Welcome to the Q&A Forum

Let's say Moz links to your site 5 times. That's 5 links but 1 "root domain" (moz.com)

Now, NY Times links to you ... 1000 times. That's 1005 links and 2 "root domains" - moz.com and nytimes.com

A root domain is just a domain that links to you - the first link from each domain seems to count more so the more root domains you have is better (rather than the number of backlinks total.)

To put it another way, you'd rather have 20 backlinks from 20 separate root domains than you would to have a sitewide link from 1 domain even though that may be thousands of backlinks.

Rand covers this all just a little here: https://moz.com/blog/understanding-root-domains-subdomains-vs-subfolders-microsites

Given that the word "update" is such a specific word, they should probably use another term such as "data adjustment" or "data change" or ... something. Pretty much ANY word other than "update" would be a good option.

Is it related to a time-of-day query? I see what I can only describe as QDF fluctuations sometimes.

We have run the same keyword rankings every hour for 48 hours before. All 48 were unique sets of results ... the top "100" ended up including almost 300 different URLs. A lot of the fluctuation was on page 1. We saw that a search that "deserves freshness" ended up fluctuating MUCH more. Say "restaurants near Melbourne."  In the morning we often saw different results to dinner-time.Over a couple days it's hard to say but especially on mobile, those things are changing fast.

(It's the old context question. If you type "who is the President" you get info about Obama. If you type "who is his wife" you get random info. But if you say those two questions in Google mobile, it returns Michelle Obama info.)

I do think your exact issue may be some testing, though. We have an idea that the newest/last release Penguin is coming soon. Testing makes just as much sense in your case. (The results often shift, go back, shift, go back.) I would start checking links & cleaning up in any case.

I always find these discussions interesting mainly because of the "nobody knows" factor.

What I want to point out more than any one particular thing is that there are a limited number of 'ranking factors.' Whether that number is 200, 50, 3, or 3000, it's limited. Whenever one tactic loses its effectiveness simple math says that other factors increased in importance.

Google is very, very good at messing with SEOs.

If keyword density is limited or removed, something took its place. Something that isn't "create quality content."  Most serious competitors are creating high quality, relevant content. If you think the difference in ranking one bank over another is "quality content" on a target page for "home loan" you're lying to yourself. Backlinks, age of the page, age of the site, internal links, anchor text, and some level of "keyword density" (though I think it's much more sophisticated than that" definitely helps. H1s still matter, as do H2s and H3s, tbh.

I have competitive keywords ranking with NO on-page content AND no backlinks. The page literally has a title, H1 tag and the surrounding menus & sidebars. It ranks for gambling-related keyphrases in a supposedly hard to rank niche and has ranked for months (with zero on-page text.)

SEMRush shows that same site rank for over 1200 keyphrases. It has ONE backlink to the homepage. That's it.

I wish that was the only example. But I am ranking semi-competitive marketing & SEO related keyphrases on a site with about 8 links and virtually no content. If content + links = SEO, these would never rank. So again, it's beyond that. Age of the domain? No, one is brand new. One is older. One is registered for more than a year, one for less. One is an EMD, one is not.

We've had new clients struggle & struggle to rank for really easy keyphrases with no backlink spam, technical on-site looks good and titles/content/links are all in line with other (ranking) clients. We put all their content on a new domain & it ranks just fine. NO links.

SEO is just weird. Let's face it - we're all attempting to do the best we can for clients but at the end of the day, none of it truly makes that much sense.

I would definitely argue in favor of this. Cleaning up broken links, changing the copyright date on websites, adding new content - it all sends signals to Google that the site is maintained regularly and has active management. A site that is regularly updated is more valuable than one that is created and then left to rot.

While Matt Cutts said in 2013 (eons ago in SEO) that broken links weren't a ranking factor, the Google Search Quality Raters Handbook says they are a factor for manual review.

They actually say:

Webmasters need to make sure their websites function well for users as web browsers change. How can you tell that a website is being maintained and cared for? Poke around: Links should work, images should load, content should be added and updated over time, etc. Exercise caution relying on dates: Some webpages automatically display the current date. Rather than just looking for a recent date, search for evidence that effort is being made to keep the website up to date and running smoothly.

When the Raters Handbook says that, I fix broken links.

There really isn't a typical amount of time as it depends on links to your site, crawl frequency of your current page, and how many resources they choose to use on you at that time.

There are, however, ways to speed up the process.

1. Claim your site in Search Console (Webmaster Tools)

2. Submit a sitemap

3. Use the "Fetch" tool in Search Console.

4. Build/earn some new links to the pages you care about.

All of those will help your pages be reindexed more quickly. Also, if you want the "very fast, slightly cheap" way of doing it, post the links on Google+, ping them with Indexkings and build internal links from new blog posts. All of those definitely help with index status.

Like Deacyde I use the content in quotes in Google itself "to test if this is duplicate." But I also use a couple tools:

• http://www.copyscape.com/
• http://www.siteliner.com/
• http://www.webconfs.com/similar-page-checker.php
• http://www.plagspotter.com/ (undergoing a bit of renovation currently)

If you wanted to build this into your own tool, PlagTracker has an API

http://www.plagtracker.com/api.html

(As do a few of the ones above.)

Moz has a recommended provider list though I believe it doesn't change a lot over the years.

My advice on this is typically the same every time - browse through Q&A, find people others agree with a lot and who seem to know a lot more than you about SEO, look into their profiles and hire from there.

There aren't a lot of great, reliable ways to test how much someone knows or what they can do for your business so a lot of this is based on trust and results. Just be careful in your selection and be reasonable in expectations for the budgets provided.

Could you be looking at the wrong (www vs non) version of the site?

Did the dev block robots.txt from all bots (or Mozbot) before transfer?

Are you still cached & indexed in Google?

There's also an ongoing issue with some sites as seen here.

Google will do this somewhat in the search but I wouldn't do it on your site. You won't be telling Google what the page is about, you'll be asking Google what the page should be about. That's a very weak position to come at things from.

With the very little amount that Google actually passes you the search query, it's probably not very useful to attempt to implement this on pages.

On your Adwords, for sure. But not on your actual website, imho.

I'm going to take a very neutral approach to this until I see how it plays out.

Not requiring a G+ account may make things a bit easier for detractors to hate on a business. So it's a bit of one, bit of the other. I'm monitoring all my important reviews closely at the moment just to make sure nothing slips through that I may have in the past missed or let go.

(Taketh the good, taketh the bad. Taketh them both ... )

1. It's not bad to have pages that are not directly linked to your site. In fact a good lot of PPC landing pages aren't typically linked from anywhere within the navigation or website itself.

• Keep the page in the sitemap if you want it to remain indexed but if you only care about the direct traffic, it doesn't need to stay.

• You can definitely 301 the old page to the new one but you said you previously hid the parent? I would only 301 redirect if you're 100% sure this is the way it will stay forevermore.

So it really depends on why you're removing and what you essentially want to happen.

I would actually add a canonical tag and then handle these using the Parameters section of Search Console. That's why it's there, for exactly this type of site with exactly this issue.

https://support.google.com/webmasters/answer/6080550?hl=en

Well, Moz Content isn't something I subscribe to yet so I'm just sort of giving you an outsider's perspective but I like to do the searches and look for content that's winning. Here are some things I use in my analysis using Moz Content:

1. All the related keywords on the top content. They are often different from each other and it shows me how the keyword ties into its related terms. For instance, if you do a search for "email marketing" you come up with a Clickz article that has: b2b email, email marketing, b2b email marketing, b2b, b2b email marketing tricks. But the 2nd result is Marketing Profs and theirs has email subject lines, subject lines, worst performing words, etc. So you can find a variety of topics & keywords this way.

2. Blog research. Want to know what someone is searching for in your niche? Try "problem with SEO" - you will find some of the top articles that answer these questions and can put your own spin on the answers, as well as find related questions and think of ways to use the keywords in question. For example, under "Worst performing words" in the email example above you could do a post of "20 Best & Worst Performing Email Marketing Words" - new content, new-ish topic, perfect blog material. You can also then put "worst performing email" into Moz Content and find the top performing content for that exact search.

3. Use Moz Content + Open Site Explorer. Take that top performing Clickz link and put it on OSE. Now you have all the links back to that content. These are places that may be interested in your new article and help you build backlinks, as well.

4. Age of great content - the older it is, the more something is likely to have changed. You can update these articles for your own audience, look for broken links (broken linkbuilding), and when you do update, go back to all the links they had and show those site owners your new & improved content.

Lots of ways to use Moz content - these are just a few ideas right off the top of my head and none of these even require the paid version. Once you get that you open up auditing your own URLs, getting CSV exports of results, having a content inventory to find gaps in your material and gaps your competitors left open, topical analysis and more.

I love the way you phrase things. lol

I agree with EGOL here - it's definitely not helpful, it's definitely not needed, it could be harmful, and it's a sign of other potential issues. I would definitely work to remove it from the title.

Plus,
.... I mean what's it trying to accomplish? This tells me it's auto-generated or (stealing EGOL's word) "dumb." because it's a line break. You can't line break a title. I mean I could see some example of code that are meant to do something in a title (for instance, **for bold) but not BR? It makes no sense. **

Exactly right. When you say launching a new site, it's vague so John's answer is 100% correct.

• New domain - new UA
• Old domain with new design - old UA

So it just depends which part is new.

You probably don't want to be going 200 > 301 > 200 too often. Although it may work and may rank again, once you 301 by definition that's a "permanent" redirect. Especially since the property could then go off again ... 200 > 301 > 200 > 301 > 200 ... and somewhere in there, all is lost.

I would make these 302 links. First, if it's 302 long enough, the goodness will pass through the 302 anyways. I don't know if this is 90 days, 180 days or something really unexpected like 102 days ... but eventually it does pass through (same if you 302 a bad domain to your main domain, the penalties will eventually pass through.)

But a 302 would serve you as a temporary redirect, keep the juice on the referring page (unless it never comes back) and then either long term rank the original URL or pass the juice.

I am assuming they have hardcoded it themselves into their site using the same idea as the plugin does. When you said you wanted to copy it, that plugin I linked is probably the best way unless you get a developer to hard-code it for you like SEJ appears to have.

This question was recently asked over on Inbound and someone linked out to this: https://wordpress.org/plugins/read-metrics-for-google-analytics/

It seems to be brand new so this may be the best at the moment.

It really depends on your current metrics. As Sergey said things like views, videos uploaded, shares & minutes watching are all ranking factors over there so the more you have already the better.

If your average time spent watching is really low, I'd consider moving them - but an account with history is more valuable than a brand new one so if your numbers aren't completely shot, I'd keep the channel and rebrand it.

By far the most SEO-friendly platform overall is going to be Wordpress.

While other platforms can be made to do what Wordpress does, almost a quarter of websites online are powered by Wordpress which gives it a huge community to draw resources and information from.

I don't mind Joomla (i'm the one guy, I know) and I have friends who prefer Drupal or other things for security reasons (the biggest always gets attacked most, right?) but Wordpress has Yoast SEO plugin and that is a huge help.

I have a few ecomm stores online and I use Woocommerce with Wordpress for most. I have a Magento store but ... shew. It's so much harder, it's so much more resource intense and it's not as good for anything unless you're a developer.

Good find on the homepage - I saw it was Wordpress and assumed that would take care of the problem but you're right, something is wrong.

Also, the canonical tag is set to non-www  and needs to be fixed.

These things resolve themselves over time. Google (and Moz) have a record of what your site looked like. If they have 100 "pictures" of your pages and they see the www one, well that's new. We'll put that .. oh wait, it's duplicate.

They haven't gone back to see that the one it's a duplicate of is no longer there and is now redirecting. Once the bot crawls the new site entirely, those type of issues go away.

Your rankings should bounce back as soon as Google gets your domain crawled again. In the meantime if you don't have Search Console setup, do that. Go into the section that says Fetch as Google and get them to recrawl your site. Also, if you have a sitemap.xml on your site you can submit a sitemap through Search Console to help them crawl more of your site.

Those two things should get you back on track a bit quicker.

You said you don't like the idea of redirecting the root domain to a sub dir - but why? It's passing the full value of the UK site to the new location of the UK site. Makes perfect sense in my head. I would definitely do so.

I would only recommend the /US version because of the hreflangs. You can do it the other way, it's just a bit more technical to build the hreflang tags. You can use this example I've used before, though:

" hreflang="x-default" />
" hreflang="en-uk" />
" hreflang="en-de" />

This will do the trick, I believe. Then will display that in the header of every page and the request URL will grab the URL you're on so you'll end up with hreflangs for the two URL structures.

You said you could migrate .co.uk to .com/UK asap - do it as soon as you can, as long as the URL structure won't change. If you're moving from say .co.uk/about.html to .com/UK/about-us/  then you can't do it until those URLs exist. I wouldn't move the existing site over now if the URLs are going to change. This is what I was saying above - you don't want to go from .co.uk/about.html to .com/UK/about.html (and then in August) to .com/UK/about-us/

Make sense?

Awesome find!  I did a little checking and this would be my guess as well.

If it were my site, I would move the UK site to .com/UK, then have .com/DE and .com/US.

Then I would 301 redirect all the UK links back to their counterparts on the new site, keeping as much authority as possible. Point the main homepage at .com/UK and all the inner links as they currently are.

Doing this now will definitely help your expansion in the future as well.

I would do all this after you've rebuilt the site, though, because you don't want to do 301s to 301s. Figure out what the new site will be, do the 301s to that one directly. No intermediate jumps ... it's a longer time to implement but you're better off in the long run.

Links you should be getting (related industries, etc.) shouldn't hurt you and can definitely help depending on the link source.

US links are generally .com, .net and .org.  I can't see Google devaluing those even if they point at a .co.uk  Imagine a NYTimes article pointing a link at BBC ... that would be the highest quality connection, right? But they're related industries and across the pond.

Would I get one from a .us site? I mean ... maybe? If it's a really decent link, sure. But I wouldn't want 50 .us domains pointing at me. That's like .ru, .ro, .ar and .it - the more of those other country extensions you have may indeed show signs of spam (there are great .ro links and great .ar links ... but if you see them, there are usually 50 of them and they were bought links.)

Some of your pages lack enough content or unique content for Google to care:

This: http://carpetcleanertn.com/carpet-cleaning-specials/ isn't indexed. But then, why would it be?

http://carpetcleanertn.com/carpet-cleaning-spring-hill-tn/ this isn't indexed - but ... yeah.

No content except an image on this either: http://carpetcleanertn.com/gift-certificates/

None at all here: http://carpetcleanertn.com/great-nashville-area/

Your pricing & services page aren't being indexed but that's because they're near-copies of the actual chemdry page. See attachment.

I think if you wrote clean, new copy for the pricing & services page you'd see them get indexed but since they're super-close to the original, Google has chosen to only index one.

vLWyTlO.jpg

Are you getting the message in Search Console that there were errors crawling your page?

This typically means that your host was temporarily down when Google landed on your page. These types of things happen all the time and are no big deal.

Your homepage cache shows a crawl date of today so I'm assuming things are working properly ... if you really want to find out, try doing a "Fetch" of your site in Search Console.

Crawl > Fetch as Google > Fetch (big red button)

You should get a status of "Complete." If you get anything else there should be an error message with it. If so, paste that here.

I have checked the site headers, cache, crawlability with Screaming Frog, and everything is fine. This seems like one of those temporary messages but if the problem persists definitely let us know!

Do you know that the read more tag on that page actually displays less and read less actually displays more?

When I put that page through Pingdom Melbourne I get a 5.09 second load time and 1.6 MB page size. That's a bit long and you get a 45/100 (red) score on Google Developer's Tool for mobile and a 70/100 on desktop: https://developers.google.com/speed/pagespeed/insights/?url=https%3A%2F%2Fwww.yellowoctopus.com.au%2Funique-gifts%2Fgifts-for-men

The site at #1 gets a 79/100 mobile and 92/100 desktop and is only 209 kb and loads in 2.37 seconds on Pingdom: http://tools.pingdom.com/fpt/#!/bol6kH/http://www.giftsforblokes.com.au/

You have what looks to me like some link spam going on. There are 30 domains that link to you on just one IP address and another 20 from a second IP. That's 50 domains across just two IP addresses. The odds of that being natural are astronomical and very, very unlikely. https://ahrefs.com/site-explorer/referring/v2/ips/subdomains/fresh/all/1/refdomains_desc?target=yellowoctopus.com.au

You also have a staggering number of links from .org, .net, .uk, .ru, .de, .info and so on. This is also a high indicator of link manipulation. The site at #1 doesn't have this issue at all.

The vast, vast majority of your backlinks come from images and the anchor text is your brand, not anything to do with men. 45% of your links over time have been deleted, also a very bad quality sign: https://majestic.com/reports/site-explorer?folder=&q=yellowoctopus.com.au

The site I see at #1 is roundly kicking your site's tail at every corner. The site I see at #2 loads a slightly larger page than you but does it in much less time: http://tools.pingdom.com/fpt/#!/dW9TDy/hardtofind.com.au  Most of the sites above you also have a better (clean) backlink profile. Hardtofind has over 23,000 links showing up in Majestic, which is huge. Only about 15% of their links are images where yours is closer to 80%. Only 20% of their links are nofollow while 50% of yours are.

Your page is the only one in the top 6 that uses meta keywords and you use them a little bit "old school" - they can generally only be a negative ranking factor and it seems they aren't helping you.

The content across all 4 pages is quite the same so it shouldn't matter that much.

I would say your slightly spammier link profile, slightly slower load times and slightly fewer dofollow, good links are pushing you down. Their stuff is, frankly, better at the moment. And I see DJs and Kmart above you - bad luck about that but if they want your category you're going to have to spin miracles to beat them.

I would check Search Console for a penalty notification although this is most likely an algorithmic adjustment. Those links aren't counting for you. If someone did a disavow on them they may have also gone "too far" with it. It's hard to say. I would check search console settings, including claiming all versions of the site and make sure the HTTPS and www version are your preferred versions.

Then I would try to draw non-image links to the page, maybe add some new content above or below products, work on your link profile overall and push the domain as a whole as well.

PM the URL or post it here so we can take a look. Generally this is a case of:

1. No indexed links yet

2. Duplicate content

3. Robots meta tag or robots.txt

4. Content not actually in the site (iframes, other coding issues)

5. Content in an image/flash/other unreadable or nearly unreadable format.

It's a very tough question and one that is common with a lot of e-commerce.

The only really complete solution I have for you that addresses each of your needs is to not base the page "content" on the specs.

Make specs a table on the page but put in enough unique content about each model and variation that it has its own truly unique content.

I know this solution means writing at least say 200-300 words of unique content for every model but 100k words solves the whole issue. It just depends if it is worth them all ranking. But this solution gives you:

a) unique content

b) chance for every page to rank & no canonicals back to one page

c) much more long tail search volume

d) specific searches for every one of your potential customers.

That's really the best I can do ... it takes the duplicate content issue away and solves every problem except the one of having to create this much content in the first place.

Generally a huge difference in local vs organic rankings is one of citation authority. You need to build more citation strength with full NAPs (name, address, phone) - things like what Moz Local does.

Doing that type of citation-building should help move you up the local charts. Make sure your GMB page is optimized for the location, you've added photos to the listing, the keyword is prominent across the description and such ... consistent NAP on other sites that list the business, and then just build more. Should rank fine after that.

I would have a main page about each service topic and then link off from that.

Bathroom remodeling page

Bathroom remodeling content

We do bathroom remodeling in: blah blah blah, linked page, linked page, linked page.

Or naturally in the content if possible, of course. Or even a list at the bottom:

Our service area includes: town(link), town(link) etc.

This works very well for clients I've had in the past with many suburbs & locations.

If you do more than one type of thing per location, I would silo them:

• example.com/ashburn/basement-remodeling
• example.com/ashburn/kitchen-remodeling
• example.com/sterling/basement-remodeling
• example.com/sterling/kitchen-remodeling

You likely won't outrank them just with the MLS contents. But can you put unique content around each result?

If you can make your page look more like:

• Header
• Unique Content
• MLS
• More content
• Footer

then you can definitely compete. Using this formula and strategy I was outranking sites like Amazon, Ebay & Polyvore for clothing-related searches. It's fine to have some duplicate content. It's not fine to have ONLY duplicate content if you want to rank against some top sites. The good news is they aren't doing much in the way of unique content so if you do, you have a good shot.

Agree with Jordan - block the parameter for search in robots.txt and forget it. It won't bring search traffic in, it shouldn't get crawled but if it does, it's always a negative.

I would say the most important is that you only have 1 link and I always say it like this: you need to be listed where you need to be listed. Local sites, wedding sites, where are your competitors? search their URL in Google or using a tool like Moz and find their links. See what you can easily mimic and expand from there.

Content & blogging will come and need to be done - but the "in between times" are great for link building. (Or, hire a wedding/photography link builder and do the blogging/content yourself.)

Anchor text on backlinks will help Google sort it out, as will the content itself. Generally you won't want to rank two pages for any keyphrase but say in the SEO example - because you're pushing those pages so hard sometimes one falls. It's not that they replace each other in a bad way, it's because one is "sort of" being pushed too hard that it may fall and get replaced by your backup.

In a plumbing example, you'd target the homepage to say "Melbourne plumber" because of Google My Business but over time you may add services and swap it to another page.

If your site is only in English you don't need to use hreflang tags at all.

Well, you have two .com.com in there ... so that's in general wrong.

But yes, it's en-gb, not en-UK.

You just have to have all the tags on all the pages. You can use Flang to see what's wrong as well.

Hi Jon,

Former wedding photographer here myself so I know exactly what you're going through. It can be hard to watch someone jump ahead of you or start to make huge progress when you aren't.

That said, let me address a few differences on the sites to help you a bit more.

He has 185 pages of content indexed, you have 19. That's 90% more content. Most of this is long-tail keyworded traffic going to things like Waimea wedding photography blog posts, Poipu beach blog posts, etc. That sort of content really helps and fleshes out your site more. It definitely establishes him right off the bat as more experienced / expert in the area.

His homepage title is much better & cleaner. Yours has two pipes - one of which leads to nothing. | <- that thing.

Many of his titles include the word "photography" where most of yours include "photo & film."  Spamming the same keyword over & over isn't necessary but since he's varying the locations & all that, it's working well for him. In several cases your over-use of the photographer/y keywords actually get you into some trouble.

According to Majestic, I see exactly one live link for you: https://majestic.com/reports/site-explorer?folder=&q=balihaiphoto.com  That's not much authority to rank you for anything.

Your domain also has some old Chinese spam/links attached to it. http://2.1m.yt/KSwroiP.jpg mostly from it's history: https://web.archive.org/web/20020926015757/http://www.balihaiphoto.com/

You have some duplicate content going on:

• http://joshuagrantrose.com/nicole-ryan/
• http://joshuagrantrose.com/gallery/nicole-ryan/

Which will be problematic when you have this little authority & this few links.

Anyhoo - that's quite a lot of info to start with. If you want, add me on Facebook where I share quite a few industry-specific tips including this marketing guide for photographers which is a bit outdated now but the concepts should still apply. Happy to help if you need more.

Generally when you get the no return tags error it means that you put a hreflang for one page but not the referring page.

Let's say your pages are:

• website.com/en
• website.com/es
• website.com/fr

If you put hreflang tags on /en but not the rest, you'll get the no return tags error.

So just put the full hreflang tag set on every page that uses any one of the languages and you'll solve this error. Also see Webmasters Support for more info.

Well ... are you checking based on the same keyword? That's the first question. They both use 27 data points and I do think there's quite a lot of overlap, if not 100% ... so they should be fairly similar in terms of what's missing/what you should fix.

I just don't think they're graded the same way with the weights. As an example:

If I check my site http://mattantonino.com/ in the Page Optimization tab for my name, I see a score of 79 because I have addressed 22/27 factors.

Using the on-page grader I have also addressed 22/27 factors but ... I get an "F" for optimizing this keyphrase.

But the grader rates the importance of each factor - Critical! My title doesn't include my name. And two of the factors are "high importance."  The grader says that as well - "We weight the items graded by their importance level, which we determine by our industry experience and correlation research."

I think the Page Optimization is not weighted the same way - it's more of "how many boxes did you check."

As long as your SEO is done fairly well, the on-page grader AND the Page Optimization should only be used as a guide. I used the above example to illustrate the point as I'm currently ranking #2 behind my well-optimized LinkedIn profile for my name. Yes, if I put my name in the title I would shoot to #1. But #2 and the page got an F on the grader.

I chose the #1 ranking site for "SEO" here locally and it receives a "B" grade with the keyword used 193 times. The sites ranked #2 & 3 get an A and the one ranked 90th gets an A as well.

A link can't be direct to your site, then refer back to shareasale, then redirect back to your site unless you've set it up that way somehow.

So if someone builds a shareasale link like:

Link to your product

This is going to shareasale. If you quit shareasale and they said it would still go directly to you, they're just not planting their tracking cookie on the intermediate step. Everything else would remain the same.

If the link is something like:

Link to product and somehow that goes from your site, to shareasale, back to your site - well that's definitely not helping you very much for SEO and if you quit, yes that would then go directly to your site - but only if you reconfigured how the links were working since this type of linking would require a lot of special work to make it happen. It's more likely it's the 1st example.

In any case, they both pass through Shareasale and, per Matt Cutts, Google is attempting to not give you credit for that link. Whether or not they do, I can't say - but they're attempting not to. I'm not sure which part of my previous answer may not be "completely true."

I've seen many pages optimize both the homepage and an inner page. Think about how most SEO agency pages are organized. Homepage has some SEO content - and most pages that rank for "SEO" are homepages. But then you have the same sites also inevitably have an /seo/ page on them or /services/seo as well.

So doing both is sort of standard operating procedure for some businesses but it really depends.

• Best for ranking would be as Eric said. Optimize as if you're writing a book. Homepage should be optimized for the title / main topic.
• Best for conversions may be to have a dedicated landing page for each topic and use your homepage to tie them together.

These type of links are generally demoted/do not pass juice in the first place.

Matt Cutts said in the past in an interview with Eric Enge:

Matt Cutts: Typically, we want to handle those sorts of links appropriately. A lot of the time, that means that the link is essentially driving people for money, so we usually would not count those as an endorsement.

To me, Shareasale is a huge provider of affiliate links so I would assume Google is well onto those links and doesn't count them. You aren't going to get a ranking benefit from these links, IMO.