Ever wondered how using AdSense affects page performance or SEO? Or you maybe trust Google blindly to deliver optimized content that does not hurt your site’s performance? This post will not only give you an empirically proven answer but will also show you how to keep an eye on the behavior of external scripts to maintain your site as optimized as possible.
AdSense Basics
AdSense is one of the outputs for Google AdWords advertisers. This means that by creating an AdSense account and adding a property, you join Google’s partner (display) network (this is different than Google Partners for AdWords). This network is exclusively used to show ads, relevant to a site’s content and site visitors’ interests (revealed by cookies).
AdSense is therefore a service, allowing publishers to generate income by showing commercials on their sites. Even though there are strict rules for ads and a thorough approval procedure, it’s the advertiser that provides content, is responsible for its optimization, and has the possibility to modify or add additional tracking tags to URLs. It’s therefore wrong to assume that it’s Google’s responsibility to deliver optimized ad content.
AdSense can easily be added to page code without plugins by manually adding it to Text widgets (for arbitrary text or HTML) or pasting it directly into the page’s content (this is done in Text mode). In some cases, where the structure of your site or the functionality of your theme doesn’t allow for this, you can still choose to install AdSense or Ads plugin, giving you additional flexibility and speed of execution.
Until 2017, Google did also provide such a plugin but has now deprecated it, as they don’t deem it necessary to use a plugin. Instead, focus is moved on offering maximum functionality and flexibility of ad setup in AdSense accounts.
When adding a single ad unit to your site, you have to copy-paste a piece of JavaScript, which you can choose to make asynchronous. This piece of code works by pulling a number of external resources, such as images, text, tracking, etc. in order to deliver the full ad.
AdSense And Page Performance: The Experiment
In order to assess the way AdSense possibly affects page performance, the following experiment was carried out: two identical (in terms of structure, text length, and page elements, such as sidebars or number and size of images) post pages on WPBloggingNerd were a part of the test. On one of the pages, three AdSense adds were added into the sidebar. The other page remained unchanged. The only difference between the two pages at that moment were the 3 ad units.
Afterwards, page performance tests were run on GTMetrix. Firstly, for the individual pages and then a comparison test for the two pages. At last, an additional test was performed on the page with 3 ads after the removal of two of them, so that only one ad was left, in order to understand the precise implications of using AdSense.
Limitations
This experiment has certain limitation, related to the fact that AdSense ads can differ in terms of optimization, depending on the advertiser. This is, however, not something you can control.
Additionally, page performance is dependent on different factors, such as server performance. While it was not possible to eliminate such factors, they were minimized by carrying out the experiment on the same server, on the same website at almost the same time, with only minutes between the tests, in an effort to achieve precision. The tests were run a few times per page to make sure the results are consistent.
Last but not least, the tests were run on identical pages (same template, same sidebars and widgets, same formatting, and similar text size) and not on one and the same page, meaning that there might be a very slight difference in performance, caused by the fact that the pages feature different images. This was done in order to run simultaneous tests in real environment, without generating duplicated resources. According to the test results though, there was 1 percent point difference for the image optimization factor between the two pages, which is insignificant in terms of the overall score and cannot affect the key optimization areas that are in focus in connection to the experiment.
Page Performance: No AdSense Ads
The performance of the post page with no AdSense ad was measured as follows:
Page performance report of a post page with no ads by GTMetrix
The performance score shows that the page is well optimized, with only a very few possibilities for further optimization. Pay special attention to Leverage browser caching and Defer parsing of JavaScript. The report shows that JavaScript is effectively deferred and browser cache is almost optimally set. The reason why cache isn’t up at 100 points is that the site uses Google Analytics. The analytics script is set by Google to be cached for 2 hours only, which is a bad caching practice.
Pay also attention to Fully loaded time, Total page size, and Requests at the top of the report. The green arrow to the right of these values means that the page performs better than average.
Page Performance: 3 AdSense Ads
In contrast to the report above, the performance score for an identical page, where the only difference is that 3 AdSense ads were added into the sidebar, looks like this:
Page performance report for a post page with 3 AdSense ads by GTMetrix
Notice that the overall performance (the top line) has decreased dramatically. Especially alarming is that the values for page load time, total size of the page, and number of requests are much higher than the corresponding values for the page with no ads. Remember that the only difference from the page, analyzed above, are the 3 AdSense ads.
It’s important to note that the report reveals that all of the values in the picture are almost exclusively affected by AdSense. To exemplify this, we’ve expanded one of the sections, showing which JavaScript resources are render-blocking. As you can see, they are all ad-related. Even though there are only 3 ads on the page, the actual number of render-blocking JavaScript is higher than that.
Comparison: AdSense vs. No AdSense
To clearly document the difference between the performance of a page, featuring 3 ads, and a page with no ads, a comparison test was carried out. Here you can see the results, accompanied with a graphic of the two pages.
Here are shown the comparison test’s results of a page with 3 (to the left) and without (to the right) AdSense ads.
As a result of adding 3 AdSense ads to the page, the performance grade has decreased with 29-34% percent, the size of the page has been doubled, and the number of requests has got 4 times higher. The Fully loaded time is also monstrous for the page with ads.
If we look at the results in detail, it seems that it’s much more than render-blocking JavaScript that gets affected by adding AdSense ads to a page.
The report to the left shows performance stats for the page with 3 AdSense ads. The report to the right shows the stats for the page without any ads.
Especially in terms of avoiding redirects, serving resources from a consistent URL, deferring JavaScript, leveraging cache, and minimizing request size, Google AdSense performs poorly and hurts page performance considerably.
All of these stats affect not only page performance but also SEO. However, the most serious problem in terms of SEO and user experience seems to be the horrible increase in page load time. Most people bounce, as long as a page doesn’t load in 3s. The chances that site visitors will wait more than 20s for a page to load are almost not existing.
Note! Previous tests on the same website have established that ads in the sidebar perform better than ads, placed directly in the text. Thus, the above tests are made on a page with optimized ad positioning.
Page Performance: 1 AdSense Ad
Even though the test above clearly shows that displaying 3AdSense ads affects page performance seriously, the questions is: Is it just as bad to display 1 ad per page?
Here’s how the report looked for the page with 3 AdSense ads, after removing 2 of them:
Page performance report for a post page with 1 AdSense ad by GTMetrix
As GTMetrix starts the report with the lowest scores, you can see that almost all of the critically low scores from the report above are gone. The only 3 areas, affected by AdSense, are serving scaled images, leveraging browser cache, and deferring JavaScript. Note that there are 4 ad-related render-blocking JavaScript resources, even though the page features a single ad.
Even though, as noted in the paragraph on limitations, some ads might be less harmful due to better optimization on the part of the advertiser, it does seem consistent that you lose about 3-10 percent point of page performance grade by adding a single AdSense ad. Each further ad costs you additionally around 10%.
Monitoring External Scripts
As external scripts aren’t hosted on your website, they cannot be optimized by you. Therefore, it’s crucial to monitor for such scripts and evaluate the way they affect page performance vs. their benefits.
Most often, such scripts are related to external services, as analytics software, tag manager, advertising, affiliate marketing, and streaming. However, sometimes they can be added to your code by a plugin that uses external resources.
It is therefore a good idea to monitor your site regularly and especially after adding, removing or altering plugins, services or code. You can easily monitor your site by carrying out performance tests with GTMetrix, Pingdom Speed Test or Google PageSpeed Insights.
As you won’t get a direct answer to which service or plugin might be causing troubles, it’s important to continuously carry out tests and react to changes. The tests will provide you with a list of problematic resources that often can be attributed to a specific service by reading their URLs. In rare cases, the URLs won’t reveal information about the service they are connected to. In such cases, you should follow your steps in revers by deactivating/removing recent changes to identify the source of the script.
Once you’ve identified the service, you should perform an analysis of its advantages and disadvantages. The rule is that problematic scripts should not pile up, as this would destroy your site’s page performance.
