Welcome to the Q&A Forum

Matt Cutts repeatedly said many ccTLD are recognized by google algo as a country specific signal, while few ccTLD and generic TLD are just... generic and treat as such.

I don't think the TLD of your customer is the real issue.

How much you pay for the certificate has nothing to do with performance.

Google doesn't want your website to load under .25 seconds, they never gave a specific number above which they penalize or below which they prize. The general consensus is a TTFB (time to first byte) below 1s is ok, below .5s is great.

Keepalive is standard for browsers and servers these days, page load under https will be affected only by the initial handshake.

I never used wpengine, I have no idea how https is going to impact a website hosted on their platform, they should know if they have other customers who moved.

In my experience with the websites I moved from http to https the performance didn't degrade at all, both TTFB and full page load stayed the same, TTFB under .4s and page load below 5s; but no one of these websites were using either WP or shared hosting.

10$ vs 250$, the only differences are the brand selling them and their customer service, the brand theoretically affect conversions, if you believe your visitors are going to check the certificate authority and trust one more than another, I don't and I wouldn't go for the more expensive.

But I would go for the EV certificates, just because the browsers will show the website company name just before the url, and I have the impression that does instill trust. But it's just my opinion and googling around you can find opposite views on the subject.

single domain, wildcard or multi-domain, again from a technical perspective are all the same; multi-domain sounds very attractive but do have few drawbacks, first whenever a user check the certificate only one company name and one domain is listed, for all domains the details will be the same; furthermore (and more important) to use UCC certificates you also need a webserver supporting SNI, not all web servers do, and last but not least you need visitors using browsers supporting SNI (which rules out anyone still using windows XP and anyone with old browser versions)

with regard to SEO, are all the same, it's not the certificate by itself which matters, but to follow the right procedure when moving from non-ssl to ssl: http://moz.com/blog/seo-tips-https-ssl

Tracking ROI is not an issue. Branding without ROI is an issue.

Cutomer service for me is no marketing, you plugin any CRM software into your FB account and you have an additional channel to stay in touch with your customers and that's ok.

So we have one thing called branding, take the brand you think is more appropriate for FB branding, Coca Cola, Burberry, Tesla, whatever you want. And write down a formula which calculate the ROI of your branding investement on FB for that brand. I would love to see that figure.

I may sound naive, but in my balance sheet what it matters is sales, revenue and profit. Branding has a value as long as it bring customers, otherwise is charity to FB.

If you have any numbers showing your branding spending on FB translate into sales and thus can be called investement, I would gladly see those numbers, otherwise it's charity.

After googling about it and thinking better I choose not too. I think you are absolutely right, too inconsistent.

We don't use Comscore. Analytics transparently kept tracking everything without any change. We don't use Tagmanager url matching tracking, but unless you have not defined rules which include the url protocol it should not need any attention either.

Yep!

https://www.youtube.com/watch?v=UpK1VGJN4XY

Is it legit to show different content to http request having different referrer?

case a: user view one page of the site with plenty of information about one brand, and click on a link on that page to see a product detail page of that brand, here I don't want to repeat information about the brand itself

case b: a user view directly the product detail page clicking on a SERP result, in this case I would like to show him few paragraph about the brand

Is it bad? Anyone have experience in doing it?

My main concern is google crawler. Should not be considered cloaking because I am not differentiating on user-agent bot-no-bot.

But when google is crawling the site which referrer will use? I have no idea, does anyone know?

When going from one link to another on the website, is google crawler leaving the referrer empty?

In addition to what Leonie just said, one technique I first spotted on zappos.com and I often use is to put the excerpt of the content on top of the page with a read more button pointing to an on-page anchor to the full text to the bottom of the page.

It's not something you can always do without impairing UX, and I have no idea if google algo prefer that when compared to hiding the same content behind a js click.

But I also remember Matt Cutts discussing this in a video and say to do not worry too much because they do understand in today ui there's plenty of valuable content show through tabs, carousel and so on.

But you know, seems "don't worry too much" is one of his favourite sentence.

In my opinion is just a matter of waiting for google crawler to visit each page individually and update serp. In the past when I had similar issues with bad redirect was just a matter of fix it and wait for google to notice the change. Check last time google crawler visited each url, if you have historical data on google crawler visit and you know on average how often it does visit each page you also know how long you have to wait.

In case you choose the most expensive EV certificates as we did, for whatever is not directly visible, like the cdn serving js, css and images you can just use cheap 8 $/€ certificates.

One thing I forgot, if your server support SNI, don't use it.

We did initially, but soon found out some price engines could not read feeds, moz crawler could not crawl, and everyone on XP+IE was left out. So we disabled it.

Have fun.

It's a couple of weeks I experience this from different computers. If I write a short answer like this one I have no problem. If I spend some time writing is more likely than not to happen.

If I was you I would check keywords using semrush.com but seems you did a nice job already analyzing them.

I actually agree with what Monica said, theoretically is better to have two pages on serp for CTR.

But if I look at serp for those keywords and I open the two pages I feel a bad taste in my mouth, I have the feeling I have been cheated because the meta title before to click and the content after the click seems too similar (I didn't read through all the content I confess), but I may be biased since I know what you are trying to do, a regular visitor my feel different.

Nevertheless I think the theoretical benefit of having two pages in serp could be balanced by the ranking benefit of merging the two pages; iif a big jump was possible (http://moz.com/ugc/click-through-rates-in-google-serps-for-different-types-of-queries). Which seems highly unlikely, but you can test it.

To merge or not to merge?

I changed my mind, I would not merge. I would start to slightly change one of the pages, both meta-title and content, to remove that bad taste in my. And monitor ranking variation after each change.

That's what I am doing when I remember doing it. But I think someone at moz should look into the error.

What keywords are driving traffic to each page?

If semantically they share the same traffic, and if after merging the content and skimming the fat you don't loose too much, I would merge them and 301 one url into the other.

If each page is having traffic from different keywords you have the option of just do nothing and leave them the way they are.

Imagine being a visitors of those pages. Would they better serve your need for information split or merged?

I did it a month and half ago for a couple of websites.

Transition was smooth. I had to buy more ssl certificates than I thought for the many domains serving js css and so on... But was not a big hassle.

Just after moving from http to https I didn't notice any ranking change, and to have a good level of accurancy I monitor the same keywords with both moz ranktracker, proranktracker and semrush.

But in fact google is slowly recognizing the move few urls at time, each day you will notice some google serp start serving https url in place of http ones.

After a month we had a big jump in ranking, around +30% more keywords in the top 100 and a general increase in ranking for all the keywords already in top 10, top 30 and top 50.

But I have no idea if it's connected with the shift to https since we also constantly do many other things, get backlinks, improve on-page, etc...

At least it didn't seem to penalize the websites.

This is driving me crazy and making me wish to destroy everything I can reach with my hands.

I like to write long curated answers to people questions on here. And evidently there's some sort of timeouts in Moz Q&A software.

Because every time I answer to someone I get this fu#@!ng error message...

http://imgur.com/ypVQ2hD

So I go back, I lost everything and either give up or write one sentence.

ypVQ2hD

Everything went back to normality in GWT after 48 hours.