Why You Should Update Content – Or Risk Losing The Traffic You Have [Case Study]
[Free Consultation] Are you spending money on advertising but not getting the results you want? Are you looking for more sales and leads but have no idea where or how to start? Get help from our world-class marketing experts in a free consultation call.
Click Here To Schedule Your Free Consultation Now
Content refreshing (or content updating or content upgrading) has become increasingly popular among content marketers.
- HubSpot released statistics stating that updating old blog posts increased traffic to old posts by 106%.
- Neil Patel also said that he has three full-time employees that update close to 90 posts per month.
Here at Single Grain, we also update our content. However, as someone who updates a portion of that content, I noticed that not all posts perform exponentially better after the update. In fact, not only do some posts remain flat after an update, but sometimes the traffic even tanks.
So is updating content worthwhile? And if it is, what’s the most efficient way to update content?
I sorted through 42 updates that the Single Grain team did over the past couple of years and uncovered patterns in posts that perform incredibly well and those that perform poorly after an update.
In this post, I’ll go through what I did and what I discovered. You can use these statistics to share with your team and update your content refreshing system.
How I Gathered the Data
I chose to use only data from the Single Grain blog, as this ensures that all blogs were updated under the same guidelines and reduces variables in the data.
The tools I used are: Ahrefs, Wayback Machine and Google Analytics.
Here are the Single Grain updating guidelines:
- Rewrite the intro and conclusion
- Delete/combine any weak sections
- Research and write as much as you feel would add to the value of the piece
- Update any stats/facts/quotes and make sure all links to sources are no more than ~2 years old
- Replace any outdated or irrelevant images
- Replace any examples/case studies with new, current ones
I also predicted that a website’s domain authority could impact post-update performance, so by using data from just one website, we eliminate additional variables.
Once I had a list of blog posts that had been updated, I sorted the data to find answers to questions like:
- How long does it take to see results?
- Does changing too much copy harm your rankings?
- What kinds of posts tend to perform the best?
- And much more…
Updating Blog Posts Increased Traffic by 96%
Ultimately, updating old content increased our traffic by nearly 96%. In other words, it nearly doubled the traffic of the old posts.
Out of 42 posts that were updated, within six months:
- 25 posts generated more traffic
- 17 posts generated less traffic
The total traffic these 42 posts generated:
- the day before the updated version rolled out was 9,207 visits combined
- exactly six months after the update was 18,030 visits combined
Therefore, we gained 8,823 monthly visits from these 42 posts, nearly doubling our traffic, just by updating content.
The verdict? Content updating does work.
Learn More:
* How To Write Data-Driven Posts
* How We Doubled Our Traffic (Without Writing Any New Blog Posts!)
* Do You Really Need to Write 1,890-Word Blog Posts to Rank on Page 1?
Don’t Worry About Ruining Your Posts and Losing Traffic
Earlier, I mentioned that some of the posts I updated actually performed worse after being updated. However, the 8,823 monthly visits we gained from these posts also includes the loss of traffic from some of the posts.
Furthermore:
- Traffic decreased by just 21.5% for failed updates
- Traffic increased by 135.8% for successful blog post updates
In other words, while we lost 504 monthly visitors from failed updates, the successful updates earned 9,327 additional visitors.
If you’re still worried about ruining your posts by updating them, follow the Single Grain updating guidelines mentioned above, as these are the results they produced for us.
61.9% of Results Came From 11.9% of Posts
If you do enough content updating, you’ll begin to see a pattern:
Some updated posts perform well and others perform poorly, but a handful of posts will perform outstandingly well and produce the majority of your results.
These posts, which we’ll call unicorn posts, both doubled their traffic and achieved over 300 monthly visits six months post-update. Of the 42 posts Single Grain updated, five of them qualified as unicorn posts.
I also noticed that these five unicorn posts alone were responsible for 61.9% of the overall traffic increase.
We’ll dive deeper into what makes a wildly successful update below.
Only Update Posts That Have Traction
Similar to writing a blog post, much of the success of an update is determined before you ever begin updating it. If you choose the wrong blog post to update, you may not see any results.
For example, our results showed that despite using the same updating method on each post, posts that already had traction performed significantly better after the update than posts with little to no traffic before the update.
In fact, 45.2% of the posts we updated had less than 20 visits per month pre-update. After the update, these posts only increased the total sum of traffic by 14.8%. In other words, 85.2% of the traffic increase came from posts that already had over 20 monthly visits.
Therefore, my advice is to only update posts that previously performed well and are now steadily decaying.
Unfortunately, you’ll likely find that the majority of your posts receive very few monthly visits. This is particularly true for younger websites that are still building authority. So what should you do with them?
The best solution is usually to combine several related posts, and then 301 redirect the URLs to the new combined page.
We did this to create our huge guide, The Ultimate Guide to Growing Your Business with Influencer Marketing. It currently consists of 10 chapters, each of which was previously an individual post:
While the 10 posts performed poorly independently…
The new updated and combined post generated roughly 4,600 monthly visits just 3 months after being consolidated.
By combining and pruning pieces that performed poorly, you’ll cut the dead weight from your blog, and your average post engagement will increase. This is not only useful to the visitor, but it also makes Google look more favorably upon your website.
The Percentage of Text Changed Doesn’t Affect Performance
As I was updating posts, I always wondered if I was changing too much of the text. For example, because we update all posts to include new examples, most list posts consist of almost entirely new copy. Therefore, 90% of the post’s text might be new.
So would Google look at a heavily updated post as an entirely new post and drop its rankings?
To find this data, I used the Wayback Machine, a digital archive of the World Wide Web, to compare the before and after text from the updated posts. Unfortunately, only 33 of the 42 posts had pre-update screenshots, so that’s the data size I used.
As you can see from the data screenshot, there was a wide range of identical text matches, from just 0.8% to 72% identical text matches:
Note: A 1% identical text match means that just 1% of the text from the pre-update version exists in the post-update version. Similarly, a 72% identical text match means that 72% of text from the original post exists verbatim in the updated version.
Ultimately, there was no correlation between the percentage of text changed and the success or failure of a post.
Despite a wide range of identical text percentages (0.8% to 72%), the average identical text match was similar between the successful and unsuccessful posts:
Therefore, while this data shows that Google probably doesn’t mind if you change a lot of the text, there is a caveat: If you’re updating old examples with new examples on a similar subject, Google doesn’t mind how much text you change.
This proved to be accurate with one of our unicorn pieces that jumped from 296 visitors to 495 visitors after being updated.
The piece is titled 11 Companies That Are Doing Mobile Advertising Right, so to update it, all 11 companies were changed. As a result, the post is almost unrecognizable with just a 0.8% identical text match and 11 completely new companies that do mobile advertising well:
In the above example, while the text is changed and the examples are changed, Google sees that it still serves the same search intent. Even though the examples are different, they are still examples of companies doing mobile advertising well.
However, if you change your post’s meaning so that it serves a different search intent, you might risk losing traffic.
For example, let’s say your original post is titled “Statistics on Instagram Engagement,” and the post is essentially a bullet list of Instagram engagement statistics.
Now that you know Google doesn’t mind your changing much of the text, you may be tempted to structure it in more of a how-to style on Instagram engagement with some statistics sprinkled in.
Unfortunately, this would likely cause your traffic to tank because you changed the search intent of your post. The original search intent served people who want a quick statistic on Instagram engagement. But the updated piece would serve an entirely new audience that wants to learn more about increasing their Instagram engagement.
Therefore, focus more on how the text will change your post’s meaning rather than how much of the text is identical.
Learn More:
* How to Understand Searcher Intent and Use It to Boost SEO Rankings
* What Is Content Decay and How It Affects Your SEO
* What You Should Do With Old Content That’s Not Getting Traffic
* What’s the Right Content for Each Stage of the Marketing Funnel?
List Posts Perform Best After Updates
Another interesting piece of data that emerged from this study is that list posts performed significantly better than ultimate guides, how-to posts, and standard blog posts (for example, What’s the Right Content for Each Stage of the Marketing Funnel?).
Although only 26% of the posts updated were lists, 84.20% percent of total traffic gained came from list posts.
Additionally, four of the five unicorn posts were also list posts.
I assume list posts perform well because they are among the easiest to expand while maintaining the same search intent. For example, one of the unicorn posts, 42 Digital Marketing Trends You Can’t Ignore in 2023, originally started as “11 Digital Marketing Trends”.
When we expanded it with an additional 31 marketing trends, the traffic skyrocketed from 115 monthly visits to 3,640 monthly visits just six months post-update:
However, you don’t necessarily have to add more examples to make the post perform better.
For instance, one of our other unicorn posts, 22 Brands with the Best Content Marketing Campaigns, jumped from 258 to 878 monthly visits post-update. Instead of adding more examples, we just swapped out the examples for fresher ones and made each explanation a little more in-depth and actionable.
Further Experimentation
Although this data is certainly a great start, I still think there is room for more experimentation. Here’s what I found.
Does Adding More Keywords Increase Traffic/Rankings?
One idea I wanted to create a statistic around is the correlation of keywords and traffic/rank increase. Unfortunately, I couldn’t pull up historical keyword data, so I’ll have to use an example instead.
My prediction is that adding keywords to the text won’t really help increase rankings.
For example, this post ranks number one (well, technically zero) for the keyword “video editing guide”:
However, if you go to the post, you’ll notice that doing a control + F search, the keyword “video editing guide” is nowhere to be found on the post:
In fact, when I looked up the post in Ahrefs, only one of the keywords it ranks in position one for exists verbatim in the blog post:
Therefore, my prediction confirmed that adding these broader keywords won’t boost your post’s performance.
Instead, Google has become savvy at distinguishing the meaning of a post. Therefore, Google will rank the best article on video editing for all related keywords, including “guide,” “tips,” “how-to,” etc.
Learn More:
* SEO Keyword Research Made Easy in 2023
* Why You Should Use Long-Tail Keywords in Your SEO Campaign
* 12 Types of Interactive Content to Drive Better Engagement
* How to Create a High-Performance Content Marketing Strategy in 2023
Does Adding a Section Help a Post Rank for Long-Tail Keywords?
Additionally, we often talk about adding sections to a blog post to make it rank for more long-tail keywords and increase traffic. While I would have liked to run an experiment where I add a section to a piece, wait six months, and then see if it ranks for related long-tail keywords, I didn’t want to wait six months to write this.
So instead, I’ll share the insights I found in several examples.
I assume that adding sections will likely not make your post rank in positions 1-10 for competitive (Ahrefs KD 25 plus) long-tail keywords. However, adding sections will probably make your piece more competitive for main keywords (such as “video editing guide”).
For example, this same post on video editing has a section on color correction:
Though it does appear that adding this section gained the post an extra two keywords, these keywords rank in such low positions that they don’t really increase the traffic from people looking for color editing:
It appears that Google still maintains its stance of serving the user the most comprehensive guide to the keyword searched. Therefore, my prediction is that Google pits the one section you added to your post against other complete posts on color editing.
For example, it might look like this:
In this scenario, it’s unlikely that your single section is more in-depth than the competitor’s entire blog post on color editing. And that’s fine. Your target user (a beginner or top of funnel person looking for general video editing advice) doesn’t want to sift through a massive guide that’s overwhelming and too in-depth.
Just don’t expect that by adding a section on color editing, your video editing guide will now also rank in positions 1-10 for color editing terms. In other words…
Adding sections likely won’t increase traffic for long-tail keywords, but adding sections will make your post rank higher for your main keywords.
While it may sound like I’m advocating against adding sections (like video color editing), the opposite is actually true.
You shouldn’t expect to suddenly rank for long-tail terms like “video color editing” by adding that section, but you will rank higher for your main terms like “video editing guide.” This is because Google sees that your piece is more comprehensive than the other video editing guides that don’t include a section on video color editing.
The only exception to this rule is if you add long-tail keywords with extremely low difficulty scores or if you answer featured snippet questions.
For instance, the question “how fast do you lose weight on the paleo diet” is a featured snippet question with low keyword difficulty:
This question requires a concise answer, and Google will therefore create a featured snippet and give it to the post that answers the question in the briefest and clearest way:
This is an easy way to win a long-tail keyword almost overnight.
Again, while I’d like to measure this with data, I simply can’t add featured snippet questions to all of our posts, wait six months and then show you how the traffic changed. 😁
SERP Analysis Still Matters
Another factor that I wasn’t able to pin a statistic to is SERP analysis. This is perhaps the most important factor when choosing to update a post, as rankings are ultimately a competition between blog posts.
Therefore, if you notice that your updates aren’t increasing your traffic, try Googling the post’s main keyword and compare your updated post to the top-ranking posts. Specifically, you should look at:
- The domain authority
- The post’s depth
- The search intent
- The backlink profile
If you notice that your post falls short in any of these categories, it might not perform well. In fact, you should do this exercise before every post you write.
Of the four categories listed above, post depth is the only one you can instantly change through content refreshing. Therefore, if you find that your DA, backlink profile and search intent match the top-ranking posts in Google’s SERPs, you can probably add more depth to the post and make it rank.
Unfortunately, the other three categories are a little more difficult to change.
For example, I updated a post called Essential WordPress SEO Plugins: The Marketer’s Toolkit. The updated piece was actionable, had a great user experience, and totaled over 2,500 words.
Unfortunately, it performed abysmally.
It had just one monthly visitor before the update, and the highest number of monthly visitors it ever received post-update was only four:
So if it’s a quality post with a decent word count and lives on a strong website, why did it perform poorly?
I decided to dive into the top two positions:
When I plugged them into Ahrefs, I noticed that both of the top-ranking posts had a much stronger backlink profile than ours:
I assume that regardless of how well this post is updated, it won’t gain any traction unless it earns some backlinks. By missing the mark on just one of the four categories, the entire post failed.
In this case, I probably wouldn’t combine or prune the post, since it does meet the quality standards and domain authorities of the two top-ranking posts. Instead, I’d promote it heavily and give it some time to earn backlinks.
This is another reason why it’s essential only to update posts that already have some traction.
Your Next Steps
I created this data-driven post to understand not just that content updating is effective, but why some pieces perform well after an update while others fail.
This post should guide you in the right direction, but you should still monitor your own updates and look for patterns in what’s working and what isn’t.
If you’ve tried updating a piece of content and it didn’t work out, don’t be discouraged. These statistics should show you that even though not every piece will perform well after a refresh, some of them will perform exponentially better and make up for those that did not.
