Google Gemini tips: how to prompt the AI assistant for better answers

Why Gemini works best when it is set up first

Google Gemini is the company's AI assistant, a chatbot that answers questions in plain language and helps with everyday tasks. According to Engadget, it is not the most feature-rich assistant on the market, but its close ties to Google Search and the company's other apps make it a serious rival to ChatGPT and Claude. The site argues that getting good results starts with setup rather than clever questions.

The first step Engadget recommends is connecting the apps a person uses most. This is done through the customisation menu inside Gemini. Once apps are linked, the assistant can pull in information from them and help streamline daily workflows instead of working in isolation.

How to write prompts that get better answers

Engadget's core advice on prompting is simple: specific, detailed instructions with added context perform far better than short, generic ones. The site suggests treating the assistant's personality as flexible. A prompt can tell Gemini to behave like a close friend, a mentor or a professional from a particular industry, and the tone and depth of the answers will change accordingly.

If a prompt is hard to phrase, Engadget recommends the opposite approach: set restrictions instead. Limiting which sources Gemini may draw on, or how it should present its answer, is described as a quick way to tailor responses without writing long instructions.

The guide also highlights a feature called skills. These are prompts that can be saved and then run with a short slash command. Google provides a few premade skills, and users can write their own through the customisation menu. Engadget suggests creating one skill for each role Gemini regularly plays, so switching between them takes a single typed command rather than a rewrite of the prompt each time.

Chat settings that change the depth of a response

Beyond prompting, Engadget points to options that appear when starting a new chat. Clicking the plus icon near the chat bar reveals several settings. The first options let a user upload documents or images for Gemini to work with.

The more consequential choices sit at the bottom of that menu, according to the guide. One is deep research, which instructs Gemini to behave like an authority on a subject and consult multiple sources before answering. Engadget describes this as a good way to get detailed responses on niche topics that are difficult to research through ordinary search. The other is guided learning, a mode framed as a study aid for people learning a new subject, which shapes responses into something closer to a lesson than a plain answer.

Bringing chat history over from other chatbots

Engadget notes that many people use more than one AI assistant, and Google appears to have designed for that reality. Gemini can import chat history from other chatbots. The process involves exporting chat data from the other service, then using the import option in Gemini's settings menu.

Once imported, that history gives Gemini a starting memory to draw from, which the site says helps tailor its output from the outset. For anyone switching assistants or splitting their time between several, this removes the need to re-explain preferences and past projects in every new conversation.

What this means for everyday users

The overall picture from Engadget's guide is that Gemini rewards a little preparation. Linking apps, saving skills for recurring roles and choosing the right chat mode turn a general-purpose assistant into something closer to a set of tailored tools.

None of these steps require technical knowledge. They are all menu options within Gemini itself, and most take only a few minutes to configure. For readers who already use Google Search and the company's apps, the guide suggests that this integration is where Gemini holds its advantage over competitors, and that unlocking it is largely a matter of knowing which settings exist.