How to get started with Google Gemini in your first 30 minutes

Open Gemini on the web or on your phone, stick with that one place for thirty minutes, and run three real prompts that each say what you want back. Typing “hello” teaches you nothing about the product, while a short task with a clear output shape teaches you a lot.

You may have tried this already. You signed in, stared at an empty chat box, typed “hello,” got a friendly paragraph back, and closed the tab because you were not sure what to do next. That was a greeting, not a first session, and this post gives you a better plan.

Where to open Gemini on web and mobile

Gemini is Google’s AI assistant, and you can reach it in a few different places. This first session uses only one of them, so the choice is simple. The rest of this post walks you through a timed path that ends with a saved draft you could actually send.

Pick one place and stay there for half an hour

For this practice session, use the web version if you can, because the larger screen makes files and copying text out much easier. Use mobile only if that is truly where you will work, for example on a commute or for camera tasks. Skip the Workspace side panel for now unless your only allowed access is through work mail, since it deserves its own post.

Before anything else, click your avatar or account menu and confirm which Google account is active. Mixing a personal hobby account with a work browser profile is how people paste the wrong things into the wrong place, and one glance prevents it.

The thirty-minute path

Minutes 0 to 5: land and orient

1. Go to gemini.google.com or open the mobile app.
2. Confirm the account and glance at any plan badge or upgrade prompt. You are not buying anything in this session.
3. Notice where to start a new chat, attach a file, and open settings or activity controls. The activity controls decide what Google keeps, and a later post on Gemini privacy covers them in detail.
4. Write your goal in one sentence on paper or in a notes app, such as “Produce a one-page outline for our weekly review.”.

Minutes 5 to 12: one real task

Pick a task that is real but low risk. Here are a few examples.

• Outline a 30-minute team meeting for a process that already exists.
• Turn rough bullets into a polite status email.
• Explain a concept you must teach someone else tomorrow.
• Summarize a public PDF or a document you wrote yourself.

Then write a prompt that has a shape, instead of a vague wish. A prompt is just the instruction you type into the chat box.

Role: You are a careful assistant for a busy professional.
Task: Create a one-page outline for a 30-minute weekly ops review.
Audience: five people who already know the project.
Constraints:
- 6 sections max
- each section has 2 bullets
- plain language, no hype
- flag anything that needs a real metric from me
Output: markdown headings I can paste into Docs.

This works because each line does a job. The role sets the tone, and the task is a single thing so the answer stays focused. The audience line keeps the draft from turning into generic filler, the constraints force a structure you can scan, and the flag line invites the tool to admit which numbers it does not have.

Minutes 12 to 20: revise once, like a colleague

Do not accept the first draft as finished, because a second pass is where vague claims and wrong numbers get caught. Send one revision request, the way you would ask a coworker to tighten a memo.

Revise the outline:
1) Cut anything vague
2) Replace adjectives with actions
3) Add a 5-minute timebox per section
4) Keep total under 200 words
Return only the revised outline.

If you have a safe sample, such as your own agenda from last week or a public article, paste a short excerpt and ask, “Align the outline to this context without inventing metrics.” Even then, verify anything that looks like a fact, because the tool can sound sure about things it made up.

Minutes 20 to 30: save it outside the chat

1. Copy the final outline into Google Docs, Notes, or an email draft.
2. Add one line at the top, such as “Drafted with Gemini on [date]; metrics not verified.”.
3. Delete the thread, or avoid leaving anything sensitive in it, if you pasted something borderline.
4. Write a two-line retrospective: what worked, and what you will try next time.

A chat thread is a poor place to keep the only copy of anything, because threads get messy and hard to search. Your notes app is boring, and that is exactly why you can trust it.

Prompt patterns that beat “write something good”

Files and images in the first session

If the screen offers an attachment button, start with something you would forward to a colleague without flinching: a public PDF, a document you wrote, or a screenshot of a non-sensitive error. Do not upload customer exports, ID scans, or password screenshots. A later post on photos, screenshots, and PDFs goes deeper on staying safe with images and files.

When you attach a file, ask for a structured answer: “Summarize in 8 bullets, then list 5 open questions, then quote 3 short passages that support the summary.” Structure makes checking the work easier than a smooth essay does, because you can compare each bullet against the source.

Mobile-specific notes

On mobile, voice input is useful for brainstorming while you walk, but the transcript may come out messy. End every mobile session by pasting the good parts into a note. The camera is powerful for whiteboards and packaging labels, and it is also how badge photos and private screens end up in a chat by accident, so aim it on purpose.

What success looks like after 30 minutes

• You have one finished piece outside Gemini, such as an outline, an email, or a checklist.
• You revised it at least once.
• You know which account you used.
• You did not paste anything you would regret in a phishing report.
• You can name the next real task for tomorrow.

Common first-session mistakes

• Only testing with “write a poem about my job,” which is fun but does not build a skill.
• Accepting the first draft without a revision pass.
• Leaving the only copy inside the chat.
• Switching accounts mid-session and losing track of where your data went.
• Treating upgrade banners as emergencies.

Practice for tomorrow (20 minutes)

Repeat the path with a different deliverable, such as a customer FAQ, a project kickoff agenda, or a study plan. Keep the same prompt skeleton so you can compare which constraints helped most.

A second worked example: a status email in one pass

Suppose you need a Friday status note for a manager who skims on a phone. After the outline drill, try this as your second piece of work in a later session.

Write a status email.
Audience: my manager who has 60 seconds.
Inputs:
- Shipped: onboarding checklist v2
- Stuck: waiting on legal for vendor DPA
- Next week: pilot with two internal teams
Constraints:
- under 110 words
- no hype adjectives
- one clear ask
- subject line under 6 words
Return subject + body only.

Then revise with a follow-up such as “Make the ask a yes/no question. Move the blocker to the first bullet.” Copy the result into Gmail or your mail program. That is the whole method: a structured request, one revision, and a human who presses send.

What an empty chat is for, and what it is not for

An empty chat is a scratchpad for thinking. It is not a project tracker or a customer database, and it is not a safe place to keep the only record of a decision. If a thread becomes valuable, move the decision into a document with a date and an owner. If a thread becomes embarrassing, delete what you can under the activity controls and tighten your pasting habits, which the later post on Gemini privacy explains.

You also have a choice between a new chat and a long thread. Long threads carry useful context, and they also carry early mistakes that the model keeps repeating. When the conversation goes sideways, start a fresh chat with a clean brief and paste only the final constraints. That feels wasteful, but it is usually faster than arguing with a confused thread.

Settings worth a two-minute tour

• Theme and text size, if you will work in long sessions.
• Linked apps or extensions (turn on only what you understand).
• Activity and history controls, which you should bookmark for the privacy post.
• The model or mode picker, if you see one (keep the default on day one and change it later on purpose).
• The help or release-updates link, so you know features keep moving.

Do not spend day one customizing ten toggles. Spend day one producing one finished piece of work, because customizing is a week-two problem.

If you get stuck or the answer is fluff

Fluff usually means your prompt had no audience, no length limit, and no definition of done. Add those three things. If the model still rambles, ask, “Return only a bullet list with 7 items, each under 12 words.” Constraints are a kindness to your future self, who has to edit the result.

If the product shows an error, switches models, or refuses a request, note whether you hit a safety boundary, a plan limit, or a temporary outage. Each one has a different fix: rephrase the request, upgrade or wait, or try again later. Do not assume every refusal means you are being scolded.

Quick recap

• Confirm your account, pick one place to work, and run the timed path.
• Give the role, the task, and the constraints, revise once, and save the result outside the chat.
• Start files and photos only with low-risk material.
• Next up is Gemini inside Gmail, Docs, Sheets, and Drive.

Series notes

This is Part 3 of Learn Gemini from scratch. Previous: product basics. Next: photos, screenshots, and PDFs.

Sources

• Gemini app
• Gemini overview
• Gemini Apps Help Center

Merch on Etsy

Open store →

Git Version Control Shirt View on Etsy → Standard Deviation T-Shirt View on Etsy → Git Version Control Shirt View on Etsy → Apache Iceberg T-Shirt View on Etsy →