Welcome to the Q&A Forum

The trick is usually not to migrate from one CMS to anther, but instead to combine the best elements of each. For example I think Magento is a pretty big steaming pile from an SEO POV, but the way it handles eCommerce data is very efficient and regimented. So you'd want the back end on Magento and the front-end on WordPress. I imagine that with BigCommerce it would be much the same thing, WP is so established now from an SEO POV that it's hard to beat so you'd at least want to retain it at your shallow front-end, even if you powered the commerce system(s) a different way

It could easily be possible. Google usually takes the most specific directive when multiple directives contradict each other

If you previously had your site targeting 'international' in GSC and you had US hreflangs, Google would have still targeted to USA as US directive is more specific

Since you say you had targeting set to USA (and still do) and you may have had US hreflangs which were removed, this is a bit odd. Even without hreflangs, the GSC target US users directive is still more specific and thus Google 'should' default back to that

That being said, hreflangs might be a bit of a harder directive as they are actually coded onto your web pages. It may also be Google ignoring GSC directives for some sites to try and 'encourage' more webmasters to embrace hreflangs and proper internationally targeted websites

Google won't look at your bounce rate I don't think. Certainly not from Google Analytics (as that's your own data and it's also really easy to manipulate, e.g: you could use JS to detect what country people are connecting from and when you serve the pages you could strip out the analytics tracking script, thus lowering the bounce rate). Google use their own data, mostly the metrics from Search Console (clicks, impressions etc)

People talk about bounce rate being bad because it's bad UX (usually) and Google wants sites built with proper UX, sites that are useful to people. But the aim is to make your site better so that real people won't bounce as easily, not to get better GA bounce-rate metrics (which aren't used in Google's algorithm). Of course Google must have a way to evaluate something similar, but that number in that database wouldn't be 'the one' that they factor

I would try embedding the relevant hreflangs to support your geo-targeting and additionally checking what GSC property has your US settings in.  You can always block traffic from other countries, like how if you try to view a certain series on Netflix from the UK which is only available in the USA - it says "this show isn't available in your region". If you really care that much, just do something similar (this page is not available in your region)

The weighting of internal links is vastly inferior to the weighting of external links. I guess if you had no (or very few) external links hitting your site, the internal links might skew Google's view a little. But if some of your pages had even a few robust links coming from off-site (which are perceived as less biased links) then the internal linking wouldn't create any drastic movements. Footer links are often largely discounted, anyway - as they are not prominent and web-users don't often use them

This is good advice. Canonical tags would be a weak fallback compared with picking where you want to spend your efforts and choosing one 'main' site. As such, just like Alex has suggested - redirects and site consolidation are probably the best bet (this would also bind the backlinks for both domains, to one domain)

This could end up being pretty technical depending upon current site(s) performance. If it's done incorrectly (e.g: using 302s instead of 301s) then it could be a disaster

Alex's advice is right but just make sure you're careful how you approach this

Hi there Poppy!

For my own benefit and for the benefit of others who may wish to look at helping you solve this, just going to pop the spam results URL here:

• https://analytics.moz.com/pro/link-explorer/spam-score?site=poppyporter.co.uk&target=domain

This is the spam score which is worrying you:

• https://d.pr/i/wl6Dz0.png

And these are the links which Moz thinks are bad:

• https://d.pr/i/tNjbse.png

These are the notes from Moz on why linking sites and / or pages get higher spam flags:

• https://d.pr/i/SDKfmc.png

Basically if Moz identifies that a linking site shares common features with other sites that have factually gained Google penalties, then links from those site will raise the spam score higher and higher. For some reason Moz seems to think that Poppyporter.com fits that pattern, hence the high spam scores for those four links

These are the URLs for the offending web pages:

• https://www.poppyporter.com/racing_wings_earrings.html
• https://www.poppyporter.com/two_finger_ring.html
• https://www.poppyporter.com/large_ring_disc_model.html
• https://www.poppyporter.com/large_connections.html

If you visit those URLs, you will see that the site is not in fact parked at all. Although those pages all result in error 404s, if the site were really taken down - how come the Poppy Porter logo and branding are still there? How come featured products are still listed? If the site were parked and taken down, it wouldn't have your stuff still on it. Clearly it does and it is not parked (hence the custom 404 pages are still rendering)

If your site were really parked, the 404s would look like this instead:

• https://d.pr/i/Qqtojd.png

It would be a blank white page with no branded styling. Nothing except for the server's response! The fact that it's still all styled, shows that your old site is still online. Individual pages may have been taken down, but at large that site still exists

Interestingly, if you click any link on your old site, it moves you to the new site

Let's look at one of the 404 pages on your old site, which does not redirect to your new site and is still live:

• https://www.poppyporter.com/racing_wings_earrings.html

As you can see, there's no redirect to the new site:

• https://d.pr/i/gJ7uKT.png

If we look at the source code of a live (but 404-ing) page on poppyporter.com (say if we examine the shop link in the top-line nav):

• https://d.pr/i/dwdVut.png

You can see very clearly that, you have two sites live and one is linking to the other!

I'm not a betting man and I don't have loads of info available, but if I _were _a betting man I'd say, this is what has happened:

1. You didn't shut down your old site properly, leading to a 'broken' old site with loads of 404s and broken styling
2. You tried to 'point' your old site to your new site with some kind of link architecture update, maybe in the DB or CMS
3. Moz saw that the old site was all janky and broken with loads of broken looking and 404-ing pages
4. Moz gave your old site a higher spam score than it used to have because now it's all bork'd (broken)
5. But there are links from the old site to the new site
6. So now your new site is inheriting the negative spam scores generated by the botched move from old site to new
7. It's possible that the old site had legitimately high spam scores, but I can't say for certain. Certainly after this botched attempt to move across, you have ended up with a spammy looking (and acting, in terms of status codes) old site. Because of the links, those negative scores are being pushed from the old site to the new one
8. As stated I don't know if the old site had high spam scores prior to the botched move and you need to investigate that with some urgency, as it will entirely shape your next movements
9. If the old site was very spammy and you were trying to make a 'clean break', you have undone that by interlinking the old site with the new one. Certainly you'd want to perma-kill the old domain as a matter of urgency! Kill the links, kill the pages that are still up, do not redirect from old site to new etc.
10. If the old site had good metrics before the botched move, you want to replace all the links and live-URLs with 301 redirects. Only do this if you are certain that Google hadn't penalised the old site. This can be a bit of a technical thing to do

Anyway the long and the short of it is, yes your old site is still live - yes it is linking excessively to your new site! One way or the other, it needs sorting out...

404s are usually for pages that 'don't exist' so that's pretty usual. This is either:

1. somewhere on your site, links are being malformed leading to these duff pages (which may be happening invisibly, unless you look deep into the base / modified source code). Google simply hasn't picked up on the error yet

2. something is wrong with Rogerbot and he's compiling hyperlinks incorrectly, thus running off to thousands of URLs that don't exist

At this juncture it could be either one, I am sure someone from Moz will be able to help you further

It depends on how your unique website creates URLs and how they are formatted, it also depends upon the current contents of your robots.txt file. If you can share some examples of URLs that are blocked, which you think should not be blocked, and also the contents of your robots.txt file, someone can probably tell you what you did wrong

Thanks for your answers Tom I think I agree with most, especially the increasing utility of BERT and the increasing focus on E-A-T. I don't know about you, but I feel that E-A-T was significantly tightened in 2019, and that's set to continue

Well if a page has been removed and has not been moved to a new destination - you shouldn't redirect a user anyway (which kind of 'tricks' users into thinking the content was found). That's actually bad UX

If the content has been properly removed or was never supposed to be there, just leave it at a 410 (but maybe create a nice custom 410 page, in the same vein as a decent UX custom 404 page). Use the page to admit that the content is gone (without shady redirects) but to point to related posts or products. Let the user decide, but still be useful

If the content is actually still there and, hence you are doing a redirect - then you shouldn't be serving 404s or 410s in the first place. You should be serving 301s, and just doing HTTP redirects to the content's new (or revised) destination URL

Yes, the HTTP header method is the correct replacement when the HTML implementation gets stripped out. HTTP Header X-Robots is the way for you!

FYI!

Only her homepage 301s to her new site.

Some of her deep-link 404-ing URLs do not 301, and those are the ones which Moz picked up:

https://www.poppyporter.com/racing_wings_earrings.html

If you load the above URL, you will notice that in the URL bar it **does not **change to .co.uk. The 301 is homepage (or maybe 'key-page') level only

The domain is in fact, not parked and still hosts custom 404 pages and other content :')

Oh dear this has gotten into a bit of a mess!

For a start, there's an error in your question which may be causing your to get fewer than average responses. You have embedded two links which look like they point to different pages, but you have encoded both the links in your question to point to exactly the same site (even though the text says they should be going to different URLs). This may be confusing people whom are seeking to answer your question

These are the real links for anyone who runs into this:

Old site: https://www.selldealsmango.com/

New site: https://www.dealsmango.com/

Where you say: "I have this site https://www.dealsmango.com/ which i have selected for canonical , but google is still selecting  my old website" - no, you have not fully selected the new site as canonical, as the old site (here) does NOT canonical to your new site, instead the old site still canonicals to itself. The new site canonicals to itself to, but the old site does NOT canonical to the new site

https://d.pr/i/P70o8N.png (screenshot)

Where you write: "only one page with new site link, and also put 301 redirect" - you have contradicted yourself. You say there is a 301 redirect, but also say the page is still up with a link on it. If there were a 301 redirect, all users and bots would get redirected before the page loaded and you would never see the link (at all) as pages which redirect cannot support any source code

The fact that the page is still live with a link on it, means that there is NO 301 redirect. Maybe you have redirected some sub-pages, I don't know. But your homepage is your post important page, so failing to 301 redirect that (which you have not done) is a big error. Fix that ASAP if you do nothing else

Also make sure you use Google's change of address tool from within the old property's Google Search Console dashboard:

https://support.google.com/webmasters/answer/83106?hl=en

Even if you follow all of these steps, it may take a long time for performance to return now. That's because this migration was handled in a very strange way, what's done is done

Fix what you have implemented with a proper 301 redirect migration project down to the granular (A-to-B) URL level including historic URLs which you can pull from backlink destination data, as well as from Search Console (or analytics by blending the hostname and landing page dimensions)

You'll have to just hope for the best!

That is technically an error. If a page or post doesn't exist in a certain language, hreflangs pointing to it (from other pages) shouldn't exist. When the post is published, then the hreflangs should be written. It's probably not going to destroy your sites rankings by itself, but in SEO, the second you start making concessions on one front - people argue that you should make more concessions in other areas. Before you know it, you have a big tangle of loads of different errors. Optimisation, is about making things 'optimum'. It's not optimum to have hreflang errors like that

Thanks for your response EGOL, I always really value your comments. I think this makes a lot of sense. How many times have people tried to game the SERPs with minor tactical deployments, only to see Google drop it all within 6 months? Focusing on delivering that which Google's algorithms attempt to seek (high quality content), can only be positive

This is a really good answer. You can also look for messages from Google saying that "mobile first indexing" was enabled for your site / GSC property

I am assuming the site isn't yours and doesn't belong to you. That makes things trickier and it means you'll be working with estimates, but it doesn't make things completely impossible!

There are tools which monitor Google organic search positions and paid ads (AdWords) in order to specifically estimate the Google traffic which a site or web-page might be earning.

SEMRush is one tool (paid, though you can get some limited data for free):

• https://www.semrush.com/info/wikipedia.org

You can see a little chart here:

• https://d.pr/i/P2GoDU.png

They basically look at search and ad rankings and then estimate traffic by pitching CTR (click through rate) data against ranking position data. It's very ballpark, but it is at least something you could look at. SEMRush also has a "chart tool" whereby you can input sites and metrics and build comparative charts (quite nice). Keep in mind you can only see traffic data by Google region (e.g: USA / Google.com or UK . Google.co.uk) and getting to an all-encompassing global view can be tricky (and time consuming)

Ahrefs costs more but we have found their traffic estimates to be more responsive, accurate and easier to manage (especially if you just want to see flat global data):

• https://ahrefs.com/site-explorer/overview/v2/subdomains/recent?target=wikipedia.org

Again, it's a paid tool with some free data. You want to be looking here:

• https://d.pr/i/4KJVOM.png

The only thing you are seeing there, is organic (SEO) traffic, not PPC (paid search) traffic. Although the data is narrower from one perspective, it's much easier to get global data. In addition to that, I do find the data to be more accurate, especially since the search volumes which Ahrefs uses within its traffic estimates contain click-stream data to enhance accuracy:

• https://help.ahrefs.com/keywords-explorer/faqs-and-troubleshooting/how-accurate-is-keyword-search-volume-in-ahrefs

That being said, nothing is perfect:

• https://help.ahrefs.com/site-explorer/faqs-and-troubleshooting/why-dont-the-organic-traffic-numbers-reported-by-ahrefs-match-those-i-see-in-google-analytics

Some people have even done studies comparing different tools, including SimilarWeb's traffic forecasting which I have little / no experience with:

• https://www.screamingfrog.co.uk/how-accurate-are-website-traffic-estimators/

I heard down the grapevine a few years back, that SearchMetrics (which is quite expensive) is actually pretty good at all this stuff. What I was told was that, they buy ISP (internet service provider) data and use that to forecast and estimate traffic to different sites instead. If that is true, it would be by far the most accurate traffic data available, but would come at a premium.

I believe that earlier on in my career, I had a little play and it was pretty robust. Platforms change though, you'd have to contact them to see if they still offer similar data.

As you can see, there are different options with different strengths and weaknesses. Accurate, digital, legal competitor surveillance technology is seldom free. Even if you're not buying the data from the competitor (who would never sell it to you! Plus it might be fiddled...) - you're still paying someone. Either someone who resells ISP network data, or someone who has 'done something clever' to get more accurate Google-traffic forecasting than the rest.

Hope this helps

Check using the organic search estimates in Ahrefs, those are usually pretty good and they keep quite a long history. You can also get similar estimates from SEMRush

So from this I am assuming that you have a page with a URL string somewhat like: "examplesite.com/bike-tours". On that page you have many different regional bike tours listed so a hypothetical example might be: "examplesite.com/bike-tours/texas-bike-tour".

I don't think it's really a huge problem to leave the URL like this, particularly as it would cover the singular and plural instances which is neat. Your question though moreso referenced the on-page content and link text.

Say you have your bike tours page ("examplesite.com/bike-tours") and then on the page you have four links to four separate bike tours:

• Texas Bike Tour
• California Bike Tour
• Florida Bike Tour
• NY Bike Tour

Obviously as you stated this is probably in some kind of grid format where you have image tiles and the text links underneath.

Since at the URL layer, you pretty have "bike tour"(s) covered off, if you can find **highly relevant synonym terminology that may be advantageous **(in terms of your ranking-relevance footprint). Whilst hammering home the same thing in the way you have stated is unlikely to garner any penalty, it doesn't do 'the most' in terms of connecting with the wider search-entity (more of a thing since Google Hummingbird, but realistically entities have always been around).

What about references of "bike touring" or "Touring Texas by Bike".

Be careful. Motorcycling and conventional cycling have muddy waters in terms of their search entities and you don't want too much overlap, or to connect too much with the wrong audience. Is this cycle bike touring or proper biking / road touring?

Think of some variants. Underneath each image but before the anchor-text based link, write up your own little snippets / descriptions of each bike-tour. Use that area to really diversify and include some synonym terms.

If you don't have time for all that, don't worry - I would be highly surprised if what you have done garnered any kind of penalty. Just remember: avoiding penalties doesn't mean 'good SEO'. It's the bare minimum!

An "SEO enabled" site is not necessarily "SEO friendly" so careful how you sell it...

1.) If you mean that you are looking at the page (front-end), and the editing (admin) bar isn't there on that page

It's quite common for some sites to have the overlay disabled on the front-end, for some or for all pages. Sometimes even if the admin bar is enabled, for pages which use custom taxonomies or page-types, it won't show up. This means that you'd have to find the page in the back-end of your WordPress dashboard

2.) If you mean that in the back-end, on the admin menu - the page is missing fields to edit it

It's common for custom page-types to lack basic editing features, if the developer(s) who created that 'type' of page did not do a thorough job. In this situation, unless there's a plugin which will force-enable the fields (unlikely) - you would need to pay out for more dev work. It means your original spec or brief was incomplete

Yeah you can do language only hreflangs. But it's pure nonsense to direct Google to the very same URL and state that it is the URL for all of those different languages. At the end of the day, Google will crawl from one data centre at once which may be from one of many countries. It will see one version of the page, and assume that 'this is what the page is'

If the site structure is that you have one URL only and the contents are modified based on the user's origin, then the structure is wrong as Google will have a very hard time ranking one URL as many different URLs. People who have such a structure always end up here, always argue why it's ok and then end up 'doing it properly' later on as it just doesn't work

Also note that, if you have one version of a page served to people in different regions (e.g: an EN page which is stated in the hreflangs to be for both Canadians and Americans), Google may see that as a 'minimum effort' deployment with no value proposition. Different audiences need tailored content to suit them, so a re-write of some of the content is still expected if you want to see an increased international footprint (and you're not a giant like Santander or Coca-Cola)

The number of times I see people clone their EN site into a US folder and just 'expect it to rank' with no extra effort, just with hreflangs - is staggering. Google expect to see a value proposition when you build out your site. Value-prop ('value add'), the #1 yet never talked about ranking factor

I don't think your current implementation will work very well, if at all. You may have lots of human-brain reasons why it should - but crawlers are robots

"Firefox: Beginning around Firefox 60, the Firefox browser began adding in a class to the HTML tag: . You won't see this alteration with Firefox on Linux or Mac; it's only on Firefox for Windows." - source

It's "Moz" as in "Mozilla", not in reference to "SEOMoz" I think it's browser native and not a hang-over of the Mozbar uninstallation. I can see why you would think that, though