January 9, 2023 at 1:46 p.m.
- Google search results often show an incorrect update date for web pages.
- Our original research found that this defect results from Google parsing the date from the last comment on a web page.
- Even though you may present an accurate update date in your XML sitemap and structured data, Google will still parse the date from an incorrect source.
- Google users who see old dates in search results are less inclined to click through.
- The defect also affects rankings, as Google appears to devalue affected pages as outdated, even if they’re updated frequently.
- Besides blaming publishers, Google has done nothing to fix this defect, which has persisted for years.
Why You See Incorrect Dates in Google Search Results
Like most websites, at Appledystopia, we take Google rankings seriously because they’re a monopoly. Google is the World Wide Web. We constantly observe how our pages rank and appear in Google Search results. For years, incorrect search results have plagued our website, and I see it in others. Even major publications have gone to great lengths to convince Google that they have updated a page recently. Even so, Google will often show an old, incorrect date.
One of our top articles is updated every day. It’s about finding the best prices on Apple products at Amazon. We also have a similar article about Costco, plagued by the same defect. Even though we edit it daily and post the updated date on the site, both in content and structured data, Google will continuously show the page as being updated two years ago! Who would click on that? Despite high rankings, this page has yet to achieve an acceptable click-through rate because Google shows a two-year-old date next to our article, which is actually updated daily.
We tried adding the most recent date to the meta description to no avail. In one case, it showed both dates in SERPs. What a mess!
We checked all of our code. We looked at the structured data. We looked at our XML sitemap. Then, I finally just searched through the source code for the date that Google displayed in search results. Eureka! They’re getting it from the last posted comment.
To test this, I inserted a comment and waited a day for the Google bots to re-index our page. (If your site isn’t updated frequently, it may take longer for Google to index a page.) Finally, the search result appears with no date at all. This is better than a two-year-old date. We’d love to see a correct update date, but, in this case, we’re happy to see no date at all. Unfortunately, four days after posting a comment, Google shows that the page is four days old, even though it’s updated every day. We put in a lot of hard work to constantly update this page, and Google always shows it (and other pages) as being stale.
We also noticed that the article started ranking lower, especially for other related keywords. This seems to show that, internally, Google regarded our page as two years old and evaluated it as such — stale content.
This isn’t just a cosmetic flaw. It’s a severe bug that affects your rankings. We had a significant drop in traffic because of this defect and how it affected our top URLs. Unfortunately, Google looks at this as a webmaster problem, not a flaw in its search engine.
Google Claims This is Not a Defect; It’s Your Fault
Google provides information about dates in its Google Search Central documentation. Here’s a screenshot of where Google claims to get dates for SERPs.
Google clearly instructs publishers to use structured data for “published” and “updated” dates. This is all something straightforward with simple execution for publishers. Your SEO plugin is already doing this. Remember this because it gets a bit murky from here.
This is a screenshot from this site’s (www.appledystopia.com) source code. As you can see, we provide structured data in the agreed-upon format. We don’t do it manually. Our SEO plugin, RankMath Pro does this automatically. They’re doing it right. I’m doing it right. Google isn’t doing what the industry agreed.
Researching this issue, Google blames incorrect dates squarely on publishers and webmasters. John Mueller, a webmaster trends analyst at Google, serves as a liaison between publishers and the search giant. He claims Google works hard to find a correct date, but if it can’t, it just goes with the best option.
We disagree with Mueller. In all cases, we provided a correct date in XML in two places — structured data and our sitemap. Instead of getting the date from places where Google instructed publishers to put it, they get it from the last posted comment. Oops!
With all due respect to John Mueller, he’s not a developer. I’ve written code for 20 years, with ten years at a Fortune 5 company. I was two levels below the CIO, not some grunt code monkey. I personally know Googlers, and they’re not infallible. I don’t care if they went to Stanford and have a Ph.D. in computer science. It’s a bug!
THIS IS A DEFECT!!! We’ve tested this on a few pages and see the same behavior and how posting another comment is a workaround. Google takes the SERP’s date from the last comment, not our structured data or XML sitemap. We couldn’t make it easier for the bots to parse our date, but they still get it wrong. Should we remove all comments?
We’re not doing anything unconventional. We host with a popular service. We use the stock 2021 WordPress theme with minimal plugins. Cloudflare is our cache. I’ve cleared the cache so many times; it’s not even funny (but detrimental to site speed). THIS IS A BUG! Please fix it!
The Unseemly Workaround: How to Fix Incorrect Dates in Google Search Results
One way to fix Google’s date bug is to either wait for a new comment to come in or post one yourself. I dislike the idea of posting comments on my own site, for the sake of working around a defect. But we have no other options. Our top-ranking articles are all polluted with old dates, and inserting new comments is the only way we can fix it. We’ve tried everything else.
Unfortunately, you must post or approve a comment on your website EVERY TIME you update an article. I just checked on our Amazon article, and now Google claims it was revised four days ago, even though we modified it today with the latest products and prices. So, once again, I need to post a bogus comment because Google refuses to obtain the correct date from structured data, in-page content, or our sitemap.
Another alternative is to remove comments altogether. The whole notion of posting fake comments seemed wrong, and it’s also a lot of work. I commented out the comments to work around Google’s date bug. It’s a bug, not a feature, and not an ambiguity. I still kept our old comments and didn’t disable commenting entirely. One day, when Google fixes this defect, we’ll probably roll out comments again.
Would you click on an article about deals on Apple products that’s two years old? Of course not. That’s our dilemma. Either we forsake traffic and earnings, or we post comments on our own site, simply for the sake of a workaround. Sadly, it’s come to this. We, as a community, need to make it clear to Google that this is a defect, and they need to fix it immediately, but how?
SEO Experts Offer Bad Information
A recent article in Search Engine Journal adds more confusion to the mix. It mainly serves as a sycophantic justification for Google’s date handling. After all, Google staffs their team with a bunch of Stanford Ph.D.s, so they’re infallible.
The article even has a collection of suggestions, except for the structured data that the entire industry, including Google, agreed to implement. Yes, the “<time>” element is picked up by Google, but this is a defect. WordPress already uses this tag to define comment dates.
Anyone who knows anything about structured data would realize there are specific tags for publication and update date/time. Once we removed comments from our site, Google used these instead of the “<time>” tag. Remember the Google documentation we quoted above? It says nothing about the “<time>” tag. It tells us to use the structured data tags that we always used.
There are more specific structured data tags for supplying the date to Google. They actually work, as long as you don’t have any “<time>” tags on your page. If you’re using WordPress with its default comments, your dates will be off.
I think a lot of SEO articles intend to sabotage other publishers. If you follow the guidelines in Search Engine Journal’s articles, your SERPs will show incorrect dates. The author either didn’t bother to read Google’s documentation or is trying to mislead publishers so they’ll be less successful.
Again, let’s look at Google’s recommendation. It’s the same thing that SEO plugins do. The entire industry agreed this is how we handle dates, and a defect in Google Search abrogates this industry-wide understanding.
No Mechanism to Report Search Bugs to Google
Google has no real mechanism to take this feedback. I’ve reported it through their feedback mechanism on search pages dozens of times, and nothing happens. If you Tweet, it just gets dismissed as something you did wrong. (We canceled our Twitter account because it’s a waste of time.)
There must be a real way to report Google Search defects because their developers and QAT engineers missed this one. There are forms to report web spam and malfeasance by websites, but nothing regarding Google bugs. Their bug bounty program only allows users to report security flaws and defects in Google Play.
This is classic Silicon Valley hubris. Infallible Stanford grads staff Google Search. They could never make such a simple mistake with parsing dates in a webpage. The sad truth is, even a high school student could write better code than some of these “geniuses”. The proof is in the pudding. Go ahead and look at our page source. Look at our sitemap. Look at our on-page content. Do we have to put the updated date in a massive heading tag?
Just, please, Google, let us know what to do because you’re not getting this date from structured data, which you should. If this is not a defect, you need to communicate effectively to publishers how they can show a correct update date. Don’t blame it on us. We’re not all laypeople.
Many of us have jumped through every visible and invisible hoop that Google has put in front of us for decades. We all laugh when we hear, “just work on quality content.” Yes, that would be grand. Unfortunately, an Internet publisher must engage in the voodoo of SEO and the latest Google initiative, which has something to do with content quality, but a lot to do with jumping through invisible and visible hoops.
Google cooks up a new scheme every few years, like AMP, Core Vitals, structured data, etc. Instead of writing quality content, we fiddle with page speed, caches, plugins and get distracted from creating quality content. We must, or our traffic will fade away. I’ve opted into AMP twice, only to find it’s slower than our regular pages. I tried the second time, because I believed Google fixed this. They didn’t. AMP is dead slow, and even Google’s page speed tests show this, yet they still push this crap on publishers.
I remember when Matt Cutts was the search liaison, and he constantly advised publishers to focus on creating quality content. Yes! But that was never true. Hopefully, someday, Google will evolve to the point where they keep this promise. Google should enable publishers to publish instead of fiddling with structured data they don’t even bother to parse correctly.
I just want to write. I stopped writing code and started writing prose because I got sick of technology and ridiculous personalities. But now I’m pulled back into staring at code, XML files, and all of these extracurricular activities. Come on, Google. We want to publish, not fiddle. This severe defect forces publishers to stop writing and start fidgeting with maladroit workarounds.
We have 1267 pages on our website, most of which are indexed on Google. Instead of writing new content, we’re spending significant time checking URLs and trying to fix this issue with a kludge. We constantly hear how intelligent Google engineers are. We’d love to actually see that! If you can’t parse a date from nice, neat, standardized XML, blame it on publishers, and don’t fix it for years, there’s a severe organizational problem and a serious technological defect. This whole affair sheds new light on Google’s competence.