Elena Terenteva

11 Most Common On-Site SEO Mistakes - SEMrush Study

If you want to make sure your webpages rank well in search engines, it's essential to have a basic knowledge of search engine optimization (SEO). But the truth is, many people don't. SEMrush’s research shows that website owners often struggle with technical SEO. That's no surprise, because the average SEO checklist contains more than a dozen issues to address before a site can truly succeed.

But which SEO issues do you really need to focus on?

We decided to find out using real data. We collected anonymous data on 100,000 websites and 450 million pages using SEMrush’s Site Audit tool to determine:

  • top SEO issues

  • how many sites these issues affect.

In this article you will find a list of the most common on-page, technical SEO and website issues and information about the way they can affect your search engine rankings. The infographic below summarizes our key findings, but you’ll learn even more if you read on.

1. Duplicate Content

Google defines duplicate content as "substantive blocks of content within or across domains that either completely match other content or are appreciably similar."

According to our research, the most common SEO issue affecting websites is duplicate content, which we found on 50 percent of the sites we analyzed. During a recent Google Q&A session, Andrey Lipattsev, Search Quality Senior Strategist at Google, stated that there is no such thing as a duplicate content penalty. But that doesn’t mean it’s something you can just turn a blind eye to.

First of all, by having duplicate content on your website, you will lose your capability to choose which page you want to rank for. Search engines won’t know which pages you want to be considered landing pages in SERPs, and those pages can even start to compete with one another. Second, search engines are specifically designed to make the web a better place for users; valuable, unique content is highly appreciated by both search engines and users.

2. Missing Alt Tags and Broken Images

Images are an important part of content marketing, but they can also create major SEO issues. Our research concluded that 45 percent of sites have images with missing alt tags, and another 10 percent have broken internal images. Both of these are problematic.

Let's start with alt tags. They’re important to image search. Although search engines have gotten pretty clever, alt tags help them understand what images are about.

In other words, alt tags provide textual descriptions of images, which makes it possible for search engines to categorize them. That's one reason why your image alt tags should contain your SEO keyword phrases.

Alt tags are also useful for visually impaired people using screen readers, as the readers will use the information in alt tags to describe images to the web users. Search engines are very concerned about user experience, and images without alt tags can hardly be considered a sign that that a website provides value to the user. They can lead to a higher bounce rate and, as a result, they might be one cause of poor search engine performance.

Broken images cause the same issues as broken links, which we will look at in more detail below. Broken links are dead ends for users and search engines, and they can cause search engines to downgrade your website because they create a poor user experience.

3. Title Tag Issues

Search engines use title tags (page titles) to determine what pages are about. Title tags appear at the top of search listings, which means they also help web users decide whether to follow your link. Title tags are also one of the most important SEO elements on your page – correctly optimized title tags can really impact your rankings in a positive way.

At SEMrush, we have seen four major issues related to title tags:

  • Duplicate title tags affect 35 percent of sites.
  • Overly long title tags impact 15 percent of sites.
  • Missing title tags are problematic on 8 percent of sites.
  • We found title tags that are too short on 4 percent of sites.

As mentioned earlier, Google always wants to deliver unique content to users. But missing and duplicate title tags don’t provide users or search engines with relevant information about a page’s content, nor do they communicate that a page provides value.

The length of your title tag matters because it affects how much of your title is visible in search results. According to the latest news, depending on the device that’s being used, Google may show 70-71 characters, so it's a good idea to keep the key information (including your chosen key phrase) within that range.

Most good SEO tools will help you to identify duplicate, long, short or missing title tags. For more help, check out our recent guide to optimizing title tags.

4. Meta Descriptions

Meta descriptions showing up in search results help web users decide whether to visit your site or not. Although meta description doesn't influence page ranking directly, the relevance of meta description influence page CTR, which is important. 

Our research shows that 30 percent of sites have duplicate meta descriptions and 25 percent of sites have no meta descriptions at all.

Our Site Audit tool will help you identify both issues, but it's important to go in and fix them manually, if necessary. Again, if you are using WordPress, the right SEO plugin will help you set up rules to ensure that all your content has meta descriptions.

When you're ready to create new meta descriptions, here are some resources that can help you:

5. Broken Internal and External Links

Broken links can be a big onsite SEO problem. As your site grows and you update resources, one or two broken links might not be a problem. If 404 page is properly set up, they’re not a problem at all, but what if you have hundreds of them?

This makes broken links a potential danger. There are several reasons for this, but, first of all, if a user sees a 404 page instead of the useful information they wanted, it causes a traffic dropdown. Also, users will perceive your site as low quality.

Second, broken links are a waste of your crawl budget. Each time search engine bots visit your website, they crawl a certain number of pages, but not your entire site. If you have a lot of broken links, you risk diverting the bots’ attention from your pages, which actually matters – your pages won’t be crawled and indexed.

Our research showed that 35 percent of sites we crawled have broken internal links that return bad HTTP status codes (70 percent of those return a 4xx - page not found or similar - code).

Twenty-five percent of sites we’ve analyzed had broken external links. In the long run, this issue can reduce the number of pages that will appear in the search engine results and impair page authority too, so it's something you need to fix.

You can always identify broken links via our Site Audit tool or a link checker plugin, and then fix all of them. You can also find broken inbound links and approach webmasters to suggest a new resource on your site they can link to.

For more help with crawl errors and broken links, check out 5 Steps to Get Your Website Crawled Faster and ultimate list of crawlability mistakes - 18 Reasons Your Website is Crawler-Unfriendly: Guide to Crawlability Issues.

6. Low Text-to-HTML Ratio

We saw the warning low text-to-HTML ratio on 28 percent of the sites we analyzed. This means that these sites contain proportionally more back-end HTML code than text that people can read. We suggest an acceptable lower limit beginning from 20 percent. Often, this warning is a sign of other SEO ranking issues you need to address. For example, a low text-to-HTML ratio can be a sign of:

  • A poorly coded website (with invalid code and excessive Javascript, Flash and inline styling)
  • Hidden text, which is something spammers do, so it's a red flag for search engines
  • A slow site – the more code and script pages contain, the slower they’ll load, and page load is an important SEO factor.

Fix this issue by checking any pages on which this warning appears and:

  • Removing unnecessary code to reduce page size and increase speed
  • Moving inline scripts and styles into separate files
  • Adding relevant on-page text where necessary.

7. H1 Tag Issues

Historically, header tags have been a hugely important part of SEO, indicating the most important content on the page. There should be just one H1 tag on any page, often for the title of the content. Even though the use of HTML5 has changed how header tags are used (you can now have more than one H1 on a page), header tags still create a useful hierarchy for both search engines and web users.

Our research showed that 20 percent of sites we analyzed had multiple H1 tags, 20 percent were missing H1 tags, and 15 percent had duplicate information in their title tag and H1.

So, it's important to know the difference between title tags and header tags. The information contained in your title tag appears in search results; header tags are what your reader sees on your page. 

As we mentioned, you can have multiple H1 tags on a page, but only if you are using the right HTML5 markup to distinguish between sections of equal importance.

This tutorial from Tuts+ provides more information. In all other cases, avoid overusing H1s . And ensure that your H1s are consistent with your title tags (for the purpose of SEO relevance) without being identical.

8. Low Word Count

Eighteen percent of the websites we crawled had a low word count on some pages. Word count is a complicated SEO metric, because on one hand, there's no minimum word count for a page. On the other, Google is known to rank content with more depth higher, and longer content is one indicator of depth, especially if you avoid fluff.

In onsite issue number six, we mentioned the importance of including relevant on-page text wherever possible. You should do anything you can to give your content depth and make it more valuable for readers. Think about it: when you see an infographic, don't you appreciate it when its creator has taken the extra time to provide more context? We do, and so will your readers.

9. Too Many On-page Links

Linking is hard to get right. That's probably why 15 percent of the sites we researched had too many on-page links on some pages. While Google dropped their requirement to keep the number of links on a page under a certain number a while ago, good SEO means having a natural link profile that includes relevant and high-quality links.

Too many links can dilute the value of your page and send most of your traffic away. But if those links are relevant and useful, then your site will still rank well.

To solve an on-page link issue, perform a link audit and make sure that all links on the page in question really add value. Otherwise, get rid of the ones that don’t to improve your SEO and provide a better user experience (which is also good for SEO).

10. Incorrect Language Declaration

On the web, our audience is global. That's why it's important to include a language declaration on the page to declare the default language of the text in the page. In our research, 12 percent of websites got this wrong. A language declaration is an important tool that:

  • Informs browsers of the content’s language (which useful for translation and page display)
  • Ensures those using text-to-speech converters hear your content read in the correct dialect of their native language (for example Castilian or Latin American Spanish)
  • Helps with geolocation and international SEO.

Use this list of language codes to ensure that you get language declaration right. Google will use language declaration to ensure that the right people see the right content. While this may not directly affect your SERPs, remember that it will help improve your pages’ relevance score, which is an important part of SEO.

11. Temporary Redirects

Redirection is an excellent way to let search engines know when a page has moved, so that you don't lose any page authority. However, there's a big difference between permanent (301) and temporary (302) redirects in terms of SEO. Our analysis showed that 10 percent of sites we looked at contained temporary redirects.

A 302 redirect can cause search engines to continue to index an outdated page while ignoring the page you are redirecting it to. That's why if the change is permanent, it's worth implementing a permanent redirect. According to Moz, that's the best approach.

Of course, in the end, Google may recognize that a 302 redirect is permanent and make it into a 301, but to avoid poor optimization and loss of traffic, it's better if you take control of the process.


These 11 common issues are hurting many sites' optimization efforts. Our research shows that a lot of websites have very serious on-page SEO issues. Yes, not all of them impact your rankings equally. However, bad SEO is not always about a direct violation of search engine rules, it sometimes results from simply not taking proper care of your website and your users.

Yes, manual temporary redirects may not affect your rankings as severely as other on-page issues, and, yes, of course, there is no such thing as a duplicate content penalty or a title tag issue penalty. But, you should definitely be aware of such issues, because they affect your traffic and your user credibility, and, as a result, lead to profit losses. And there is always the chance you could get to a critical point where a lot of small mistakes can accumulate and snowball into a litany problems that can destroy your website’s rankings in SERPs.

Are you fully aware of all of your website’s issues? Forewarned is forearmed. We hope you find this data valuable and helpful on your way to the top of the SERPs!

The article is extremely helpful in understanding the details of SEO and has been explained in a lucid easy to understand manner.
Hi Elena Terenteva! Let me ask you an advice: I don't have programming background, I'm only at the basic level. I know what alt tags are, but I don't know how to redirect a page! Is a SEO analyst responsible to redirect a page, or he can ask a programming professional for help? At the begining of your career, how did you learn how to redirect a page?
Thank you for this awesome article!
The title should be big and easy to read. This is more important than ever. (Many people will first encounter your cover on a screen, not on a shelf.) This is such a well-worn cliche of cover design that I have a designer friend with a Facebook photo album called “Make the Title Bigger.”
Don’t forget to review a thumbnail image of the cover. Is the cover compelling at a small size? More people are buying books on a Kindle or mobile device, so you want the cover to read clearly no matter where it appears. You should also anticipate what the cover looks like in grayscale.
Do not use any of the following fonts (anywhere!): Comic Sans or Papyrus. These fonts are only acceptable if you are writing a humor book, or intentionally attempting to create a design that publishing professionals will laugh at.
No font explosions! (And avoid special styling.) Usually a cover should not use more than 2 fonts. Avoid the temptation to put words in caps, italics caps, outlined caps, etc. Do not “shape” the type either.
Do not use your own artwork, or your children’s artwork, on the cover. There are a few rare exceptions to this, but let’s assume you are NOT one of them. It’s almost always a terrible idea.
Do not use cheap clip art on your cover. I’m talking about the stuff that comes free with Microsoft Word or other cheap layout programs. Quality stock photography is OK. (iStockPhoto is one reliable source for quality images.)
Do not stick an image inside a box on the cover. I call this the “T-shirt” design. It looks extremely amateurish.
Avoid gradients. It’s especially game-over if you have a cover with a rainbow gradient.
Avoid garish color combinations. Sometimes such covers are meant to catch people’s attention. Usually, it just makes your book look freakish.
Finally: Don’t design your own cover. The only people who should consider designing their own covers are professional g
This is a great blog, however, needed some examples on lang declaration, word count,text to html ratio
Great article. Short, insightful and informative. Knew about some of these, but appreciate the advice on more technical issues. Thanks!
Great research. Although I have a bone to pick with SEMRush. I have a site with articles that are between 1K-2K words long and they have the same percentage of text to html ratio than other sites (that I also manage) with shorter articles (600 words), same theme and even more plugins. Do images count as content? (They should and in this case, the first site also wins with 25+ images per post vs 4+ images per post on the second.
Havi Goffan
Hi Havi! Thank you for your comment. First of all, you gave us the idea to write a descriptive material about low text to HTML problem. Will do! Second, I can't tell you exactly what the problem is without taking a look on your project. Can you please write me on twitter - I'll send you my email and we can solve it together. Thank you!
Good article
I need help about how to solved low content and what is the minimum requirements of content website?
Hi, good article,
However I can't help but think there are some issues that proponents of the "don't focus on ranking #1" crowd keep missing.
1) Given a system where the public conducts a search query that results in a list of links pointing to pages on sites that are ranked according to relevancy, it's only natural that the searchers focus on those top search results. (it is an organic search after all)
2) Given that, "the top listing in Google's organic search results receives 33% of the traffic, compared to 18% for the second position, and the traffic only degrades from there." http://searchenginewatch.com/a...
3) Given that the "I just want people to know that if the only goal and KPI is rankings, then they’re setting themselves up for failure" can be said about any other metric (be it bounce, CTR, conversion, etc.) however we never see same argument for not focusing on those other KPIs.
4) Something that nobody else seems to cover, which is opportunity cost with respect to organic traffic market share. If we are not in that #1 slot, then our competition will be. Not a pleasant thought, especially for ecommerce.
It just seems like a strawman to keep bringing up the "don't focus on #1 spot" argument. Yes, if it were THE only metric, sure, but with all of these SEO sites, articles, posts, blogs, etc., I'm not seeing that as THE only focus. A major focus, yes, but not the only focus. Do we obsess? Yes. But we obsess on all of the metrics . Heck, I've spent days focusing on just title, H1 and meta description alone and now we have Social (all umpteen) along with LSI (not_provided) to worry about as well.
From what I gather, the trick is to make them all work, for if you let any one of the metricses (metricki?) fall, then you're still not optimizing results... http://www.seoindiarank.com/
Hi Jeffy! Good point - focusing on one metric just couple on-page elements won't bring you to the top. And from this research, we can see that a lot of website owners are missing some on-page elements they consider to be irrelevant or insignificant. And it's one of the ideas we want to share - keep the balance and conduct regular audit of the site, to be sure that all wheels are turning and your website is running like a car, not a bicycle.
50% is duplicate content issues - this is cos semrush includes pages like your blog home page, which will have a summery of maybe the last ten blog posts. And the blog categories, again contain summeries of the posts in the categories. Same with tag categories too. Does this duplicate content have an affect of SEO, and is there EASY way to fix this??
I would like to throw in image size as another issue for websites. Lots of sites have images that are far too big which slow down a page's load time. Great post though!
Greg Kristan
Thank you for your comment!
thanks for the article! Will check my site. the infographic is amazing!
Sergei Eremeev
Thank, Sergei! This was our main goal - to point out that SEO health is important. Seems like we achieved it :)
how to do mobile optimization ?
Dataentry Outsourcing
I guess it's a topic for another big article :)
Dataentry Outsourcing
As a developer myself, mobile optimization comes hand in hand with desktop, but there is particular focus on the USER EXPERIENCE. Leaving out unnecessary elements, such as images that will take up too much estate on a small device and chunking of information helps so users can find what they are looking for quickly. Its all about usability, scalability, efficiency and speed to mention a few.
Hi Guys, Nice infographic. This is what Google has to say about Duplicate Content: "Duplicate content on a site is not grounds for action on that site unless it appears that the intent of the duplicate content is to be deceptive and manipulate search engine results. If your site suffers from duplicate content issues, and you don't follow the advice listed above, we do a good job of choosing a version of the content to show in our search results". So should it be given such a high priority? I always fix it if I can, but does it warrant a lot of time? Interested to hear people's thoughts.
Great article. Thanks.
Great article, thank you. Duplicate content seems a real issue for me, but only with regards to duplicating portfolio categories and tags, so I'm not exactly sure if that is actually an issue that would penalise me or not. Still searching for clarification on this but the Semrush audit says it is bad.
Michael Douglass
Hi, Michael! No, there is no such penalty, as "duplicate content penalty" - http://www.hobo-web.co.uk/dupl..., so don't worry :) But still, it's is an issue. Not the worst one, but bad under certain circumstances.
Elena Terenteva
Hi Elena, great advice, that link cleared things up for me. Very helpful thank you.
Great post guys. Duplicate content issues can arise if you use Wordpress and you are using a post template to display images, and your images have the same name. We are working with that issue right now, 2 or three pages, which are actually images that are part of a portfolio. It's not actual duplicate content, but because they are images that are named pretty much the same thing, Google sees duplicate content.
Joel Black
Thank you, Joel!
You guys have access to such a large amount of data! You should be listing issues and how they have effected rankings. Eg the same groups of keyword in multiple cities with similar search volumes and competitiveness, similar backlink profiles, one has poor alt tags another has them perfect, what's the difference in rankings?

I was expecting to find which of these issues had the biggest effect with real world data as opposed to just what issues are out there with no data backing up what effect this had on rankings. I've used your API before, many times, you guys definitely have the data, and internally it wouldn't cost you guys an arm and a leg to produce a report like it would for me via buying API credits.
Paul Rogers
Paul R - the only tie to rank correlation lies in the fact the the data derives from Google Top 100 results. Elena did a good job stating prevailing on-site tech problems. I agree that more impact would result from a comparison of issues afflicting domains with the most #1 rankings, the highest proportion of #2 & #3 rankings, #4--#10, or perhaps relying on SEMrush rank status. Maybe next time!
Paul Rogers
Thanks for your comment, Paul! I see your point. This kind of research would take much more time and effort, but it certainly would be interesting. I'll consider your comment as the request :) We are definitely would love to work on something like this.
Thanks semrush..

tips terkini
Watchy Jelekz
You are so welcome!
Are these stats from all websites or from websites that experienced a decline? If they are from all websites then maybe they have no negative effect on sites?
These stats are from all kind of websites, so we state that all issues, described here, as the most common, but not most harmful. You right, some of these issues, might not have direct negative impact.
Elena Terenteva
It would be interesting to compare the percentage of these issues in all sites compared to those that declined after updates such as Panda. Its a great analysis but if these issues are present in all sites I'm not really sure what we can learn from that. Maybe they have no impact on rankings at all.
Sounds like a plan for me :) This might be timely, but I'm sure this would be a great research. Thank you for suggestion!
From these common issues you have identified, which would you say has most negative impact on your pages in regards to rankings?
Greg Martin
This is the big missing part of the story. Most of the stuff here is repetitive and basic. I was hoping for something more juicy.
Mart Shark
Thank you for your comments, guys! This is just the first step. We are not going to stop and we thought already about another study about the most harmful issues. Stay tuned!
Thanks for the statistically valid information. I loved the infographic, too, and retained it for reference. GD.
You are welcome! It took us awhile, but it worth every minute of it. Glad you found this data useful!
Great inforgraphic and insights. Thanks +SEMrush!
Jeannie Hill
Thank you for a feedback, Jeannie!
Add a comment