Marketing

How Businesses Use AI to Create Content

Original in Russian: blog.skillbox.by
How Businesses Use AI to Create Content / Skillbox Media

Contents:

Free education: "Practical course on neural networks"

Find out more

Recently, we shared stories about digital experts who fell victim to attempts to replace them with neural networks - and these attempts were unsuccessful. The use of artificial intelligence resulted in a decline in reach, publications failed to satisfy audiences, and ultimately led to project closure.

We spoke with entrepreneurs and managers who have successfully integrated artificial intelligence into their companies. The next section presents five examples demonstrating the effective use of neural networks.

  • The blog was scaled up, resulting in a tenfold increase in traffic.
  • The neural network was trained, which reduced the time spent on drafting by 30%.
  • With the implementation of artificial intelligence, the creative process was significantly accelerated: now it takes only one hour to develop ten ideas instead of a whole day.
  • Editors were provided with training in the use of artificial intelligence, which freed them up from 80% of their daily duties.
  • A copywriter was replaced with artificial intelligence, which freed up time to focus on new projects.

"Increase in blog traffic: growth from 8,000 to 80,000 unique users per month"

"A sharp jump in blog traffic — a tenfold increase in unique visitors"

"The blog's popularity has increased: 80,000 unique visitors per month instead of 8,000"

"Impressive growth in blog traffic: from 8,000 to 80,000 unique visitors per month"

"The blog is experiencing a boom: a tenfold increase in the number of unique visitors"

Co-founder and CEO of the online learning platform Zerocoder.

In 2023, we were inspired by the first examples of ChatGPT use and decided to test our hypothesis: is it possible for content created in collaboration with a neural network to be perceived as naturally as texts written by a human?

Our corporate blog, "I am Zerocoder," was chosen as the experimental platform. Over the past two years, all content was created by professional copywriters. The editorial team published about 30 articles monthly, spending 500 thousand rubles on this. Despite significant investments, the average number of readers remained at 8 thousand per month. As a result, this blog did not rank among the ten most effective marketing channels of our company in terms of traffic volume.

In the spring of 2023, we took steps to transform our blog into a full-fledged product. One of the key stages of this process was the hiring of a neurocopywriter—a professional who works with text in tandem with artificial intelligence. Their responsibilities go beyond prompt creation: they develop content structure, refine model inputs, fact-check, and adapt the text style.

Over the course of six months, we significantly increased our publication volume: from 30 to 100, and then to 300 articles per month. As a result, our blog traffic increased tenfold—from 8,000 to 80,000 unique visitors per month. Meanwhile, time spent on the site, engagement, and bounce rates remained high, confirming that scaling is possible without sacrificing content quality.

As a result, costs were reduced. Copywriters, who had previously been responsible for writing large volumes of SEO content alongside other tasks, now focused on larger, high-quality projects and more complex texts. This approach helps them avoid professional burnout without reducing their income.

The cost of one article has decreased from 2,500 to 800 rubles. If we write 300 articles monthly, this allows us to save approximately 240,000 rubles on their production alone. This amount should be supplemented by savings on author and editor fees, which would have been necessary under the previous work model. Thus, the total savings amount to approximately 1 million rubles per month.

We recognize that users are able to recognize signs indicating the use of artificial intelligence in content creation. In this regard, we regularly conduct open tests: during webinars and in our Telegram community, we encourage participants to try to determine which text is human-written and which was generated. As a result, approximately 70% of respondents are unable to detect the difference.

The main conclusion is that a neurocopywriter is not a replacement for authors, but, on the contrary, serves as an addition to the team. It frees you from routine tasks, accelerates production processes, and allows you to focus on creating deeper, higher-quality content. As a result, the economic impact can reach one million rubles per month.

Over two years of collaboration with artificial intelligence, reach rates have remained stable, content perception has not deteriorated, and the user churn rate remains stable at 11%. While in 2023, neural networks required refinement in half of the cases, today more than 80% of created materials require only minor edits. Since July, we have been testing fully automated text generation.

To improve editorial efficiency, we strongly recommend access to one or two text neural networks, such as corporate ChatGPT or GigaChat with the "smart editor" feature. It would also be useful to have at least one visual neural network, such as Midjourney, for image generation.

The situation with the neural copywriter serves as a clear example of how the correct use of artificial intelligence can not only reduce costs but also increase efficiency. Instead of cutting staff, we took a different approach: we redistributed roles within the team and created an editorial ecosystem in which humans are responsible for strategic leadership, while AI helps scale processes.

Co-founder and CEO of Zerocoder Internet University.

In In 2023, we were impressed by the first examples of ChatGPT use cases and decided to conduct an experiment: could content created in collaboration with a neural network be perceived as naturally as human-written texts?

We chose our corporate blog, "I Am a Zero Coder," as the experimental platform. For the past two years, content had been created exclusively by copywriters. The editorial team published approximately 30 articles monthly, spending 500,000 rubles. Despite significant investment, our audience remained at only 8,000 readers per month. In terms of traffic volume, our blog did not rank among the top ten most effective marketing channels.

In the spring of 2023, we began transforming our blog into a full-fledged product. One of the most important stages was hiring a neurocopywriter—an expert who works with texts in tandem with artificial intelligence. Their responsibilities go beyond generating queries: they shape the content structure, refine input data for models, perform fact-checking, and adapt the text style.

Over the course of six months, we managed to significantly increase the volume of article publications: from 30 to 100, and then to 300 per month. As a result, our blog traffic increased tenfold—from 8,000 to 80,000 unique visitors per month. Key metrics such as time spent on the site, engagement rate, and bounce rate remained high, confirming that content can be scaled without compromising its quality.

At the same time, costs decreased. Copywriters, who previously handled numerous SEO texts alongside their other responsibilities, now focus on developing large, high-quality projects and complex materials. This approach helps them avoid burnout while maintaining their income.

The cost per article has decreased from 2,500 to 800 rubles. If we write 300 articles per month, this allows us to save approximately 240,000 rubles on writing alone. This amount should also include savings on author fees, as well as editors who would have been required under the previous system. All in all, the total savings amount to approximately 1 million rubles per month.

We recognized that users can recognize the characteristics of text created by artificial intelligence. That's why we periodically conduct open tests: during webinars and in our Telegram community, we challenge participants to determine which text is human-written and which is artificially generated. Approximately 70% of participants cannot detect the differences.

The key idea is that a neural copywriter is not a replacement for writers, but, on the contrary, becomes a valuable addition to the team. They take on routine tasks, speed up processes, and allow you to focus on creating high-quality content. As a result, savings can amount to one million rubles per month.

Over the past two years of working with artificial intelligence, we have seen stable reach, and content perception has not deteriorated. The bounce rate remains at 11%. While in 2023, neural networks required refinement in half of the cases, now over 80% of materials require only minor edits. Since July of this year, we have been testing fully automated text generation.

To improve editorial efficiency, it is advisable to provide access to one or two text neural networks, such as corporate ChatGPT or GigaChat with a "smart editor" function. It's also worth having at least one visual neural network, such as Midjourney, for image generation.

The situation with the neural copywriter vividly illustrates how the correct use of artificial intelligence can simultaneously reduce costs and increase efficiency. Instead of reducing the size of our team, we decided to redistribute responsibilities, creating an editorial ecosystem in which humans are responsible for strategic planning, and AI helps in scaling processes.

"Artificial Intelligence as an Assistant: Liberation from Routine Tasks for Deeper Analysis"

Head of SEO at the digital agency INET Studio.

Artificial intelligence can indeed act as a copywriter, but the process of using it is not a simple "turn it on and forget it." Achieving optimal results requires extensive model tuning and training tailored to the client's unique goals, requirements, and constraints.

We use AI:

  • to create short texts - product descriptions, specifications, answers to frequently asked questions;
  • technical content involves developing small programs designed to automate mundane tasks, such as data parsing, processing information, and changing formats.
  • analysts monitor traffic anomalies, examine vast amounts of data, and detect patterns.
  • preparatory work involves developing the foundations for upcoming articles, as well as collecting and organizing the necessary information;
  • writing materials on simple and accessible topics, such as date ideas, explanations of dream meanings, and horoscopes.

However, even in this case, the process is not limited to simply clicking a button - text creation is carried out in stages. Artificial intelligence acts as an assistant that takes over routine tasks, allowing us to focus on analytical aspects.

In the process of creating expert content, we conduct experiments with an internal agent, training it and testing its writing abilities. However, we do not currently offer this service to clients, as the risks remain significant. However, it is safe to say that this format will become relevant as early as 2026.

Currently, the main goal of a copywriter is to perfect the content created by artificial intelligence. This includes fact-checking, adding expert opinion, and infusing the human touch that algorithms do not yet possess. Thus, the emphasis has shifted from creating text from scratch to editing, fact-checking, and enriching existing materials. Naturally, the final result is largely determined by the quality of the initial request.

The use of artificial intelligence has significantly increased work efficiency. The time required for preparing drafts and collecting source information has decreased by approximately 30%. This, in turn, makes it possible to create more content within the established deadlines.

Artificial intelligence significantly simplifies the process of detecting trends and anomalies, allowing for quick changes to the content strategy. Although exact data is difficult to determine, the savings are impressive: instead of traditional manual analysis, we get automated solutions that do not require programming knowledge.

Head of SEO at the digital agency INET Studio.

Artificial intelligence can indeed act as a copywriter, but the use It does not imply a simple “set it and forget it” principle. Achieving optimal results requires careful tuning and adaptation of the model to the specific tasks, requirements, and constraints of the client.

We use AI:

  • to create concise texts - product descriptions, their characteristics, as well as answers to frequently asked questions;
  • technical content involves developing code to create small programs that help automate everyday tasks, such as data parsing, processing, and format conversion.
  • analytics is the process of monitoring traffic deviations, examining vast amounts of data and discovering patterns.
  • preparing materials includes forming the basis for upcoming articles, as well as searching and organizing information.
  • developing materials on accessible topics (ideas for romantic meetings, dream interpretation, astrological forecasts).

However, even in this case, the process is not limited to simply clicking a button - text creation occurs in stages. Artificial intelligence acts as an assistant for us, taking over routine tasks, allowing us to focus on analytical work.

In the process of creating expert content, we conduct experiments with an internal agent, training it and testing its writing capabilities. However, we are not yet ready to offer these materials to clients, as the risks remain significant. However, it is safe to say that this format will be relevant by 2026.

Today, the role of a copywriter is to bring the text created by artificial intelligence to the final version. This includes fact checking, adding a professional perspective, and introducing a human touch that algorithms currently lack. As a result, the emphasis has shifted from creating text from scratch to editing, verifying information, and improving the content. Of course, the end result is largely determined by the quality of the initial request.

The implementation of artificial intelligence has made it possible to significantly increase work efficiency. The time spent preparing drafts and collecting source information has decreased by approximately 30%. This opens up the prospect of increasing the volume of content produced while maintaining the same deadlines.

Artificial intelligence significantly accelerates the process of detecting trends and anomalies, allowing you to quickly adapt your content strategy. While exact figures are difficult to estimate, the cost savings are impressive: instead of relying on manual analysis, we gain the ability to automate without requiring programming skills.

"Artificial Intelligence: Creatives that Outperform Human Efforts in Audience Engagement"

"AI Content: Sometimes More Successful in Attracting Attention Than Manual Creativity"

"AI Results: Creatives That Cause a Storm of Emotions in Viewers"

"Creative Comparison: How AI Outperforms Manual Creation in Audience Reaction"

One of the founders of the Chad AI platform for neural networks and the digital agency Advertalyze.

We're not sure that artificial intelligence will be able to completely replace content creators. According to McKinsey research, organizations using AI in marketing can achieve a 10-15% increase in efficiency. However, this effect is achieved precisely through the interaction of humans and machines.

We actively use artificial intelligence to develop advertising content, including banners, graphics, and text. For image creation, we use tools such as Midjourney, Nano Banana, and Flux, and generate text using various language models.

The process works as follows: initially, a copywriter develops the main idea or text, after which, using artificial intelligence, we create multiple alternative versions of this content for further analysis. The results obtained with AI often outperform the original texts in terms of effectiveness. AI is also becoming an indispensable assistant in the brainstorming process—different models suggest a variety of slogans and creative ideas.

We haven't abandoned copywriters—their roles have evolved. Nowadays, they not only create texts in Word, but also actively use artificial intelligence tools. Copywriters have a clear understanding of their goals, develop synopses, and use their own AI models, which allows them to create content faster and more efficiently.

The main result is a significant reduction in the time spent on content generation. Previously, developing ten creative ideas for an advertising campaign could take a full day of collaboration with a designer. Now, thanks to artificial intelligence, the same amount of work takes just an hour. This is especially evident in banner creation—finding attractive creative has become much easier and faster.

According to research, the use of artificial intelligence allows companies to reduce content creation costs by up to 70%. However, determining the impact on reach is difficult, as precise data is not yet available. Creatives developed with the help of AI sometimes receive more views and responses than those created manually, but it is premature to draw definitive conclusions at this point.

Nevertheless, artificial intelligence clearly has a beneficial effect on the results of advertising campaigns, allowing for a faster response to trends and providing the opportunity for rapid A/B testing of creatives.

Co-founder of the Chad AI platform for neural network solutions and the Advertalyze digital agency.

— We are confident that artificial intelligence cannot completely replace content creation specialists. According to McKinsey research, organizations that implement AI in their marketing strategies can achieve a 10-15% increase in efficiency. However, this increase is achieved precisely through the interaction and collaboration between humans and machines.

We actively use artificial intelligence to develop advertising content, including banners, visual elements, and text. To create images, we use tools such as Midjourney, Nano Banana, and Flux, while we form the text using various language models.

The procedure is as follows: first, a copywriter develops the original text or concept, after which, using artificial intelligence, we create various versions of this material for subsequent testing. AI-generated versions often outperform the originals. Furthermore, AI actively participates in the brainstorming process, suggesting various slogans and creative ideas through its models.

Copywriters remain a vital part of the process, but their roles have evolved. Today, they go beyond working in Word and actively utilize AI-powered tools. These specialists have a clear understanding of what they want to convey to the audience, develop synopses, and utilize their own AI models, allowing them to generate content with greater speed and efficiency.

The main result is a dramatic reduction in the time required for content development. Previously, creating ten creatives for an advertising campaign could take an entire workday when a designer was involved. Now, thanks to the use of AI, the same amount of work takes just an hour. This is especially evident when working with banners, as finding engaging creative has become significantly easier and faster.

According to research, the use of artificial intelligence allows organizations to reduce content creation costs by up to 70%. As for changes in reach, precise data is still difficult to obtain. Creatives developed using AI sometimes garner more views and engagement than those created manually. However, it seems premature to draw final conclusions at this stage.

Artificial intelligence certainly has a beneficial effect on the results of advertising campaigns, facilitating a faster response to current trends and providing the ability to quickly conduct A/B testing of creatives.

"We aimed to automate 80% of routine tasks"

"Automation: our priority in reducing routine work"

"We aim to transfer 80% of routine processes to technology"

"Artificial intelligence to the rescue: 80% of routine tasks under machine control"

"Work optimization: delegating 80% of routine duties to machines"

Head of the digital reputation management agency "Nevidimka".

Currently, we create core content for a number of our clients using artificial intelligence technologies. This includes, for example, writing SEO articles for corporate blogs and managing Telegram channels. Previously, a staff of several copywriters and editors worked on these tasks.

The problem was not the level of quality, but the pace and volume of work. Our editors are true professionals, but the majority of their time—up to 80%—was spent on producing standard copy. Only the remaining 20% ​​of their resources were available for what they truly excel at: developing strategies, creative ideas, and fact-checking. We hired highly skilled specialists to perform repetitive tasks.

We sought to change this ratio: transferring 80% of routine tasks to automated systems to free up 80% of our employees' time for more complex and creative tasks. Instead of cutting staff, we redesigned their roles. Copywriters were retrained and now work as "AI editors."

In essence, they acquired new skills at the intersection of editing and analytics, mastering the art of breaking down creative tasks into their component parts and communicating with the machine in its own language. Now, their primary task is no longer creating text from scratch, but rather managing neural network systems: creating detailed prompts, coordinating the generation process, and ultimately refining the raw content to the required level of quality, adding style, factual data, and expertise.

As a result, the content drafting process has accelerated by approximately 70%. The time senior editors spend on day-to-day writing tasks has decreased by 60–65%.

This freed-up time has allowed us to implement two new, more sophisticated services that we previously lacked the resources to implement: proactive reputation monitoring and communication strategy development.

The key lesson we've learned from our experience is that the focus shifts from task execution to task management.

Head of the digital reputation management agency "Nevidimka"

We currently create initial content for a number of our clients using artificial intelligence. This includes, for example, creating SEO articles for corporate blogs and managing Telegram channels. Previously, this work was handled by a team of several copywriters and editors.

The problem wasn't the level of quality, but the pace and volume of work. Our editors are true professionals, but they were forced to spend the majority of their time—up to 80%—creating standard texts. Only 20% of their efforts were left for tasks where they truly shine, such as strategy development, creative approaches, and fact-checking. We hired highly qualified specialists to handle routine tasks.

We sought to change the ratio, transferring 80% of routine tasks to machines to free up 80% of employees' time for more complex tasks. Instead of layoffs, we changed the roles of employees. Copywriters were retrained and now act as "AI editors."

In essence, they acquired new skills at the intersection of editing and analytical thinking, mastering the ability to break down creative tasks into their component parts and interact with the machine in its own language. Now their main task is not to create text from scratch, but to manage the neural network: formulating well-thought-out prompts, coordinating the generation process, and then refining the raw materials to the required level, including elements of style, factual information, and expert opinion.

As a result, the speed of content draft creation has increased by approximately 70%. The time senior editors spend on routine writing tasks has decreased by 60–65%.

Freeing up the experts' time has allowed us to launch two new, more complex services that we previously lacked the resources for: proactive reputation monitoring and communication strategy development.

The key takeaway from our experience is that the focus is shifting from task execution to task management.

"Quick Solutions to Business Problems with Ready-Made Prompts"

"Efficient Business Copywriting in Seconds"

"Ready-Made Prompts: An Instant Solution to Business Text Problems"

"Text Strategies for Business: Fast and Effective with Ready-Made Prompts"

"The Secret to Quickly Writing Business Copy with Ready-Made Prompts"

A luxury e-commerce content specialist and managing partner of an event agency.

I use artificial intelligence to generate content in three areas: managing two personal Telegram channels on various topics, creating artistic descriptions for brand products, and curating materials for event agency, including its website and social media.

First of all, artificial intelligence significantly reduces the time I spend. Previously, researching information for posts would take all day. Now, AI processes large volumes of data in minutes and provides accurate answers to my queries, eliminating the need to spend hours browsing through various sources.

I write 500-character product texts that convey the atmosphere and highlight the unique selling proposition. I develop a custom prompt for each client, which allows me to quickly create descriptions in the required style, avoiding the perception of "plagiarism."

At the agency, content-related tasks were often postponed due to time constraints or a lack of inspiration. Now, thanks to ready-made prompts, I quickly receive texts that help solve business problems in just a few seconds. This also simplifies staffing issues: the need to search for and hire specialists for small or occasional assignments disappears.

With the availability of a staff budget, I was able to attract not just a contractor, but a true manager. Such a professional, with a deep understanding of the tasks, will be able to effectively formulate assignments for artificial intelligence, as they will be clear on the exact result required.

Now let's move on to the results. A regular publication schedule has been established on personal channels. Currently, there is a series of materials planned for several weeks in advance. Increased activity has allowed us to attract a regular advertiser.

The agency founders have been freed from the routine work of writing texts. In addition, we have begun actively managing our Pinterest account, something we previously lacked the time for. As a result, views increased by 186%, engagement by 620%, and the number of active users jumped by 300%, even though only 30% of the planned content was published.

For me, artificial intelligence is an important strategic tool. It can take over routine tasks, allowing resources to be redirected from task execution to task management, thereby increasing both personal and business efficiency.

These stories, like the previous ones, demonstrate that neural networks are not "job killers" and do not possess unlimited intelligence. In fact, they are a tool that can lead to both positive and negative results. In order for artificial intelligence to meet our expectations, it is necessary to master the skills of effective interaction with it.

A content specialist for luxury e-commerce, managing partner of an event planning agency.

I use artificial intelligence to generate content in three key areas: managing two personal Telegram channels on various topics, Creating creative product descriptions for brands, as well as curating materials for an event agency, including its website and social media.

First of all, artificial intelligence significantly reduces the time it takes me to complete tasks. Previously, collecting information for publications took me all day. Now, AI can process vast amounts of data in minutes and provide accurate answers to my queries, freeing me from the need to endlessly browse various sources.

I develop 500-character product descriptions that capture the atmosphere and unique selling proposition. I create a custom prompt for each client, which allows me to quickly generate texts in the required style, avoiding a sense of templatedness.

At our agency, content-related tasks were often postponed due to lack of time or a lack of creative inspiration. Now, thanks to pre-prepared prompts, I can quickly generate texts that effectively solve business problems. This also eliminates the staffing problem: there's no longer a need to search for and hire specialists for small or irregular projects.

Now that I have a budget for staff, I can attract not just a line worker, but a true manager. This professional, with a deep understanding of the tasks, will be able to competently formulate assignments for artificial intelligence, as they clearly understand the required end result.

Now let's move on to the results. Publication has become a regular process in personal channels. Currently, a stockpile of materials has built up that covers weeks in advance. Increased activity has attracted a regular advertiser.

The agency founders have freed themselves from day-to-day copywriting. Furthermore, we have begun actively managing our Pinterest account, something we previously didn't have time for. As a result, views increased by 186%, engagement by 620%, and the number of active users increased by 300%, despite only publishing 30% of the planned content.

For me, artificial intelligence is an important strategic resource. It frees you from monotonous tasks, which makes it possible to redistribute energy and attention from performing operations to management, thereby increasing both personal and business efficiency to a new level.

If you are interested in the topic of neural networks, we recommend that you familiarize yourself with the materials from Skillbox Media.

  • We will share ways to start earning money using neural networks, even if your experience with them is limited.
  • Studying neural networks and abandoning the 5/2 schedule: the story of one of the pioneers in the field of neuroart.
  • Thirty powerful neural networks for solving a variety of problems.
  • The DeepSeek neural network: capabilities and reasons for its popularity in discussions.
  • Artificial neural networks for marketplaces: 4 tools that can significantly speed up work processes.

Practical study of neural networks

Find out more