Playbooks · Everyday tasks · last verified 2026-10-23

Draft a simple business plan outline for a side business

For an aspiring founder × drafting a one-page business plan outline for a side business × with Gemini, in plain chat — numbers you did not supply stay marked as questions.

Before you start

You need four things written down before you open the chat: what you will sell or offer, who it is for, how you would reach those people, and what it costs you to start (only costs you already know — rent of tools, materials, fees you have actually priced). If you do not know a cost, write UNKNOWN. An outline built on invented numbers looks finished and is not.

Time and cost: NOT_MEASURED. No revenue, profit, or success claim is made here, and no market size is stated.

Do not paste bank details, tax IDs, or other people's personal information into the chat. A plan outline needs your offer and your audience, not your accounts.

What "done" looks like

At the end of one sitting you have a one-page outline with eight short sections (below), every number in it either supplied by you or marked as a research question, and a list of the three facts you most need to check before spending money. You have not registered, priced, or launched anything — the outline is the result.

Steps

1. State the business in two sentences. Paste:

   I want to start a side business. What I offer: [one sentence].
   Who it is for: [one sentence]. How people would find it:
   [one sentence]. Help me draft a simple business plan outline.
   Use only the facts I give you in this chat. For any number
   I do not give you — prices, costs, market size — write
   RESEARCH NEEDED, not an estimate.

2. Ask for eight sections, one paragraph each. Paste:

   Draft the outline in eight sections: (1) what I offer,
   (2) who it is for, (3) why someone would choose it over
   what already exists, (4) how I reach customers, (5) what
   it costs to start, using only costs I gave you, (6) how
   I charge, if I stated a price — otherwise RESEARCH NEEDED,
   (7) what I do in the first month, (8) what could go wrong.
   Keep each section to one short paragraph.

3. Turn every RESEARCH NEEDED into a concrete next check. Paste:

   For each RESEARCH NEEDED item, write one sentence saying
   exactly where I would check it — a supplier quote, a
   competitor's public price page, a local rule, or a person
   I would ask. Do not answer the question yourself.

   This step is what separates an outline from a wish list: every gap gets a named check, not a guessed number.

4. Cut anything you would not defend. Read the outline and delete any claim about customers, competitors, or demand that you cannot back up from something you know. A shorter outline you believe is worth more than a longer one you do not.
5. Copy the outline out of the chat. Save it in your notes or documents. It is now the document you take to the research checks in step 3 — not a plan to act on until those checks are done.

Check the agent's work

• Every number in the outline is either one you supplied or marked RESEARCH NEEDED. A number with neither origin is a number to delete.
• The "why someone would choose it" section names something specific about your offer — not a generic claim about quality or service.
• The first-month section lists actions you can actually do with the time and money you have, not hires, ads, or premises you have not budgeted.
• The "what could go wrong" section includes at least one risk you recognise from your own situation, not only generic risks.
• Nothing in the outline states a legal, tax, or licensing requirement as fact. Those are research checks, and the outline should label them as such.

What can go wrong

• It invents market size or revenue. This is the most common failure in this task. The RESEARCH NEEDED rule is the guard; if a number appears without a source you gave, treat it as invented.
• It writes a plan for a bigger business than yours. Hiring, premises, and ad budgets creep into generic plans. The first-month check — actions you can actually do — catches this.
• It flatters the idea. A draft that only lists strengths is not a plan. If the "what could go wrong" section reads as thin, ask for three specific risks for a business of exactly this type and size.
• It states a rule as fact. Licensing, tax, and food-safety rules vary by place and activity. A chatbot's statement of a rule is a lead to check with the official source, never the check itself.

Limits

• Gemini's chat is a drafting tool in this playbook. It cannot register a business, check a licence, price your market, or confirm that customers exist for your offer. Those are your research steps.
• No financial advice is given or claimed here. Pricing, costs, and whether the business is viable are decisions you make from the checks in step 3, not from the outline.
• No file upload or web search capability is claimed on this page. The outline is built entirely from facts you type into the chat; if you choose to upload a file in Gemini, its help page describes that separately, and this playbook's steps do not depend on it.

Verification

Checked 2026-10-23. This page was not run end to end; its capability claims rest on provider documentation fetched this run. The only product capability this playbook relies on is conversational drafting in a chat — typing a request and refining it with follow-ups — which is the ordinary interaction described across the provider pages consulted: Google Gemini Apps Help — Upload & analyze files in Gemini Apps (Gemini Apps as a chat product you ask for answers, summaries, and insights about content you provide; file upload is available but not used by this playbook's steps); OpenAI — Research with ChatGPT (chat follow-up pattern, consulted for the general chat interaction and set aside where this playbook does not use search). Omitted: no claim is made that Gemini can register a business, verify licences, size a market, or give financial advice — those steps were cut because they are not grounded in the fetched pages, and the playbook says so in Limits. Limits of this verification: provider pages describe the chat product, not whether any given business idea is viable — that check is the "Check the agent's work" section above and your own research checks.

Back to: Everyday playbooks hub · Skills index