Choosing a Text Optimization Strategy for Ecommerce Sites: Generate or Unify

Choosing a text optimization strategy for ecommerce site

Choosing a Text Optimization Strategy for Ecommerce Sites: Generate or Unify

Them Text optimization remains one of the basic criteria for indexing pages and determining their compliance with certain search queries.

Maria Baglikova, a leading specialist in the research promotion department of digital agency Nectarine, explains how to choose the optimal content optimization strategy for a specific site and save resources.


specialist in the SEO department

of the digital agency Nectarine

Electronic commerce is a category of sites selling goods or services. The search engine optimization success of these web resources largely depends on business factors that cannot be influenced by an SEO specialist: assortment, price point and ability to order, competition in a niche, brand awareness, availability of USP, and favorable conditions. for the customers.

Based on this, SEO specialists strive to achieve maximum effect on the aspects that they can influence. In particular, it is about the optimization of the text – the elaboration of the text zones of the pages that the search engines take into account when ranking before the request: title and titles H1 – H6, Meta Description tags, small text fragments, long blocks of text, as well as internal and external link anchors.


Single Text or Template: Compare and Choose an SEO Strategy

An important feature of commercial sites is a large number of similar pages – product sheets and catalogs (listings).

For them, two SEO approaches can be distinguished: a one-off

  • study of the content of each page – we will conditionally call this uniqueness approach;
  • template optimization – setting up the generation of template-based tags and texts, where key features of a particular page act as variable parts.

Pros and Cons of Each Approach

At first glance, it seems that using templates is inferior to preparing unique content. But each approach has its pros and cons – let’s take a look at them.

Unique Text

🟢 Text looks more human and answers questions about a specific product or category – for example, it contains selection tips.

🟢 Hit more precise in the main requirements for occurrences and volume of text.

🔴 Preparing unique texts for a large number of pages takes a long time to write and agree on the texts + is expensive.

🔴 The substitution of quantitative information in these texts (number of products, price range) is a resource-intensive development task.


A single sample text

Text template

🟢 Speedy site-wide content preparation.

🟢 Put in place in a practical way the substitution of quantitative indicators in the texts: the number of products, the price range, the minimum price, etc.

🔴 The text contains less useful information about a specific product or category.

🔴 Less exact match in terms of the number of occurrences and volume of text.


Sample Placeholder Text

Despite the compelling benefits of uniqueness, cost and speed of content preparation play an important role in decision making. For this reason, the use of templates is often a more attractive option.

To maintain a balance between quality and resource efficiency, you need to consider whether the effort of preparing unique content will be more effective for search rankings than generating tags and texts. Therefore, an important part of the job is the analysis of competitors who have already secured positions in the top search results.

Consider the choice of a stepwise strategy.

Step 1. Selection of pages to analyze

We need a semantic kernel with a breakdown of requests by site pages, frequency, positions, and URLs in Yandex and Google search results. This will make it possible to form small samples of priority lists and product sheets to be analyzed and, on this basis, to select an optimization method for pages of the same type.

The main criteria for adding a page to the sample are a high frequency of promoted queries and the presence of positions in the top 30, as these pages have more potential to lead to the top 10 and generate search traffic.

Then choose the search engine for which you need to optimize pages in the first place – Yandex or Google. If one search engine has positions in the top 10, and the other does not, then it is best to focus on which one is behind and try to move up the positions. If there are no top 10 positions in any search engine, it is worth focusing efforts on the search engine, where the positions are on average higher – there we will get an effect most important.

In case of doubt or in case of equality of visibility, in our opinion, it is better to choose Yandex, since this search engine reacts more strongly to text spam and, therefore, requires more attention.

A combined approach is possible: for example when a part of the pages of the site is optimized for one search engine, and others for another.

Step 2. Select Competitors for Comparison

The basic concept of text optimization, followed by many SEO specialists, is “first do as above, then try better”. This is why you should analyze your competition in the top 10.

However, there is one caveat that needs your attention: it is not always possible to compare all the search results of a query with the promoted page. .

There are the following reasons for this:

  • The output can be heterogeneous – it can contain not only business results but also news, multimedia, and the like. This is typical for e-commerce topics with low product knowledge, where users may need questions answered rather than buying them.
  • There are not enough quality competitors on demand. This situation occurs in subjects with low competition, such as low profile sites and monopolies. The result can be poor quality if search engines don’t have enough statistics to understand which competitors to show.
  • The query can be interpreted as a mark for another site, so it gets a ranking advantage and even takes multiple positions in the search results at once, but that has nothing to do with text optimization.

If none of these signs are present, feel free to move on to the next steps.

In another case, you need to identify the competitors for comparison. Here’s how to do it.

  • Find competitors with high positions in the target semantics. To do this, you need the top 10 search results analyzers – for example, Key Collector or free Arsenkin Tools – or services for selecting competitors by keywords and position suppressors – for example, Rush Analytics.
  • Identify the strengths and weaknesses of your site and your competitors: the size of the assortment, the number of product categories, the availability of multi-brand products, the popularity of the company (brand application), the number of office regions, ease of use of the site, etc. Therefore, you should only leave the competitors closest to the promoted site for these metrics. For example, it makes no sense to compare to giant aggregators, sites with a large Type-in ​​- direct and browsing traffic – due to the high demand for the brand, if there is players of your level on the first page of search results – you can work hard to get them.

It is advisable to get a list of 5-7 competitors at the exit – a kind of whitelist, which will be useful later. However, the more suitable the competitors, the more accurate the results of the more in-depth analysis will be.

Step 3. Choose an Optimization Strategy

At this point, we turn to consider the competitors and determine the method they use to craft the tags and texts – the uniqueness of each page or the preparation of templates. .

Here you need a text analyzer – a service that checks the number of keyphrase mentions on the promoted page and among the competitors at the top.

Selecting a text analyzer

For the sites described in step 2, in our opinion, it is better to use the Rush Analytics tool: it allows you to manually select for analysis the competitor pages that correspond to one that is promoted for the intended purpose. For example, it makes sense to compare a catalog only with directories. Other types of pages should be excluded.

For other sites, services without the ability to filter competitors are suitable – for example, the Just Magic text analyzer.

Starting a Project

Popular text analyzers have a limit of no more than 10 search queries per page. To make the scan as efficient as possible, you should select the queries so that they include as many frequently mentioned words as possible.

We recommend that you use the following keyword combination: key query, 1 to 2 most frequent synonyms, remaining queries with different tails, such as “buy”, “price”, “cost”, “photos and price “, and others. It is important that these queries belong to the same cluster in the semantic kernel, otherwise, the text parser will check each of them for different results.

In the Rush Analytics tool, you can select a setting that allows you to count only competitors in search results for all queries at the same time:

An important feature – the service allows you to view and adjust the list of competitors before the launch. Additionally, you can manually specify the competitors to analyze each search phrase.

After starting the project, we perform an Excel download.

Offload Analysis

Here are the results of a text parser test for a page in the “Hobs” category, which was taken as an example: The

the tool shows the difference between the number of occurrences of certain words in the semantics and the amount of text between landing pages and competitors in search results in various elements of the page: SEO – text – any large block of content is taken as is, text fragments – small blocks of text on the page, link anchors, H1 headers and the Title tag.

Pay attention to the value in the SEO text difference column and in the word count row: the


table shows that there is no text on the promoted page, but the competitors have it, and its length average is 72 words or 504 characters. Therefore, for the promoted page, you need to prepare a block of text of approximately this size.

Since the challenge is to determine an optimization approach for sampling, you need to examine the results of the text analysis for each page that is included in it.

Then we assess which texts are used by competitors – unique or generated.

Analyzing Competitor Text

Now you should go through competitor pages and find this very long text.

It is not always possible to immediately determine that we have a generator in front of us – a text prepared on the basis of a template. For example, if the model is different for different categories, and not only the name of the category, but also some additional characteristics acts as a variable part – for example, for washing machines, this is the capacity of the drum , the number of washing modes, the maximum number of revolutions.

The main signs of a generator on the same type of pages:

  • approximately equal volume;
  • the same structure of the texts, the number of paragraphs, and the construction scheme of the sentences;
  • the use of common estimated characteristics – for example, assortment size, price range, store USP.

To determine which type we are dealing with, it is necessary to analyze the competitors at a time for an entire sample. For example, you can do a summary in Excel like this:

Summary conclusions:

  • To list cooking surfaces, you need to prepare a block of text consisting of two parts: a single text about the product selection – and a general description stating the number of suitable products and the minimum price. The total volume is 150-200 words.
  • For electric and gas cooktop listings, it is advisable to prepare a unique text and template for 150-200 words, but you can first adjust the template text to 100 characters and see if it will be enough for the growth of positions.
  • To list induction surfaces, you can use a template of up to 100 words or not write the text at all.

If most of the competitors in the example configure the generated text, you can use the same approach.

If more than 60-70% of competitors write unique content, it makes sense to adhere to the same strategy, otherwise, there is a risk of not even making it into the top 30.

Competitor Tag Analysis

At this point, It is convenient to analyze the Title and Description tags of the same type of competitor pages – for example, using the Screaming Frog SEO Spider or Netpeak Spider desktop programs – and see how the Title and Description tags of the meta tags look. Typically, template tags are visible to the naked eye: they are not very consistent text, often have end-of-case or letter-case errors, rarely contain synonyms for keyphrases.

An example of the generated Title tag: “buy hobs from XXX online store, low price, owner reviews.” Large catalog, description, characteristics – Moscow. ”

An example of a handwritten title label:” buy a baking tray in Moscow, more than 250 products at the price of 13,000 rubles, fast delivery of a new hob, guarantee, and refund in YYY. ”

There now remains to determine the number of individual models to prepare for the site and to select the best competitors for the benchmark. Different options are possible: for example, a single model for the same type of pages or one model per category until the 1st, 2nd, or 3rd level of nesting.

When preparing templates for the Title tag, you should pay special attention to the text parser recommendations on the number of occurrences.

For example, to list the category “Hobs “in the Title tag, you must use an occurrence of the keywords” hob “and” hob “in an exact or modified form:

Case: testing the method on two different sites

Consider applying the above approach in two cases from the Nectarine agency – for an online store of household appliances and a showcase site for a tire manufacturer.

The main difference between the two is the number of product categories. For an online store, there are more than 10, for a tire manufacturer there is only one.

Online store of household appliances

It is a young brand site without a branching structure. Each category – washing machines, refrigerators, and the like – has a list, not to mention the pagination pages. The product line includes 300 to 400 items.

Home appliances are a popular topic that never loses in demand, and therefore very competitive. Yandex’s top 10 leaders include multi-brand giants MVideo, Eldorado, Citylink and the Yandex marketplace.

The stopping factors of the project are the reluctance to implement product selections in the next six months, as a redesign and relocation of the site are planned.

Stimulating factors – assortment extension by 20% in the last trimester.

The goal is to do an initial content optimization without time-consuming and time-consuming improvements in order to achieve a slight increase in search traffic.

Analysis of competitor advertisements confirmed that individual preparation of texts and tags is required.

The main textual content of the product sheets in the leaders of the search results is a table of technical characteristics and a notice block. Some competitors have unique product descriptions, but due to the high drafting labor costs, the issue of preparing the descriptions has been postponed. To optimize the tags, we used the template generation configuration method.

The texts generated by the models were published on the site at the end of January. Since then, the share of semantic core queries in Yandex top 10 has increased from 7% to 12% – compared to 3,200 monitoring queries:

in Google, the share of queries has increased from 5% to 8%:

in Combined with other improvements to the site, which occurred at the same time, the number of organic orders increased by 36% in three months.

Pagination Pages – the second and subsequent pages of a list of products in the same category; is used when all the product sheets of a category do not fit on a single page


SEO specialist


  • will accompany you on the path of a new profession – obtain organic traffic and increase the visibility of your site in searches engines
  • we will teach you how We will teach you to think, find non-standard ways to develop projects, and make decisions quickly

Tire manufacturer website

Despite the presence of only one category – tires, the site has an extensive structure lists of products by diameter and size ⟶ a large number of similar pages difficult to work manually.

Analysis of the research results confirmed that competitors configure tags based on patterns and place blocks of generated text – both on listings and on product listings.

In this case, you cannot use simple templates with a single trigger, because the uniqueness of such texts on a site with thousands of ads will be too low.

Search engines are willing to forgive the negligence of aggregators who have earned their trust due to other factors – for example, brand awareness, product line, and the ability to place an order. But small single-brand sites need to do everything they can, especially when the main source of non-brand traffic is low-frequency query traffic to full-size ads.

One way out of this situation was a variable template for text: it not only changes the trigger size for the lists and the name of the card product but also uses various synonymous expressions.

Since the implementation of the tags and texts – in December and January – the number of requests in the top 10 Yandex has increased from 54% to 60%:


And in Google – from 49% to 55%: as

a result of the introduction of the tags and texts template – as well as the elimination of a number of technical errors – the number of targeted sessions on the site increased by 30% in two months.

Trigger – part of the text template that changes depending on the target page

To sum up

With the right approach, competitor analysis will help you determine where you can save writing effort and cost, and where you will need to individually develop the content of the text.

For sites with several pages of the same type, general is often used.

However, even if you choose templates as the primary method of optimizing content, you can make sure that the generated texts are highly informative and unique. The former can be achieved using quantitative indicators – such as the number of products or the price range. And uniqueness is achieved by using synonyms in the model itself.

For sites with a wide range of products in various categories, it is more likely to use a combined approach: prepare unique content for top-level pages and configure tag generation and text generation for child pages. On these web resources, we recommend that you analyze each section separately and pay close attention to competitors.

Whichever approach you choose, after making changes to the site, you should perform a position growth control measurement and, if necessary, adjust the strategy. A competent post-analysis will allow you to consolidate the positive dynamics of visibility and search traffic, and therefore to bring more requests and orders.

Choosing a Text Optimization Strategy for Ecommerce Sites: Generate or Unify

Leave a Reply

Your email address will not be published. Required fields are marked *

Scroll to top