Welcome to the Q&A Forum

If I look at your backlink profile with ahrefs, there's few I don't like, I am referring to anchortexts stuffed with keywords which doesn't look natural to me, and I assume google would not consider them natural either.

Your competitor pinkcherry doesn't look much better, but that should not give you much comfort. And btw they have much more linking root domains so they beat you anyhow.

It's the on page which is really an issue in my opinion. Just starting with following advices from MOZ page grader would help a lot.

Take this page http://www.cupidboutique.com/vibrators which I assume from your early comment you would like to see ranked better than your competitors for the keyword "vibrators", and look at the source code. Starting from the head there's a lot I don't like, if you want to rank for vibrators, trim the title, that megamenu is overkilling with external links, the keyword density is overwhelming, and 50 h2 tag are simply too much. And content is thin.

You should rethink it. If you look at hushcanada.com page source for their best ranking results for the keyword vibrators you will see it's much much much better.

Pinkcherry rules because of the links, hushcanada beat you because of their onpage.

Change the name of the login page, I mean in addition to having a strong password of course.

Those automated scripts look first for the known wp login page if they don't find it the will give up, if they do they will keep trying forever and ever, an unecessary load for your servers.

Your opinion is reassuring. Thanks.

Hi Rand,

it's helpful.

But I can narrow my doubts/question to this: what is a safe range for the ratio of deeplinks vs rootlinks to get the best results?

If a website is getting a lot of deeplinks and few links to the root domain, google could perceive those pages as highly relevant but do not reward the domain as a whole?

I know what you are thinking, since google algo is not on github, we don't know exactly.

But if you at MOZ calculate DA as you explained, is probably because you think google is doing something similar, and thus rewarding a domain as a whole for it's deeplinks?

Hence, in terms of the broader SEO concept of Domain Authority, should we assume a website with 0 rootlinks and many deeplinks is getting a similar authority as if it had all rootlinks?

I know this is highly speculative, that's why at the beginning I was asking if anyone had some data/analysis on this subject.

Thanks a lot of your time!

I am a little surprised, because having those pages as "noindex, follow" should not bring GWT to flag them as errors.

Monica is correct in addressing google flag anything than 200 as errors, but... Your page with "noindex, follow" should return a HTTP code of 200. If it is returning anything else, it's probably wrong, and you should analyze why is doing it.

My religion has a law saying that GWT should return no errors, point. I have also witnessed few times a correlation between lowering GWT errors count to 0 and an improve in SERP ranking; but I have no proof one is causing the other.

I can see the total number of referring domains decreasing for cupidboutique.com while it's increasing for the two others. That's not good.

Usual metrics, like DA/Trust etc... are definitely better for pinkcherry who also has more backlinks, but cupidboutique values are supposely better than hushcanada.

But... The count of positions in SERP top 50 is surprisingly low for cupidboutique (209) when compared to pinkcherry (1.2k) and hushcanada (823), something is definitely wrong with your SEO on-page in my opinion; but it could also be your backlink profile, you should check anchor text diversity and backlink diversity.

In addition to all the good answers you already got.

Putting everything on a single page you lose opportunities to place keywords in url, <title>and <h1> or diluting <h1> value.</p></title>

I would be surprised.

Google knows a lot, but not everything. Unless GA tracking code is installed google shall not know about things such a user click.

If they were passing page juices only for clicked backlink they would be ruling out a too big chunk of the web. It doesn't sound logic to me.

Also it doesn't sound realistic to analyze all users click in the world when refreshing google index, they do have a lot of metal, but not that much.

Seems to be working. Today had no problem.

Sorry to report that, but it's either still broken, or broken again.

http://imgur.com/h3gWIui

Got this few times in the last couple of hours.

I have been using OSE over the weekend and it was still broken. Today it seems to work.

Thanks.

I must be misunderstanding something, how can you have 80-100 completely relevant keywords for each page?

they cannot be relevant for each page if each page is targeting only one keyword

of course each page will have other semantically near keywords in it, but 100? Doesn't make sense.

anyway, if you are targeting 100 keywords, and you have 100 pages, each page with content designed around one keyword

given that little information you are fine and I don't see any reason to cut the number of keywords/page, I don't see any logic in it and I am wondering why are you in doubt, what have you experienced or read to think otherwise?

How many words per page?

Of course adding unique content is the only way.

add unique categories description

if you have a blog, last excerpt of posts related to the category

alternative categories info

try to figure out possible user generated content, category related

define an algo to present different set of products for each category, even if many products fall on multiple categories you don't have to show always exactly the same

multiple product descriptions for the same peoduct, one for each category...

do some brainstorming and you will come up with more ideas

What make you believe limiting the number of keyword you are targeting would bring any SEO value?

unless you are stuffing to much keywords into the same pages I strongly believe it doesn't matter how many keyword you target

what it really matter is the backlink profile supporting those keywords

assuming you don't have unlimited resources, of course you cannot focus on unlimited number of keyword from the very beginning

make a list of all the potential keywords, filter by conversion rate and competition and choose a reasonable set of keyword to focus on at beginning

focus on agroup doesn't mean ignore the others, I mean I would still monitor other groups and I would still generate good content for those keywords if it make sense, if source materia is available if the topic is hot, etc...

Here you have it:

http://imgur.com/xNrxEuw

It's true sometimes it works, let's say 1 out of 3 queries, my guess is a load balancer is sometimes redirecting the query to something working.

xNrxEuw

Getting errors all the time.

A complete Bible then.

It's not complex to do, it's not harmful if done right, and it helps ranking.

If you want to lower bounce rate in analytics you can just trigger an event after a 30 seconds time out using one line of javascript.

But be careful in examining which figures is more interesting in your case, even after 10 minutes someone who leaves the landing page without browsing deeper should be considered a bounce in the majority of the cases in my opinion.