â€œAnalyze your competitors and do betterâ€ is a well-known SEO principle. The study of similar sites of the same subject allows you to analyze trends in a niche, identify features and assess the potential of the project.
In SEO, the analysis of competitors in the search engine results is widely used. At the beginning of the promotion, it helps to select the most successful of comparable competitors in a basic, enlarged manner and assess the traffic potential, and later on, in detail, cluster-by-cluster analyze both individual landing pages and general elements of projects.
In the first case, the general parameters of competing projects are collected: the age of the site, the number of external links, the range (number of pages) and its presentation (listings, tables, pages of the landing structure), etc.
In the second, common elements are searched for among the most successful competitors pages and the possibility and necessity of placing these elements on the pages of the promoted project is considered.
The specialist considers individual correlations – links where the influence of individual factors manifests itself only as a tendency (on average) in the mass observation of factual data – and their strength. On the basis of these links, hypotheses are built, which are further tested and implemented on the site.
An example of analyzing landing pages by a cluster of search queries
We take a ready-made cluster of semantics, for example: Extermination of cockroaches
Using the service http://coolakov.ru/tools/most_promoted/ or any clusterizer that displays the list of URLs in the TOP, we identify the list of competing pages:
Getting rid of Yandex services (they do not fit in the format of the site and the format of the pages), we get a list of the most successful competitors in this cluster. The numbers next to the URL are the number of cluster requests for which this URL is in the Yandex TOP.
We take everyone who has more than 40 queries out of 74 in the TOP (those who cover more than half of the queries we are interested in):
- https: //ses-moskva.rf/disinsection/tarakanov
Next, we analyze the group of factors of interest to us.
You can evaluate any group of factors: cross-cutting elements, detailed on-page optimization, link optimization, approach to the formation of meta-tags. You can also assess commercial factors in the context of the entire site: the presence of mandatory pages in the subject matter, for example, Warranties, Certificates and licenses, Employees, Examples of work, etc.
Let s take an example of on-page optimization of landing page data:
The number of blocks and their composition will depend on the topic of the site.
From the above table, you can conclude about the priority of the presence of certain page elements and arrange possible edits in order of importance:
- Price table.
- Block of benefits.
- Questions and answers.
- Our clients.
The item “Questions and Answers” was ranked above the block “Our Clients” due to the ability to cover LSI-phrases (phrases that set the subject) in the text of this block, while the block “Our Clients” often contains only images (company logos) …
This example shows the main blocks without a detailed immersion in the topic. With a detailed examination, you can pay attention to such issues as:
- the composition of the price table, the availability of prices not only for apartments, but also for cottages;
- the presence of not only text, but also video reviews;
- the ability to increase certificates on click (lightbox);
- the presence of a real, atypical USP in the block of advantages;
Disclaimer: above are deliberately examples of using a free service. In fact, it doesn t matter which particular service you use. The main thing is to understand the principles of work and the scheme for forming the final results.
Both types of analysis are applicable when developing a new site:
- General analysis is applicable to determine the structure of the site and the potential of the project.
- A detailed analysis of clusters also helps at the MVP design stage to create prototypes of landing pages, identify common, cross-cutting blocks and necessary commercial pages of the site.
However, like any method of statistical evaluation, such an analysis has features that must be taken into account in order to avoid â€œblind copying of all decisions that competitors have in the TOPâ€.
Limitations of Competitive Correlation Analysis
1. Correlation and causation
The apparent simplicity of the correlation analysis pushes to make a false intuitive conclusion about the presence of a causal relationship between the factors under consideration and the site s position in the search results.
In fact, there may be a correlation, but there may be no correlation.
Also, the correlation of two values ??may indicate the presence of a common cause for them, although the values ??themselves are not interrelated.
Example: Most of the competitors in the TOP may have a primary key in the domain, however, the success of these sites in SEO-promotion may well be due to the age of these sites.
Simple logic: in competitive topics, most of the domains with direct inclusion of the main key have been sold out long ago, therefore, the age of these domains will be impressive and can have a strong impact on their presence in the TOP, and the very presence of the keyword is just a “nice bonus”. At the same time, in less competitive topics, the opposite picture can be observed: the presence of keywords in the domain will play a greater role than the age of the domain.
2. Weak correlation coefficients, negative correlation, the presence of correlations with a difference in intents
Availability weak correlation with a presence in the TOP-10 (if we look not only at the TOP-10, but consider a factor within the TOP-50), it can indicate that, say, in the TOP-30, the factor is already a generally accepted practice. But this does not at all mean that the elaboration of this factor is optional or not prioritized.
Availability negative correlationon the contrary, it can speak of the â€œconservatismâ€ of the topic, indicating that the factor, which has long been worked out in most topics, has not yet been introduced everywhere. This, in turn, also does not indicate a low priority of the study of this factor.
The difference between intents can severely limit the standard calculation of the correlation coefficient. If we take a request with an implicit intent, where 50% of the TOP will be informational and 50% commercial, this is not a reason to consider the presence of a button â€œbuy in 1 clickâ€ with a positive coefficient of 0.5. Here you will either have to count within 5 commercial competitors, or get commercial competitors up to 10 units from the second issue page.
How can you benefit from looking at correlations in SEO?
Conducting a competitive analysis will be able to quickly show which elements of sites / pages have been worked out by competitors and are missing on the promoted project.
At the same time, analyzing different clusters and different TOPs, we will see that the number of competitors with an embedded element and the number of competitors without it will differ.
Based on the indicator of the strength of the correlation, it is possible to basicly prioritize the items under consideration in terms of importance and urgency for introduction. Naturally, these priorities will not be final – you always need to separate end-to-end edits and page-by-page edits, and finally streamline the introduction of certain edits, taking into account the complexity of the implementation of one or another element.
At the same time, you need to understand that the correlation link itself is not the ultimate truth.
How to live on? Experiment!
Despite the previous statement and all the limitations described in the article, correlation analysis is an important tool in the work of an SEO specialist.
It allows you to quickly identify deviations of projects “in the TOP / not in the TOP”, make a list of unprocessed factors and set a priority for each of the points.
How to work with him further?
- Supplement prioritization with cost / speed of implementation for each item. It can be hours of specialists, the cost of implementation in rubles, a conventional unit of complexity from 1 to 10. Rank points by the sum of these two columns.
- Supplement the prioritization with data from the checklist. In the case when there is a checklist with most of the possible edits for the project, an excellent solution would be to supplement the list with one more column – “priority by checklist” – and rank edits taking into account it.
- The priorities are clear! However, we still do not have guarantees that a specific change will work. We set up an experiment: we make edits to the site, measure the result.
In the case when the widespread implementation of paged edits with blocks differing in content looks time-consuming for an experiment, you can take several control sections / clusters, implement and review the result, and only if there is a pronounced influence, â€œroll outâ€ the edits to the entire site.