• 17.03.202417.03.2024
• admin

Text is the main tool of communication between a business and a buyer in the digital world , the simplest and at the same time the most complicated way of communication. It comes in various forms: inscriptions on buttons, advertising slogans on banners, calls to action, articles on category pages and on the blog.

There are two traditional ways to use this powerful tool:

1. “Let it be”

With the arrival of ChatGPT in our world, you can generate texts of average quality for free and in just a few minutes. Well, how is the main meaning of the text hidden in the fifth paragraph from the end? The main thing is that there are keywords! 

There are several signs of poor quality text: 

• missing meanings – the text talks about everything, but nothing specific; 
• lexical poverty or excessive complexity of the text – repetition of the same words or use of outdated, specific word forms; 
• outright absurdity — especially common when writing texts by artificial intelligence without further checking by an editor; 

From real examples: in the American-language text on psychology, instead of the word “grief”, the word “burning” was used. 

• stylistic inconsistencies — for example, if the text is written in a familiar and nonsensical tone for a brand of expensive men’s clothing, which maintains an official tone of communication on social networks; 
• semantic inconsistencies – phraseology should be used carefully, especially in texts in foreign languages; 
• Oversaturation of keywords can even lead to penalties from search engines. 

The use of such texts can negatively affect conversion and even adversely affect the ranking of the site in search engines. 

2. High-quality text content

Ideally, the text should meet several key parameters. We can say about quality text content: 

• well written; 
• the topic of the text corresponds to the keywords; 
• used LSI words; 
• is aimed at readers, answers their questions or solves their problem; 
• keeps in the same style with the voice of the brand;
• divided into paragraphs, numbered or bulleted lists for a better user experience;
• beautifully designed – different blocks and elements are used, maybe even quotes.

Quality text pleases the eye of the reader and helps him to maintain a pleasant impression of your site. 

Preparation of the technical task

Writing a text begins with the creation of a technical task. First of all, decide on the parameters of the future text:

1. Keywords . These are words or phrases that determine the text writing vector. They describe what the text will be about. Can be used both in a direct (that is, in a completely identical morphological form) and in an indirect form. It is important not to overload the text with keywords.
2. Text volume . Usually the customer gives a range of characters, but the ideal amount of text is a complete answer to the user’s question. You can focus on the competitor’s text.
3. Theses . Brief main topics of the text, future subheadings.
4. Uniqueness . The percentage difference of the text from those that exist on the Internet. Should be as close to 100% as possible.

After the article is written, it is important to read it and check it for compliance with the technical task.

Step 1. Checking the thesis plan

You take the finished text, read it and compare it with the theses of the technical task. There is no universal recipe for verification, but the text should have a clear “answer” to the thesis.

For example, given theses:

1. What is Easter decoration?
2. Types of Easter decor.
3. How to prepare the house for Easter?
4. 3 best ideas for Easter decoration.

After reading, you should learn about each of the theses. If this did not happen, add the text and add information.

Step 2. Uniqueness check

Uniqueness is the percentage difference of the text from those that exist on the Internet.

I will look at several uniqueness checkers.

Serpstat

If you’re an SEO professional who’s familiar with Serpstat , you may have noticed some new features. One of them is a plagiarism check. In order to take advantage of the new feature, you must be a user with a Lite subscription or higher.

To use Serpstat, you need to log into your account and perform the following steps: 

1. Click the AI ​​Content Tools button, then select Plagiarism checker. 
2. You will be taken to a page where you need to enter your Serpstat API token. It is located in the “Account Management” tab.

You can check the number of limits remaining for the “Content API” feature in the “My Profile” tab, in the “Tools” column.

The number of spent credits, Serpstat’s internal currency, for the plagiarism check is calculated in requests:

• 1 request = 500 credits for English;
• 1 request = 750 credits for other languages.

3. After the token has been entered, a window appears with the possibility of entering text for verification. The minimum number of characters is 256, the maximum is 10240.

4. You should choose the language in which the plagiarism search will be conducted. As of June 2023, 26 languages ​​are available for verification: 

• Romance languages ​​- Spanish, French, Italian, Romanian, Portuguese;
• German – English, Danish, German, Dutch, Swedish;
• Slavic — Bulgarian, Czech, Polish, Russian, Slovak, Slovenian, Ukrainian; 
• Asian — Japanese, Chinese;
• Turkic – Turkish; 
• Finno-Hungarian — Estonian, Finnish, Hungarian;
• Baltic – Lithuanian, Latvian;
• others are Greek. 

5. You can specify which sites should not be considered during the plagiarism analysis. They must be entered through a comma. 

6. Once you’ve entered your text and selected the settings for it, click the “Calculate” button.

The results are presented as a percentage of plagiarism and with links to sites where similar fragments of text were found. 

The URL score will also be given – this is a table with the pages with the highest percentage of text matches. Each URL in this block has a Score—the proportion of offers that were found on that page. For example, if the score is 0.1, it means that 10% of the sentences of the added text were found on this page.

Pros: 

• a large selection of languages; 
• there is no need to separately purchase access to the service; 
• clear interface. 

Cons: 

• expensive if there is no need to use other functions of Serpstat; 
• the feature is new, so there may be technical flaws.

Plagiarismdetector.net

For the services of the free version of the Plagiarismdetector.net service  , you do not need to register or have an account. You can just go to the site and follow the instructions.

1. Enter text into the program. There are several ways to do this:

• copy the text you need to check and paste it into the input window;

find and upload a file from the desktop using the “Upload File” function or via DropBox;

• load the URL of the text page using the Check By URL window.

I do not recommend this method of use , because not all information is pulled and you can lose an important part of the text. Parts of the description from the footer and other technical elements can also be pulled up.

2. In the “Exclude URL” field, enter the links of sites that should not be taken into account during the plagiarism check.

3. After that, click on “Check Plagiarism” and wait a few minutes.

The result is displayed at the top right, showing the ratio of original content to plagiarism. Below are the sources of plagiarism with the percentage of their use in the presented text. 

The report can be saved in a Word document. There is no difference between Download Report and Download Word Report, both links go to the PDF to Word conversion page. 

After conversion, you will receive a file with a screenshot of the verification results. It cannot be edited. 

Plagiarismdetector.net allows you to check for plagiarism in 16 languages: 

• Romance languages ​​- Spanish, Portuguese, French, Italian;
• Germanic languages ​​— English, German, Swedish, Dutch;
• Slavic languages ​​- Russian;
• Asian — Vietnamese;
• Turkic – Turkish;
• others are Hindi, Arabic, Indonesian, Filipino, Malay.

The Ukrainian language was not added as of June 2023. 

The cost of using the PRO version starts at $5 per week for 50 pages of text. Annual minimum subscription will cost from $110. 

Pros: 

• a large selection of rare languages ​​of the countries of the archipelago; 
• you can buy the PRO version for a week.

Cons: 

• cannot check more than 1000 words in the free version; 
• low-quality conversion to Word; 
• there is no Ukrainian language. 

Duplichecker

Like the previous site, Duplichecker does not require registration to use the free version. Follow the instructions to check the text.

1. Upload text to be checked in the field, use file uploads in text formats. The service has an option to check through the URL, but it gave me an error, saying that the text on the page is too small.

2. After entering the text, pass the Captcha and click “Check Plagiarism”. 

3. Get acquainted with the results of the inspection: 

• percentage of plagiarism and unique content; 
• link to the page from which the text was copied; 
• word count

4. You can save the results of the check in pdf or word formats. Unlike Plagiarismdetector.net, the conversion result can be edited.

Duplichecker supports plagiarism checking in 18 languages: 

• Romance languages ​​- Spanish, Portuguese, Italian, French, Romanian;
• Germanic languages ​​— English, German, Dutch, Swedish, Danish, Norwegian;
• Slavic languages ​​- Russian; 
• Asian — Vietnamese, Korean, Japanese;
• Turkic – Turkish;
• others are Indonesian, Arabic.

The minimum subscription price for 30,000 words is $10 per month.

Pros:

• there is a free version; 
• there are rare European and Asian languages.

Cons: 

• minimum subscription for the paid version is 1 month; 
• there is no Ukrainian language.

How to choose a plagiarism check service?

There are several important criteria that you should pay attention to when choosing a plagiarism check service. 

1. Which languages ​​should be tested? 

Each of the presented services has its own peculiarities in terms of language selection:

• Serpstat has the largest list of Slavic languages, including Polish, Slovak and Bulgarian, it is the only site that includes Ukrainian; 
• Plagiarismdetector is great if you work with rare languages ​​like Indonesian, Malay or Hindi; 
• Duplichecker.com comes to the rescue if you need to check Scandinavian or Korean languages. 

2. Subscription price.

Depending on the number of characters and words you need to check, the cost of the subscription varies. For example, the minimum cost of the package per month: 

• Duplichecker.com — $10 for 30,000 words (it is possible to pay weekly rates);
• Plagiarismdetector.net — $20 for 71,000 words;
• Serpstat – uniqueness check is included in the standard package of $69, but this is the total cost of access to a lot of other SEO functionality . 

The market for plagiarism checking software and devices is quite large, so it is quite possible to find plagiarism checking services for most languages ​​in the world.

Step 3. Checking the use of keywords in the text

At the stage of creating the technical task, the specialist selects keywords and phrases for the copywriter. Depending on the purpose of the text, the requirements for using keys vary.

All indicators of the frequency of key phrases are very approximate – you can focus on them during a cursory primary analysis of the text, but only a specialist can draw a conclusion about the quality of the text after proofreading.

To quickly check the keywords in the text, you can use the ISTIO program – a free service for selecting LSI words and checking the text. 

Currently, the service supports word searches in four languages: 

• English;
• Russian; 
• Ukrainian;
• german

How does ISTIO work?

1. Enter the text for verification in one of two ways: 

• copy the link of the page, in this case the service itself will pull the entire text of the page for verification (including text parts of buttons and other elements);

• copy and enter the desired text manually.

2. You can specify which keywords the service should find.

3. Click on “Key Selection”.

The downside of this service is that it does not find long-tail keywords , but breaks them into several parts. For example, the keyword “Ukrainian military” will be split into “Ukrainian” and “military”, which will be highlighted throughout the text.

4. Read the inspection results report, namely: 

• the number of keywords of this type in the text (“Count”);
• to what extent the presented keyword correlates with the topic of the text (“Relevance”);
• the percentage of the keyword in relation to all important words in the text (% in core);
• percentage of the keyword in relation to all (ie, including stop words) words in the text (% in text).

Stop words are prepositions and interjections. They are automatically included in the service and do not need to be entered manually.

The use of keys in the text can also be checked manually through the “Search” function of any text editor.

I recommend learning the ” hot keys ” for ” Search “:

• PC — Ctrl + f;
• MacOS — ⌘ + f;
• ChromeOS — Ctrl + f.

The most important quality indicator when checking a text for the occurrence of key phrases (or words) is its readability. If in the process of reading the text, in order to understand the text, you had to reread the phrase several times – the sentence should be rewritten. Remember that the text is written for the reader.

Step 4. LSI copywriting. LSI usage check

As mentioned earlier, LSI (latent semantic indexing) is the words that define the subject, excluding key phrases. After reading the list of these words, it immediately becomes clear what the text is about, although there is no key in the list.

For example, the word “stove” is killed in the search. The stove is too ambiguous a key if you look at it alone, but when you plug in the LSI “kitchen, gas, burners, electrical, appliances, delivery”, it becomes clear that it is a kitchen stove, and not, for example, a concrete one.

There are several popular free methods for selecting LSI keys.

Search hints and recommendations of search engines

By entering the desired keyword in the search bar of any search engine (Google, Bing, etc.), you will find search results that can appear as LSI words. 

Serpstat

Use automatic services for matching, such as Serpstat.

How to use Serpstat to find LSI words? 

1. Enter the keyword you need in the “Keyword” space.
2. In the panel on the left, select the item “Keyword research” – “Search suggestions”.
3. After that, select the search region at the top. For example, the USA. 
4. Optionally, you can choose tags from among the most popular words “Most popular words”.
5. done Check out the suggested word options, you can filter them and export them as needed.

The use of automatic services such as Serpstat greatly facilitates the search for LSI keys and helps to obtain the most accurate information. 

That’s it

You can also use the already mentioned ISTIO service, which has a rich functionality for highlighting LSI words in the text of competitors: 

1. The Map button shows the importance of different words in the text based on the number of times they are mentioned. The more important the word is for understanding the text, the bigger and blacker it is. Prepositions, interjections, and abbreviations are usually highlighted in light gray. 

2. With the help of the “Analysis” button, you will receive a report listing all important LSI words in the text.

3. By clicking on the “Dictionary” button , you will get access to all words, except stop words, found in this text. They can easily be copied.

4. To the right of the analysis is a help with the main characteristics of the text .

In this window you can find a lot of useful information to understand the quality of the text:

• number of characters with and without spaces; 
• the number of all words in the text; 
• number of unique words; 
• the amount of “water” — that is, words that do not carry an informational load; 
• the frequency of repeating the same words, the degree of text spam ;
• the number of unique words without stop words;
• the topic of the text (not always determined);
• language of the text;
• 10 most used words in the text.

All this information can help bring the text to perfection and achieve the set marketing goals.

Should we trust the “water” indicator of the text?

It is recommended to double-check the results of automatic text analysis, because they are quite relative and should be considered only in the general context. 

For example, ISTIO considers prepositions “water”, but it is impossible to abandon them due to the peculiarities of the English language. Such services help to see accumulations and excesses, but they should not be considered an axiom. 

Step 5. Proofreading the text

The most important stage of text verification. When proofreading, be sure to read the text carefully and thoughtfully. If you reread phrases or sentences several times, feel free to rewrite them or throw them away.

You should always consider the target audience of the site. For example, a text written for young mothers and for professional lawyers will have a completely different stylistic color.

The density of keywords, relevance to the topic, and water content — all these indicators do not give an accurate assessment, unlike proofreading.

The main rule at this stage is that the text should be easy to understand, phrases should not “cut the ears” and the article should be interesting.

Separately, it is worth avoiding the concept of clericalism – that is, complex constructions and words that make it difficult to understand the text. 

What services to use for proofreading?

The best programs for correcting spelling errors For checking textsI suggest using the following services. All of them have a paid version, but the main functionality can be used for free:

1. Grammarly  will help you find spelling and punctuation errors, as well as clerical errors in English. The Grammarly team is working on its own artificial intelligence that will help improve the text. The app is currently in beta and available to users. 

2. Hemingway App  helps to make the text “lighter” stylistically. The service highlights excessively long sentences, misuse of passive verbs and adverbs. Works only with English language.

3. Languagetool  also highlights spelling, stylistic and punctuation errors. Unlike other services, Languagetool supports many languages, including Ukrainian. 

In general, you should use any services with automatic text verification with caution, as they are not a panacea for text errors. It is better to consider them as a tool that will help prevent misprints and “sloppy looking”. 

Universal tips for improving the text

There are a few rules to follow to improve the overall quality of your text: 

1. If you can use active verbs instead of passive ones , do so. Of course, if it is not a scientific text, where the passive voice is recommended. 

Examples: 

• “lawyers defended the defendant in court” sounds better than “the defendant was defended in court by lawyers.”
• “our vitamins will improve your eyesight” is better than “our vitamins will improve your eyesight”.

2. Avoid stationery and stamps. 

Fixed phrases do not always sound beautiful. If you do not use such combinations in ordinary speech, you should think about their appropriateness in the text. 

Examples of clericalisms: 

• watering flowers = watering flowers;
• to perform a prayer = to pray;
• to have a beneficial effect = to help, to improve. 

3. Understand your reader. 

If you need to use specialized terms, professionalisms, dialectics, abbreviations, anglicisms or archaisms, explain them to the reader. If you write about fakes in a youth magazine, you can count on the readers to understand what they are. But the audience of the garden magazine may not know such words. 

I advise you to use anglicisms carefully and, if possible, replace them with analogs. For example, in regular media it is recommended to replace: 

• “skamit” to “deceive, engage in fraud”; 
• “optimize processes” — “improve processes”;
• to go to a “meeting” — “to participate in an online meeting.”

4. If you write using ChatGPT, check all unfamiliar words for stylistic features. AI often uses complex stylistic constructions and expressions instead of simple synonyms. 

For example, in an English article about real estate for Turkey, the word “covet” was used, which translates as “covet”/”strongly want”. This word has a vivid meaning and is also used in the Bible in the context of “You shall not covet your neighbor’s wife.” In terms of content, the word is suitable, but stylistically and, most importantly, contextually, it is too complex for a text in a region with a low average level of English. 

Also, constructions like “Familiarize yourself with” can usually be replaced with a simple “learn”. 

If you are not sure that you understand exactly the contextual meaning of an English word or its appropriateness, I recommend checking yourself with the help of the Cambridge Dictionary , where the formality of the word is separately emphasized. 

The feeling of the text is what should be the main thing when checking it. If the eye clings to some paragraphs, they need to be corrected. Especially if it is a small commercial text. If you are not sure of its quality, read it out loud, because it will help you track its logic and sequence. 

Conclusion

Checking texts is a creative process and there are no templates or universal rules for it. Check the basic parameters, but be sure to use common sense.

The main stages of checking texts:

1. Checking the thesis plan . After reading the text, all theses should be covered in the article.
2. Uniqueness check . Don’t waste your time on non-unique content, it won’t do you any good anyway.
3. Keyword validation . Keys should be used in the text, but without fanaticism.
4. Checking LSI words . If you write text for your reader, use them even without a list from an SEO specialist.
5. Text proofreading . Be sure to check the text for literacy, use the appropriate services and general rules.