Welcome to the Q&A Forum

This started a couple of years ago, but it still only happens in isolated cases. See this post:

https://moz.com/blog/i-cant-drive-155-meta-descriptions-in-2015

We believe it's tied to Featured Snippets and Google parsing answers from sites (they share a core engine, even though you may see long snippets on SERPs with no Featured Snippets). In cases where the snippet is deemed highly relevant, Google may present more information. It's not an across-the-board length increase, though. Most snippets are still restricted to the traditional length limits.

Our new Site Crawl can provide a full audit of your site (whether or not pages have issues) with information like crawl depth and Page Authority, but both tools have pros and cons. We use a cloud-based crawler (pardon the term) and Screaming Frog runs on your own machine -- even this has pros and cons, depending on your use-case. We now allow on-demand recrawls for Medium Tier on up. Screaming Frog has more customization, if you're looking for specific patterns and generally offers more raw data, I'd say. Our new Site Crawl is a bit more accessible to most users and helps provide insights, but doesn't have a lot of customization options right now.

If you've got a Moz plan that covers the size of your full site, then either Moz Site Crawl or Screaming Frog should be able to help you audit your full list of pages prior to an HTTP-->HTTPS migration.

I can't replicate this in live SERPs, but Matthew's explanation seems to match the evidence. It does appear that Google is doing some kind of implicit canonicalization. It's rare, but it does happen.

You could try giving the PDF a unique title in its own meta information (the document doesn't seem to have any). It's more likely to look like a duplicate if that's undefined.

The next step would be to recreate the PDF slightly and make it unique, but honestly, each step is going to get to be progressively more work. It really depends whether this is just an anomaly or it's causing you actual issues as a business.

The public metric is "absolute" in the sense that it's not relative to surrounding days. We have other metrics (not published) that are various types of moving averages, but I've found that most of them over-smoothed. Given the new normal lately, though, it may be time to revisit them.

Q1.

I'm not a domainer, but it's been my general experience that trying to negotiate directly with the owner has a much wider range of risk/reward. In other words, you might be able to contact them and get it for $250, or they might be completely insane and want $250,000 for junk, even though every other person on earth knows that's ridiculous. With the fixed-price sales, you can at least make that decision of whether it's worth the price tag for you.

That decision, though, isn't primarily an SEO decision, IMO, it's a business decision. Paying $1,000 for the right domain could be an incredible bargain if this site is your bread and butter and the name is critical. It could be a huge waste of money if this is your personal blog and a hundred different names would do just as well. It's really hard to advise you in a vacuum. If you asked me if $1,000 was a good price for an iPad, I could confidently say "no", but the value of a domain is highly subjective.

Q2.

I don't believe there's any kind of outright penalty. Hyphenated domains do often correlate with spammy domains, and it may be one signal of many Google considers, but I think they look at it clustered with other factors (that's speculative on my part). So, if your domain looks spammy on other dimensions and you've got a hyphenated, keyword-loaded domain, then yes, it might cause you problems. I don't think you'll get whacked just because of the hyphens.

I do think that hyphenated domains have a general trust issue with humans, though. It's not insurmountable, but I tend to agree with Michael -- there's always a non-hyphenated variant or an alternate TLD (even a .co) that's worth considering. Unless you're Eugene's Discount Widgets and your board will not settle for an domain but a .com with those three words in it and no other words, there are probably non-hyphenated options worth exploring.

Unfortunately, we are seeing a rate of change that won't seem to settle down. It's not clear if this is the new normal, so to speak, or if something about how we collect the data is creating more noise. There are no errors of the obvious sort (or the sort we've encountered in the past), so it's been hard to pin down. I'm hoping to know more soon.

Truthfully, I'm toying with the idea of going with our larger, but less frequently updated, data set, and focus more on things like SERP feature changes, HTTPS, etc. At this point in Google's evolution, I'm not sure how useful day-over-day flux is. The original goal was to help use flux to pinpoint what changed, but the reality is that -- while I can tell you which keywords or categories changed the most -- using that data to interpret how the algorithm moved is nearly impossible. There are just too many things changing every day, even outside of the algorithm. So, I'm actively pondering if the current role of the tool is useful in 2017.

Best of luck! Definitely interested in hearing back about how it goes.

There's no perfect solution, but Google's advice is to use rel=prev/next. This looks like pretty classic pagination. Rel-canonical is a stronger signal, but it's generally going to keep pages 2+ from ranking.

I've seen a relationship between some of those carousels and search volume, but it appears to be more complex than that (and I think there's a Knowledge Graph component). I don't think markup is having a big impact right now, except for highly specialized results (like recipes). Haven't gotten a good answer from Google, though.

Can a .ly domain rank on Google.com? Sure. Here are many that do:

https://www.google.com/search?q=site%3Aly

Do they have a ranking disadvantage? That's a lot tougher question. It's definitely in a gray area. Thanks to Bit.ly and some others, there is some precedent for these domains to be treated as generic, if other signals suggest they are not Libyan domains (geo-location, language, etc.), but there's always more risk for a new site and a chance you'll need to spend more to get the same results.

The political situation is much more worrisome, though, because none of us can predict that. If there's a history of .ly domains being confiscated and if that confiscation was considered legal (even in the short-term), then there's no guarantee that won't happen again. That's a risk we can't calculate in SEO terms. The fact that even Bitly eventually went to bitly.com doesn't bode well.

My gut feeling is that (1) there are no magic domain names, and you're going to have to advertise/market/promote any new domain, and (2) given the geo-political situation there is more than average risk (vs., say, a .co domain or one of the new TLDs). There's a lot I don't know about your situation, but I think your hesitation is warranted.

I can't do much more than speculate, I'm afraid -- I haven't seen any write-ups that seemed to do more than guess. Traditionally, the carousel results were Knowledge Graph driven, but that seems to have evolved. Some carousels are clearly based on search volume. See, for example...

https://www.google.com/search?q=best+movies+of+1984

I guess it no longer states this, but at some point I swear there was help text claiming this was tied to search volume. So, the Knowledge Graph is probably providing the movie list, but volume is driving who "ranks", so to speak.

I suspect something similar is happening with these new business listings, but I'm not sure. Google hates when people manipulate search volume, so they're very unlikely to confirm this.

I don't believe this is connected to any kind of schema or on-page markup at this point. Updating your Wikidata information is probably a good step, but not a sufficient one to get into the list (I suspect).

Honestly, we've seen such high flux in April -- including that week -- that it's really hard to pull anything useful apart. We have no confirmation of an algorithmic event (other than the usual "There are events every day" line).

There was talk last week about "Project Owl" -- Google's multi-pronged attack on fake news, but it appears that only a part of that is algorithmic, and that part has been rolling out over the past couple of months. I don't think the flux from the week in question is related to Owl.

Hopefully, someone in the community has some more insight on the vertical-specific data you're seeing.

Unfortunately, since MozCast is a side project, we haven't set up an SSL certificate yet. So, you're likely going to have problems using the widge on secure/https sites. How much of a problem it causes (warning vs. error) may depend on your site setup.

I don't have an ETA on fixing that, but we'll add it to the queue, and I apologize for the inconvenience.

Google testing tool seems to be showing a Software Application schema but not the reviews. When I pull out just the review source code, it's saying that the item being reviewed is missing. Honestly, though, trouble-shooting these can be tricky sometimes. I'd check the tool, though, and just take a look at the errors:

https://search.google.com/structured-data/testing-tool/u/0/

It is possible to have the schema correct and still not get reviews, though, if Google simply decides not to display reviews for that SERP. I'm not seeing anyone with reviews for the first give pages of that query, so that's a solid sign that this may be more than just mark-up. Google seems to ramp up and scale back which queries get reviews all the time, and there's no way to track that (other than their presence/absence). Most you can do is double-check the code and hope for the best.

Sorry, I know that's not a very satisfying answer.

Yeah, unfortunately, that's core Knowledge Graph (not a Featured Snippet), so it's coming from a partnered data source. I'm not sure where they're pulling this from, though, as the Wikidata/Wikipedia entries seem correct:

https://en.wikipedia.org/wiki/Free_Shipping_Day

There's not even a "report" option on this one, so I'm afraid Alick300 is right - you're going to have to contact Google directly, I'll ask around on Twitter, but Google isn't transparent at all about these data sources.

There's no perfectly clean break with any domain for one important reason - back-links. If the side project has solid back-links and then you disconnect it (break the 301, essentially), you'll lose the positive impact of those links. There's no easy way around that.

At the same time, the new owner will potentially benefit from those links. The only way to disconnect that benefit is to individually request that the linking sites switch their links to your site. If the new site changes ownership and drastically changes content, then Google will eventually see those links as irrelevant and devalue them, but that's probably going to take a while.

I'd worry more about your loss of benefit than about the new owner getting some unearned but probably temporary benefit, honestly. The other issue would be any brand awareness around the secondary domain - since that's tied to the name/domain, you can't really cut that loose.

Since it's a side project, I suspect this isn't a big risk, but it's really tough to know without more details.

I have seen reports that bad/old dates are impacting CTR, but the argument for how this impacts ranking is long and perilous, as we all know It does seem likely that bad dates at large scale would impact a wide range of engagement metrics.

We've seen a lot of reports, including Q&A questions, of sites showing bad dates in Google SERPs. I've verified this bug in the wild. There are also reports of bad dates being caused by YouTube embeds, with Google taking the video date instead of the page date. I can also confirm this is happening, although I don't know if it accounts for all of the bad dates.

Some people are reporting that these bad dates showing up corresponded with ranking drops. Usually, I would treat that as a coincidence (Google could easily launch an update and have a glitch on the same day), but in some of the reported cases, removing YouTube embeds led to ranking recovery soon after. I can't verify this, but I can't disregard it. There seem to be multiple reports of this recovery.

I'm in communication with a Google rep, and they are unaware of any direct connection between a bad date and ranking (such as some kind of QDF effect). I've passed along some data, and they are investigating, but there may have been multiple updates in play that are making for noisy data (even for Google).

There did seem to be heavy algorithm flux on November 10th and 18th, with some people speculating the latter spike was a reversal of the former. I have no evidence to support this, but MozCast data and chatter do seem to support both spikes.

If you've been affected by this problem and the ranking drops are severe, it's worth temporarily removing YouTube embeds (if feasible). Replace them with direct links (or maybe a linked thumbnail) and have Google re-fetch the page. I can't guarantee it will work, but the risks are low. It's easy to restore the embed.

Update (11/22) - Gary Illyes is saying on Twitter that the date problems have been fixed. If you see the proper dates cached, but have not seen rankings recover, then these may be unrelated events.

Unfortunately, they're being pretty tight-lipped on this one. Seems like a glitch, but they don't seem to think it's related to the rankings drop. Possible two events co-occurred, and there was an algo update at the same time as the glitch. Honestly, though, it's not clear at all.

There seems to have been significant movement in the SERPs around November 10th in the US (which could have hit the UK a couple of days later, although I have no data to support that). Unfortunately, we don't have much in the way of details or confirmation right now.

Are you seeing any specific pattern in the loss? I'm not quite sure I understand what the graphic is showing.