.py?hl=en&answer=49174 and watch this short video from Google: http://www.youtube.com/watch?v=K7l0a2PVhPQ

Perhaps issue a press release announcing the interview (and/or product endorsement by the legend) and that links back to the interview? PRweb.com or PitchEngine.com are well tuned for SEO releases.

You can't go crazy and do just that, or do 10,000 instances of that with no other types of link building. Yes, having comment links looks natural. Up to a point. And it depends if the links are on human moderated boards or on spam farms. The latter will produce no value. If you can get a link to pass moderation from a credible site then it has value. Your comment will need to add to the discussion and often the moderators won't allow links in comments or they will no-follow them to discourage spammers.

Best practices for on-page SEO say that the alt text for an IMG tag should be equal to the keyphrase that the page is optimized for. So if you are optimizing a page for 'blue widgets' then having an IMG with alt text = 'blue widgets' would be good. You may also want to name the image file blue-widgets.jpg (or blue-widgets.png or .gif). It's one of several on-page signals you can send Google that this page is all about blue widgets.

Ah, those are personalized search stars. Read about them here: http://googleblog.blogspot.com/2010/03/stars-make-search-more-personal.html

From an SEO point of view it shouldn't make a difference. See http://www.searchenginejournal.com/seo-internal-interlinking-relative-vs-absolute-urls/7000/ and http://www.seo-theory.com/2008/05/28/are-absolute-links-better-than-relative-links-for-seo/ and http://www.webpronews.com/absolute-vs-relative-links-which-is-better-for-the-search-engines-2004-08

There are cases when absolute is better (RSS feeds, PDF documents, or even on your site so that if it is scraped the links might stay in tact), but personally I like relative URLs on my site. If I ever have to rename a sub-directory (not often, admittedly) things go smoother.

If you want the other pages to rank for the keywords they were optimized (on-page) for then you will need to do some external link building to each of the other pages (with appropriate anchor text for each).

When I've looked at the link profile of other sites and compared the number of external links that point to the home page vs internal pages, it looks like about 40-70% of external links point to the home page. Meaning, 30-60% of a site's links point to its sub-pages. Sometimes this is because the sub-pages have awesome content that people have chosen to link to, and other times it's because the webmaster has done standard link building exercises for the important sub-pages (guest blogging, article marketing, etc) that she/he wants to rank in the SERPs.

Two places capitalization comes into play:

(1) on-page use (title, h1, body text, img alt text, etc)

(2) external anchor text

I didn't think it mattered from Google's point of view for on-page usage (is this correct?) but I notice that OpenSiteExplorer' s 'anchor text distribution' tab shows different counts for the same keyword if it's capitalized in different ways (eg seomoz.org is listed separate from SEOmoz.org). Is that just OSE or does Google treat the keyword/phrase different based on its capitalization, too?

And if so, then should I be creating external links to my site with the 'regular' and 'Capitalized' versions of my key phrases?

I can't imagine that the traffic numbers include intermediate Instant searches before the user has hit 'return'. Otherwise all the shortest key phrase would always have the highest number of searches (the sum of every other search where it is the prefix).

I don't know how reliable the traffic estimate is, but it seems odd that it's showing zero for a phrase that gets 1M searches. I just tried my own brand and I do show non-zero results that seem low but are certainly not zero.

You typically optimize 1 page for 1 phrase. If there are two phrases very close together, like 'widget' and 'widgets' then you can optimize the same page for those 2 phrases. You will have sub-optimal results if you try to rank the same page for 3 or more (or even 2) phrases.

If you have 40 phrases you want to rank for in the SERPs then you'll need 40 different pages on your site. Build a spreadsheet where you map 1 url to 1 phrase to help you keep it straight.

I have found it hard to evaluate links on a weekly basis, too. All tools give different results. You need to pick 1 tool, create a spreadsheet, and then track links with that same tool week to week. At least you should know on a relative basis if you're improving or not, even if you're not getting the exact true link count.

The Link Analysis tab of SEOmoz shows some stats (external followed links, followed links root domains, unique C-blocks) that are useful to track over time, but they only update about once a month (note to SEOmoz: weekly would be great)

The one thing you CAN track week to week is your SERP rankings, on the Rankings tab. You can track number of phrases you have in the top 3, as well as number of phrases on page 1. If your external link building is progressing then those numbers should go up week over week.

And those numbers drive your organic search visits per week, which you can view on the Traffic tab, or in Google Analytics directly.

As far as follow/no-follow goes, remember that you want a 'normal' anchor text distribution and usually a normal anchor text profile will include some no-follows. So keep some no-follows in there...

Yikes. 750 links (even if split into sections so you don't see them all at once) on 1 page is not 'human-friendly' (as I know you know). Is that really necessary? I looked at the site and having navigation sub-menus that go 4-deep is kind of a usability issue for people that aren't great with a mouse. Maybe look at ebay's menu structure; they certainly have a large db of products in many categories and don't resort to 100s of links per page just for navigation.

But from and SEO point of view I'm not sure the 750 links are hurting your on-page ranking for a phrase that the page is otherwise optimized for. If you had a page optimized for "widgets" and you did all the correct on-page things for 'widgets' and then had external links pointing to the page with anchor text of 'widgets' then I'm not sure how much you'd be penalized for having 750 links on that page. Given that on-page factors are only about 20% of the ranking equation anyway, I'm not sure it's a huge deal from an SEO/ranking point of view. It's more of a human usability thing, IMO.

Are users likely to use his name when searching for his site, or are they more likely to search for the kind of service he offers? if the later then forget his name in titles (except maybe the home page). IMO the title should match the likely search query for optimum SEO. And repeating the brand in every title is a waste, IMO. If you have 10 pages optimized for 10 different search queries then the title of each should be (or at least begin with) the likely search query that each page is optimized for.

It's a single site. It's not a weather app; I was using weather as an example. I have another (non-weather) data set that has similar characteristics -- 6 or 8 unique data points per item (a couple of which are dynamic and change as time passes). And several thousand of these items. I want to display them via template.

I guess I'll try it and see if Webmaster tools complains about dup content. Thanks for the ideas, guys...

Only if the 2 keywords are very close together, like 'widget' and 'widgets'. Otherwise, 1 keyword per page works best, or else you'll confuse the googlebot as to what your page is really about and what it should rank for.

Well, SEOmoz.org is a PR7 site. How many links do you have from PR7 sites? Even if on a sub-sub-sub page they are valuable, and hard to get. Plus, it would count as another c-block IP address that points to you, which has some value independent of relevance of the referring site.

I understand your point that relevance of referring links is critical, but there's also value from non-relevant, high-PR, new C-class IP addresses pointing to you. It's worth getting.

I've heard people ask that question many times. And the answer has always been (from many people) "no. it's a non-issue". So I don't think you have to worry about it. Here is one thread on the subject (there are others if you Google for it): http://www.webmasterworld.com/forum3/8461.htm

I would like that, too. Especially for things like new C-class IP addresses that link to my site for the 1st time, or new root-level domains that link to me for the 1st time. Would be great if OSE's data included "date link was discovered" for each link. Then we could begin to tie together new links with changes in rankings.

Thanks, David. I thought I had read that meta keywords were still used by Yahoo or Bing (definitely not Google). Is that not true? Does it harm anything to have the meta keywords if I just leave them as-is (since I already have them), or should I pro-actively go remove them?

Article submission isn't bad, but those are generally low value links. How about guest blogging on other (hopefully non-zero PR) related on-topic blogs? Offer an article to a relevant blog in exchange for a linkback from the blog article. It's kind of like article marketing except that you pick the destination based on PR and relevancy, rather than just submit to a big generic article database.

You could also create landing pages for your site with the articles, and then promote the landing pages with SEO. Then have those landing pages feed into your main site.

5C is the hex code for '' (backward slash). Maybe you have a backward slash where you mean to have a forward slash '/'? Ascii codes here: http://www.asciitable.com/

It used to be possible to use nofollow to redirect more juice to where you wanted it. But Google changed the algo so now that's not the case. See Rand's excellent pieces on it here: http://www.seomoz.org/blog/google-maybe-changes-how-the-pagerank-algorithm-handles-nofollow and http://www.seomoz.org/blog/google-says-yes-you-can-still-sculpt-pagerank-no-you-cant-do-it-with-nofollow

Paul, I assume you mean making the image file name the same as the alt text? And both for the phrase that the page is optimized for, right?

SEOmoz maintains a list of directories here: http://www.seomoz.org/directories

Depends how competitive the keyword is. If highly competitive then, yes, put phrase at start of title tag. If you're already ranking well and the phrase is not too competitive then you have a bit more leeway. See http://www.seomoz.org/learn-seo/title-tag#seo-best-practice

Can the seller provide you with historical Google Analytics data? If there is a sudden drop in traffic then the domain may have been penalized. Is there a visible site on the domain today? Does it rank for any phrases (should it? given the backlinks it has)? Maybe enter it in http://www.ispionage.com and then click on the SEO tab to see if it ranks for any phrases today. If it's on page 1 for anything then it is unlikely to be penalized.

Also, does it link out to any bad neighborhoods? Google cares more about what you link to than what links to you (in terms of giving you a penalty). If bad neighborhoods link to you then you won't get penalized but if you link to bad neighborhoods and Google identifies it as such then you will get penalized.

It's a question of whether they still have the external links pointing to them, and how you handle those references. If you correctly 301 redirect all those old inbound links (which may point to subpages that no longer exist) then you will still get the link juice. If, however, all those inbound links result in 404 errors then you will not get the link juice from them.

If you buy an old domain where all the existing exeternal links point to the root then you should be good on the link juice. But if you have a situation where you're buying an old domain that had deep links (from the outside) that you're not going to maintain then you could lose a lot of juice.

Also, keep in mind that PR hasn't been updated in a long time. Impossible to know what it's current value us (mozRank is more up to date than Google toolbar PR).

In my expience, the reports you see are the sum of two things (1) it automatically includes a report in cases where your site is in the top 50 results of the SERPs for each campaign keyword, and (2) you can manually cause it to add a report by going to the Report Card page (second sub-menu choice on the On Page page) and giving it the keyword and then the URL you want it to report on for that keyword. I believe it will remember these manual pairing additions between sessions (mind seems to). This manual process is very helpful when you are working through your on-page optimization so you can see how you're doing (from an on-page point of view) before you achieve top-50 status in the SERPs.

Query-dependent search results that redirect vs giving standard search results is kind of a confusing behavior for the user. If I'm the user then sometimes I get a typical search results page and sometimes I get a whole other look and feel for a category of results.

Instead, I would always give the search results in standard form (so all searches have the same behavior/results look/feel) and then, for certain queries that match categories you have add something above the search results that says "Looking for [query]? Here's our <query category="" page="">" or something where <> is a link to your dedicated category page. I think that's cleaner design for users and no redirects necessary.</query>

1. Free traffic. Company on a shoestring that can't afford to buy traffic with PPC (my particular keywords are $5 or more per click). Have lots of time, though. So can learn and implement correct and good SEO.

2. Almost immediate gratification of seeing higher rankings in SERPs followed by higher number of visitors/day. Direct correlation between my efforts and results.

3. Satisfaction from beating the other players (people competing to rank for the same phrases I'm targeting). It's almost game like, and Google keeps score with rankings.

Ok. Thanks.

Kind of related feature request: For off-page factors, like backlink profile, it would be nice if OSE grouped the anchor text of backlinks into categories to give you a rough % of these:

• exact match
• variations of exact match
• brand
• variations of brand
• url + variations of url
• random (click here, visit, etc)

So we could see at a glance (for a given keyword) if a page had an unnatural looking (over optimized) backlink profile.

You could also integrate this with your Report Card feature... in addition to giving a grade for on-page factors, it would be cool if it could give you a grade for off-page factors. If that's too hard then just put the functionality into OSE.... Thanks.

My understanding is that the 2nd and beyond links to the same url from the same page don't count (ie they are worth zero) and that their link juice does not get redistributed to the other links on the page. So it acts like the 2nd and beyond links to the same url are nofollow.

Then I suppose what the query was that brought them to that landing page. If someone searches for "replace my windows" or "I need new windows" and ends up on a page offering a free consultation (and assuming the company looks credible, etc) then I can imagine a pretty high conversion rate. 15% seems believable. Maybe even higher if it's just email address in exchange for product info for products related to the initial query. But again, no one is going to share this kind of data because if it's really good it will attract new competitors and no one wants to invite competition...

Since it's a weighting of many factors, everything else (the non-keyword-domain factors) should go up in relative proportion. I don't think they will explicitly turn up some other factor in order to counter the turning down of one factor.