Leave a request
SEO specialists conducting tests and putting forward hypotheses for promotion often encounter unexpected results from experiments. Despite long experience, it is difficult to get used to the fact that familiar SEO practices sometimes turn out to be harmful for organic traffic, and what was done “at random” turns out to be successful for a particular site.
Even the most strange test results at first glance can lead to important conclusions, and this is what makes testing such an important tool in the work of an optimizer.
In this section, we will look at several SEO tests, the results of which were quite unexpected.
Google has introduced a new system for generating titles and meta descriptions for snippets that is dynamic and adapts to changes on pages. However, many site owners noted problems with its operation, in particular, the appearance of meaningless or erroneous texts.
One client decided to test the 'data-nosnippet' attribute to eliminate the use of auto-generated descriptions and force Google to only show those included in the meta data. This decision resulted in a 3% decrease in organic traffic. This test showed that Google actually generates more effective descriptions and titles than those suggested by webmasters.
The experiment was conducted on an e-commerce site, where it was decided to increase the number of links to related articles on blog pages from two to four. The result was interesting: traffic increased by 11%, but what was especially surprising was that the increase was on the donor pages, and not on the pages to which the links lead.
This test showed that links help search engines better interpret page content by linking different parts of a site and improving its visibility, rather than just passing on link juice.
One UK site that entered the US market retained old terms such as "trousers" rather than using the more popular US word "pants". After localizing content to suit American preferences, organic traffic increased by 24%.
This result demonstrated the importance of localizing content and tailoring it to local market characteristics, which significantly increases visibility and attracts more targeted visitors.
On yet another e-commerce site, we checked the addition of prices to title tags. It was expected that this would increase the attractiveness of snippets for users, but the result was negative - traffic decreased by 15%. This may be because competitors weren't using this approach and adding a price attracted less interested visitors.
This experiment showed that in some cases, using prices in headlines may not meet user expectations and reduce click-through rates.
A test was conducted for the site to add alternative text with keywords to product images on category pages. Previously, the images did not have alternative texts, and it was decided to add the same text, for example, "Men's T-shirt" for all images.
Although the hypothesis was to improve page relevance, the result showed no impact on traffic and rankings. However, it has improved the understanding of image content for both search engines and users with content limitations.
These experiments demonstrated how important a testing approach is in SEO, as even small changes can have unexpected effects. To more accurately understand and implement an SEO strategy, it is important to regularly conduct such tests.
You can contact the SEO studio “SEO COMPUTER” with any question by email info@seo.computer.
ID 1182
