How to use Deep Research in ChatGPT, Gemini, and Perplexity to find topics

A practical breakdown was published on Habr on how to use Deep Research mode to find topics that can bring traffic and subscribers. Instead of a regular chatbot that paraphrases the obvious, the author suggests launching deep research, collecting a table of ideas, and then filtering them for your audience.

Why Deep Research is Needed

The main idea of the article is that the standard chatbot mode is poorly suited for finding strong content ideas. When a model doesn't have proper sources, it fills gaps with generic phrases, repeats what's already on the first page of search results, and produces a template list of topics. For a blog, Telegram channel, or media outlet, this isn't enough: you need not just ideas, but stories that are grounded in fresh market signals, discussions, numbers, and case studies.

The author considers Deep Research mode more useful precisely because it forces the model to work like a junior researcher. This mode scans more sources, compares different viewpoints, seeks confirmations, and can acknowledge when there's insufficient data.

In the material, ChatGPT, Gemini, and Perplexity are mentioned as working options: they have deep research modes that provide at least several free runs.

"If the bot doubts the answer, it will try to approach it from a

different angle."

How to Frame Your Request

The article suggests not asking the neural network for abstract "10 viral topics," but instead giving it a clear research brief. In the basic prompt, you need to describe the subject of research, the type of content, the publication platform, and the audience, then ask the model to study trends, reviews, events, and other authoritative sources from the last six months. The output should not be a stream of ideas, but a structured report with prioritization and an explanation of why these topics could take off.

• topic or market for analysis
• publication platform: Habr, VC.ru, DTF, and others
• description of audience and their pain points
• period for searching signals
• report format: table, sections, preview, evidence

The author separately recommends providing the model with your own internal data: statistics on your best posts, successful competitor publications, or examples of materials that have already worked in your niche. This is an important point because Deep Research looks for external signals, but it's the historical data of your channel that helps understand which of the found ideas will truly resonate with your readers, and which will remain just a beautiful topic on paper.

How to Choose Topics

After running the research, you need to wait about 5–10 minutes: during this time the model collects sources, synthesizes ideas, and usually produces a table with priorities. But the work doesn't end there.

The author suggests first writing down the topics you liked, then checking through Yandex search how oversaturated they are, and only then opening a new regular chat. In the second session, you can already feed the model the report, audience description, and ask it to rank ideas specifically for your task. This two-step process is necessary to avoid confusing research and editorial decision-making.

First, Deep Research answers the question of where there is potential interest and what arguments support it. Then a separate chat helps narrow down the selection for a specific product, blog, or channel.

The author also points out the block below the table where the model explains why the topic is worth pursuing: it often contains not only reasons for virality, but also material for a future article plan, headline, and angles of presentation.

What This Means

Deep Research is gradually becoming not just a feature in chatbots, but a working tool for editors, marketers, and authors who need quick content research without a week of manual preparation. But the model itself doesn't replace the editor: it helps assemble a field of options and evidence, while the final choice still depends on understanding your audience, platform, and how truly new the topic is for your market.