Guide
An AI business plan that is actually about your business
Ask a general chatbot for a business plan and you get a competent-sounding document that would fit any company. The useful version is specific: it names your customer, uses numbers you can check, and ends in work you can run.
Should I raise prices?
A higher price earns more on every sale but may cost some sales. This shows the volume you can lose and still come out ahead.
Example: raising a $50 product with $20 unit cost by 10%
Raising the price to $55 increases monthly contribution by $1,300 at a 5% volume loss.
- New price
- $55
- Contribution per unit, now → new
- $30 → $35
- Units per month after the change
- 380
- Monthly contribution, now
- $12,000
- Monthly contribution, after
- $13,300
- Volume you can lose and still break even
- 14.3%
How this is calculated (v1.0.0)
- new price = price × (1 + increase)
- contribution = (price − unit cost) × units
- break-even volume loss = 1 − old unit contribution ÷ new unit contribution
- Expected volume loss is your assumption; a staged test or a new-customer-only price measures it.
Continuing carries these exact inputs and results into Business Ontology — 3 free runs, no account needed.
Why most AI-written plans read the same
A model with no information about your market can only produce the average of everything it has read. It will invent market sizes, round every margin, and give you a strategy that nobody could disagree with — which is another way of saying nobody could act on it.
The fix is not a better prompt. It is giving the work real inputs and making every claim traceable to something.
Give it the inputs only you have
Before you ask for a plan, write down what you already know: who you have sold to or spoken with, what you charge or intend to, what it costs you to deliver, what you can spend to find customers, and what you are unwilling to do.
- Your website, if you have one, so the work starts from the real business
- Actual prices and costs rather than industry averages
- The channels you can genuinely operate this month
- Constraints: time, capital, skills, regulation
Insist on evidence and arithmetic
Any market claim should carry a source and a date, or be labelled as an assumption. Any number should be arithmetic you can recompute — inputs, formula, result — not a figure asserted in a paragraph.
When a figure is unknown, the honest plan says so and proposes how to find out. That is more useful than a confident invention.
A plan is only finished when work comes out of it
The point of planning is deciding what to do next. Each untested assumption should become a piece of work with a cost, an owner and a result that would settle it — research to run, an offer to put in front of buyers, a campaign to test a channel, or software to build when software is genuinely the answer.
How Business Ontology approaches it
Business Ontology starts from your business rather than a blank template: it can read your website, research a market and keep the sources it used, hold the economics as arithmetic you can change, and then carry out the work that follows — campaigns, outreach, hiring, or building an app when you ask for one.
Nothing external happens without your authorisation, and anything it cannot verify it says it cannot verify.
Common questions
Can AI write a business plan I can take to a lender?
It can produce the document, but the part a lender weighs is your evidence and your numbers. Bring real prices, real costs, and sources for market claims.
Will it make up market data?
A general chatbot often will. Business Ontology keeps research separate from assumption: findings carry the source they came from, and unknowns are stated as unknowns.
Does it do anything beyond writing the plan?
Yes. The plan turns into work you can authorise — research, marketing campaigns, hiring briefs, budget allocation, and app building when that is what you need.
Start with your own business
Describe what you are building and Business Ontology works through it with you — the plan, the numbers behind it, and the work that follows.