Help! My review posts disappear from the SERPs after updating them - am I overoptimizing them?

[Execute]

I have a site in the hobby niche, that I haven't worked on for the past several months, more or less since year.
I haven't updated the roundup review articles for this year 2022 (think "Best 10 X type of products" posts, eg Best 10 golf clubs). The posts were still bearing the year 2021 in title.

I decided to start updating them now.
Before updating it, this particular page was ranking 2nd-3rd for its keyword (wasn't a particularly competitive keyword).

After updating it and waiting for about a week for google to reflect the updates, it disappeared completely off the results.
Google only shows it if you search the keyword with my site name on it.
And even then, you have to click "show omitted results" at the end of the last page of search results before it shows it.
(I didn't reduce the number of instances the target keywords appeared while updating it. Rather I added more and more variants of the keyword while updating it.)

Furthermore, Google completely rewrote the page title in their SERPs listing of the page.
When I was updating the post, I changed the title such that I had repeated the keyword twice in the title using a variant of it, sort of like:
"Best golf clubs 2022: See the top golf clubs available. (Updated 2022)".
Google changed the title in their own listing by simply choosing the latter half of the title so that it reads
"See the top golf clubs available (Updated 2022)".

Now, I don't think this is just a transient volatility or transient change in SERPs while google gets used to the update, because I also noticed something similar last year, though I wasn't paying a lot of attention to the site at the time.
At that time, I noticed some of my main money pages (roundup posts) that were ranking in top 3 for their keywords suddenly disappeared completely to beyond page 100, without my touching or updating the posts. And they never came back out of that range to anywhere within the visible range, till this day.

I guess I must have done something terribly wrong for Google to yank off the pages so hard and kick them off to oblivion.

I haven't built any links to the pages so it's likely not about building spammy or toxic links.

I think I may have overoptimized some of the pages resulting in an algorithmic penalty.

Thinking back, I think the mistakes I may have made there include:

Mistake 1:
Including the keyword in too many H2 tags: I included the keyword in at least 4-5 H2 tags on some of those posts. By putting the best use of the product in the subtitle as part of the H2 tag eg:
For product number 3, 4 and 5 mentioned in the roundup articles, the H2 subtitles would be something like:
3. Tiger's Long nose Golf Club (best golf club for newbies).
4. Greenie Golf club (best golf club for pros)
5. Smiths Golf club (best golf club for those on a budget)

And so on for some of the products.
For at least 4-5 products in a roundup article, sometimes more.
I did this mainly because I had read that Google says in your roundup articles, you should make recommendations for which product will be best suited for what purposes or what niche.
So I thought, why not put it in the H2 tag subtitles, to make it more prominent to Google.
Now I'm thinking this might be too much.
My question on this would be: how many times should a keyword appear in H2 tags to be good for SEO while avoiding overoptimization?


Mistake 2:
I may have included the keyword in too many of the alt texts for the images.
There are usually about 10 images, for the 10 products mentioned in the roundup post.
I use the Amazon affiliate image link to embed the images from Amazon.
In like 5 of these images, I include the keyword or close variants of it in their alt text.
Is this overoptimization?
And how many times is it safe to include the keyword in the alt text of images. (Say for a 1500-2000 word roundup post, with 10-12 images of the products).


Mistake 3:
I may have repeated the keyword too many times in the body of the post.
How do you guys check keyword density reliably, and what is a safe range to aim for. And does this range include those in the alt text of the images (I doubt)?

NB: The long overoptimized post title of the one I updated is also an obvious flaw I think, that Google has already giving me a pointer to by rewriting it for me, so I'll correct that in the post.

Apologies for the long post, just tried to explain the situation as best as I can.
I really need help on this, and pointers to what I'm doing wrong will be appreciated.

I'm happy to provide any clarifications if needed.

Thanks a bunch in advance.

 

[Ryuzaki]

My question to you would be how substantial of an update was this? If you change enough of the text, you'll have to go into "recalculations" as I call it, meaning they no longer can believe that your content should rank where it's currently ranking. Too much of it changed and needs to be reevaluated. After that occurs, you tend to pop back in in a better spot if you optimized better.

Did you change 5% of the text? 50%? All new products mentioned? Or did you only tweak keyword usage around the optimization spots?

After updating it and waiting for about a week for google to reflect the updates, it disappeared completely off the results.

With enough of a percentage of the text changing, this isn't unusual. Just repeating that for newcomers who are skimming.

Another thing is that if you're trusting a rank tracker to tell you this, it could be wrong. It could be the rank tracker popping in with IP's from various locations, connecting to different database shards, etc. There's no real static, single SERP any more. There's a million split tests, location-based results, personalization, different unsync'd databases, and more.

Furthermore, Google completely rewrote the page title in their SERPs listing of the page.

Happens all the time, more often than not. Google thinks it knows what will get the click better than we do, and they rewrite our titles. It's annoying and there's nothing we can do about it.

In like 5 of these images, I include the keyword or close variants of it in their alt text.
Is this overoptimization?

I think this is fine. I do it every day and have zero issues. But I don't use the keyword no more than twice and others will be variations, and not all of them even have variations. It's an easy place to get greedy, and I force myself to exercise caution.

I included the keyword in at least 4-5 H2 tags on some of those posts.

This could be an issue but I've seen it done successfully many times. I've been nervous about doing it myself and haven't for that reason... it's not just an over-use but an over-use in specifically strong places that can be pretty obviously the work of an SEO.

A problem could be that by doing "best thing for newbies, best thing for pros, best thing at this budget, best thing no matter the price"... those are all different searcher intents. If someone wants to know the best one for a newbie they'll ask that question. When they want to know the "best" one without any qualifiers attached, that's what they want to know. In my opinion, all of those should be separate posts.

How do you guys check keyword density reliably,

I don't recommend checking keyword density at all. I'd consider switching your mindset over to keyword occurrences. Stick one use in the title and H1, in the first paragraph, in the last paragraph, and in a couple of H2's, maybe an H3 if you can, in one or two alt texts, and then maybe 3-5 more times spread out evenly through the article's text. Add more less in the text depending on the length but there's no need to go real high. If you want more, then it needs to be variations of the keyword.

 

[Execute]

My question to you would be how substantial of an update was this? If you change enough of the text, you'll have to go into "recalculations" as I call it.......

Did you change 5% of the text? 50%? All new products mentioned? Or did you only tweak keyword usage around the optimization spots?

I did a very substantial update that took me a few days to complete.
I replaced 5 out the 12 products reviewed in the roundup because they were no longer available on the market, so I rewrote each of those 5 sections for those products.
I also added some lines to the introduction and conclusion, rewrote the entire Buyer's guide section to give it more flesh and make it more relevant, as well as added a few more QnAs to the FAQs section to capture more related keywords and questions.

I also added a Summary section at the top somewhere in the Intro, which is a simple bulleted list of the top 5-6 products recommendations for the various purposes. Eg best thing overall, best thing for newbies, best thing for pros, best thing for those on a budget, etc. With affiliate links to Amazon in each of those bullet points.

I also made a long list of all the relevant related keywords/search queries for that topic from Ubersuggest, and tried to fit them in naturally into the text, as least as many of them as I could get to fit in naturally, which was about 1/3rd to half of the list of keywords I made.
So I ended up with about 3,200+ words total in the post, from an initial starting point of 2000+ words.
I would say I probably rewrote about 30-50% of the article, if you factor in the new lines, extra FAQs and buyer's guide sections I added.

With enough of a percentage of the text changing, this isn't unusual. Just repeating that for newcomers who are skimming.

Another thing is that if you're trusting a rank tracker to tell you this, it could be wrong. It could be the rank tracker popping in with IP's from various locations, connecting to different database shards, etc.

So in your experience, when a post goes into what you called "recalculations" phase after a substantial update, would it disappear completely off the SERPs into page 100 and beyond? Like, literally like a stone dropping into water....
I would have thought there would simply be significant volatility in the post rankings during that phase, like dropping from a top 3 ranking to page 2 or 3 or even page 10, rather than disappearing completely like a big stone in water.

And if it does disappear completely, like how long does this phase of disappearance tend to last, generally speaking?
(In my case it's now gotten to 2 weeks after I updated the article, and a week after I noticed the disappearance, and there's no sign of it appearing back anywhere in the regular SERPs yet....unless I search for it with my site name in the search term).

Also, to answer your question, I'm not depending on a rank tracker to tell me the rank. I searched the SERPs myself when I couldn't find the post in it's usual 2nd-3rd position, and realized it was nowhere to be found anymore. I trawled through the SERPs page by page as far as I could go, and still didn't see it anywhere.
And it also reflected in my stats for traffic and income, cos it was one of my site's main money pages for traffic and income.

Why I think this is more of an algorithmic punishment rather than a "recalculations phase" is because some of my other main money posts on the site disappeared exactly the same way last year, but that time was without my doing any updates or anything on the posts at that time. (They were still current for the year 2021 at that time, so it wasn't like they were outdated or anything like that). They just simply dropped off the face of the earth one day, probably after one of Google's numerous updates.
So I think while updating this post that just disappeared now, I may have inadvertently made some of the mistakes I had made in those earlier articles that made them disappear last year.
Hence my seeking clarifications to learn.

Many thanks for your detailed answers @Ryuzaki, they're very much appreciated.
I got a lot of helpful pointers from them.

 

[Ryuzaki]

So in your experience, when a post goes into what you called "recalculations" phase after a substantial update, would it disappear completely off the SERPs into page 100 and beyond? Like, literally like a stone dropping into water....

With substantial rewrites and additions and removals, especially the amount you did, it's pretty much a brand new piece of content. Even if it exists at the same URL, Google has no understanding of the new content without first running the recalculations. And they don't have any incentive to continue ranking it until they have.

So yes, I have seen them tank out of the top 100 like this. I have also seen the volatility you described where they just fall back to a later page. I think it depends on how much changes, how many articles they can surface for that query, etc.

And if it does disappear completely, like how long does this phase of disappearance tend to last, generally speaking?

Unfortunately, during big core updates, Google slows crawling and indexing way down and don't want new additions to the SERPs because they need to measure user metrics to determine user satisfaction.

It'll require them to crawl, reindex, and then recalculate the metrics from the page. I imagine that they have a lot of "offline" sub-algorithms that do this work before feeding it back to the main algorithm. Then your post has to roll back out into the live index and propagate across the data centers and all that. That can take longer during core updates, but I'd expect a couple weeks to a month usually, depending on the authoritativeness of your site. Could go longer since Google doesn't seem to have their "stuff" together like they used to.

 

[Execute]

So yes, I have seen them tank out of the top 100 like this. I have also seen the volatility you described where they just fall back to a later page. I think it depends on how much changes, how many articles they can surface for that query, etc.

Interesting, good to know...that's quite reassuring.

That can take longer during core updates, but I'd expect a couple weeks to a month usually, depending on the authoritativeness of your site. Could go longer since Google doesn't seem to have their "stuff" together like they used to.

Wow, duly noted. I'll just make the improvements and corrections you suggested, keep my fingers crossed and wait and see. (While doing the other things to improve the site generally).

Please I have a few more questions that have been troubling me regarding the whole updating old posts thingy, if you don't mind. ( @Ryuzaki and any other person that knows the answers).

Sorry, as I'm a relative newbie when it comes to all these, and Google search hasn't been as helpful as it used to be due to reduced accuracy of search results, as it gives me crappy results when I search them.

(If you prefer me to post them in a new thread for more visibility, please let me know).

1. Do you have any possible explanations for why some of my top ranking review roundup posts suddenly disappeared from the rankings last year (that's the ones that suddenly dropped to beyond page 100 without my doing any updates on them at all, and they weren't outdated as it occurred during the year of the review, which was 2021).

2. Do you Update or Republish roundup posts each year?

When you update your review roundup posts each year to tally with the new year, (eg going from Best golf clubs 2021 to Best Golf clubs 2022), do you prefer to update or republish the post?

Updating means simply updating the post without changing the Date Published to a new date for the new year.

Republishing means changing the date published to a new date in the new year, to make it seem like it's a very recently published article.

Especially as recent posts are more likely to be clicked on in search engines, increasing CTRs and possibly rankings.

Any reasons for which one you prefer would be appreciated.

3. For WordPress, any suggestions for the best plugin to use to show Date last updated on a post, instead of the original Date Published.

That shows the date well both to users and search engines.

When I browsed about it, I saw the WP Last Modified Info plugin, and used it for my site when updating a post.

However, it doesn't display the date well on my site. It doesn't alter the original published date on the page, it only shows the date last modified in a new line just before the content starts.

Also, it appears like it's not effectively submitting my Last updated date to Google. Because Google SERPs listing for that page still show my original published date as January last year, even after picking up my latest updates in their index. (I did the update for that page about 10 days ago, and it's now ranking well back in some of its old positions at number 2 position, so I know Google has picked up my updates to it. However my annoyance is that it's still showing the original published date of January 2021 in the Serps, instead of June 2022. And this will decrease CTRs as people seeing it in the SERPs will think it hasn't been updated since early last year.)

Many thanks for your help and for your time.

 

[Ryuzaki]

Do you have any possible explanations for why some of my top ranking review roundup posts suddenly disappeared from the rankings last year

Depending on the timing, you probably got caught up in the April 8th and/or December 1st Product Reviews Updates, and possibly the June and/or July 2021 and/or the November Core Updates.

2. Do you Update or Republish roundup posts each year?

Update. Some people will republish and do 301's but I wouldn't do that. I'd just update the existing URL.

Updating means simply updating the post without changing the Date Published to a new date for the new year.

Right, but your "Date Modified" is updated is any modern CMS, and you can display that. Instead of showing "Published: June 12th, 2020" you can show "Updated: June 13th, 2020". This is what I do to ensure Google understands and updates their freshness for the page.

I see you mention this next in your post. I don't know a plugin to do it as I develop my own themes, but they are out there. I think they just hook into the "published date" function and replace it with the "date modified" functionality.

Also, it appears like it's not effectively submitting my Last updated date to Google. Because Google SERPs listing for that page still show my original published date as January last year, even after picking up my latest updates in their index

You should confirm they picked up the updates by checking out the version of the content they have cached (you can do this from the SERP by clicking the 3 vertical dots next to your result's URL).

But you can't control what date Google chooses to show. It's like the meta title rewriting. It's out of our hands. They'll show the date they think will get a better click through rate. It's annoying. But it doesn't mean they didn't refresh the freshness score.

 

[Execute]

Update. Some people will republish and do 301's but I wouldn't do that. I'd just update the existing URL.

Just for clarification:
When I said Republish, I actually meant republishing on the same url, so there would be no need for a 301 redirect.
Like in Wordpress for example, after making the edits and improvements they want in the post, instead of simply clicking the Update button, they change the date the article was published to a new very recent date, and then click Publish.

So it's like a newly published post, on the same url.
When researching it, I read that Google still treats it basically as the same post since it's on the same unchanged url, and had the benefit of compelling Google to now show the recent date as the date published, as that is now the new date the post was published.
Instead of giving Google the option of choosing between the original published date and the last updated date.

I think I'll experiment with it on some posts and see if there are benefits to it.


All your other points are all duly noted.
Many thanks for everything.

 

[yuckystuff]

I've seen similar behavior for awhile which actually let to me changing my approach on these sort of "review update" pages. Put simply, I make new ones and link to the newest one from the top of the older ones. I don't 301, I let the link at the top work as intended (think "Update: to read our newest reviews for 2022 see our article "12 Best Blah Blah Blahs for 2022" and leave the rest of the content as-is. If I have 4 or 5 of these older outdated pages they all get the same treatment, and I simply change out the previous "Update" backlinks to point to the newest one.

Some of the content is re-purposed of course and the new page doesn't have the same backlink profile (obviously), but I've seen better results since changing my approach and leaving the older pages up with a link to the update.

 

[Terry Power]

I'd consider altering some of those keyword uses to variations on that main keyword. And give it a couple of weeks. Over-optimization is not a thing, Google does not penalize for it, rahter it tends to reward it from our testing, so don't worry about that. But this was, as you say, a major rewrite so allow Google to look it over. You might even hurry them along by putting your URL into the Google Mobile ready test page thus forcing a new crawl.

 

[dresden]

I'd consider altering some of those keyword uses to variations on that main keyword. And give it a couple of weeks. Over-optimization is not a thing, Google does not penalize for it, rahter it tends to reward it from our testing, so don't worry about that. But this was, as you say, a major rewrite so allow Google to look it over. You might even hurry them along by putting your URL into the Google Mobile ready test page thus forcing a new crawl.

OO is definitely a thing. Build me a page circa 2002 with same color font as the background repeating the keyword 10,000 times.

Also, you don't have to employ trickery to get a crawl. You can request indexing in GSC.

 

[Execute]

I've seen similar behavior for awhile which actually let to me changing my approach on these sort of "review update" pages. Put simply, I make new ones and link to the newest one from the top of the older ones. I don't 301, I let the link at the top work as intended (think "Update: to read our newest reviews for 2022 see our article "12 Best Blah Blah Blahs for 2022" and leave the rest of the content as-is. If I have 4 or 5 of these older outdated pages they all get the same treatment, and I simply change out the previous "Update" backlinks to point to the newest one.

Some of the content is re-purposed of course and the new page doesn't have the same backlink profile (obviously), but I've seen better results since changing my approach and leaving the older pages up with a link to the update.

Interesting approach.
My biggest concern regarding this would have been the relative lack of backlinks on the new post, but I guess the links from the old posts would pass along some of the link juice.
Also, I would be concerned about possible keyword cannibalization, where the pages are competing for attention in the Serps. But I guess this may not necessarily happen as Google will likely choose the current year over older posts.

Quite interesting, I guess this underscores the importance of testing and seeing what works for you, rather than just assuming something wouldn't work.

I'd consider altering some of those keyword uses to variations on that main keyword. And give it a couple of weeks. Over-optimization is not a thing, Google does not penalize for it, rahter it tends to reward it from our testing, so don't worry about that. But this was, as you say, a major rewrite so allow Google to look it over. You might even hurry them along by putting your URL into the Google Mobile ready test page thus forcing a new crawl.

Actually, the more I researched what went wrong with my posts, the more convinced I am it was slapped by an algorithmic penalty due to overoptimization.
I think I really repeated the keywords too much, and in sensitive places- so many h2 tags, alt texts, and a lot of places in the body.
I will "deoptimize" the articles to an extent by removing some of those keywords, as an experiment, and see what happens.