Two documents most small businesses need, both of them normally a consultant.
The business plan
A structured Australian business plan template — eight writable sections, nineteen subsections, thirty-eight things it needs to know about you — plus AI drafting for the prose.
It writes a different plan depending on who is going to read it. Tick that you are approaching a lender, or a grant body or investor, and the sections change: some are required by one audience and not the other, and a plan carrying the wrong ones reads as a template to the person you sent it to.
It starts from what the system already knows. Your company profile seeds the answers before you type anything, and each answer remembers where it came from — so refreshing from the system later cannot overwrite something you typed over by hand.
It will not invent a number that matters. Eighteen fields are on a
never-fabricate list — ABN, ACN, TFN, director credit score, ATO debt position,
revenue, EBITDA, net profit, MRR, ARR, cash on hand, monthly burn, existing debt,
debt service cover, pre- and post-money valuation, raise amount, loan amount. The
model is refused permission to suggest any of them. Where one is missing it writes
[TO CONFIRM: …] in place and moves on, and the plan shows you a running count
per section of what is still open.
A fabricated ABN or a made-up debt service ratio in a document that goes to a lender is the worst thing this feature could possibly produce, which is why it is designed around not doing it rather than around fluency.
The export is a real Word document — cover page, a live table of contents that Word populates when it opens, and every section in order. A section you have not written yet is not quietly left out; it appears, marked, because a hole in a plan going to a lender should be visible to its author.
Policies get a controlled-document workflow. Draft, submit, approve, revise, with issue numbers, an append-only history and a review date. For controlled documents the owner may not be the approver — a founder can approve their own business plan, but the owner of an information security policy cannot sign it off themselves.
Three templates ship today: the Australian business plan, an ISO 27001 information security policy, and an ISMS scope. Adding another is a markdown file, not a release. We would rather tell you it is three than let you find out.
What it is not: this is a structured template plus drafting, not a financial
modelling engine. It does not build a profit and loss, project a cashflow, compute
your debt service ratio or size your market. Every figure in the output is one you
supplied, one quoted from the template’s own reference tables, or a
[TO CONFIRM].
The marketing plan
This one starts a step earlier than marketing tools normally do — not at the campaign, but at whether you can afford customers at all.
The arithmetic, with your numbers in it:
contribution per month = revenue per customer × gross margin
lifetime value = contribution × months they stay
most you can pay = lifetime value ÷ 4
payback = most you can pay ÷ contribution
most you can bid = most you can pay × lead rate × click rate
Every line shows its working — the actual expression, not a figure in a box. No language model is involved in any of it. If the sum says the most you can pay per click is under two dollars and the going rate is eight, the plan tells you paid search is structurally closed to you, and says why: in an auction the price is set by bidders whose customers are worth several times yours. Nobody selling you advertising will put that on a slide.
Thirteen channels, scored against seven weighted criteria — does it reach the people who decide, cost per customer against what you can afford, how soon it can win work, how well it suits you, how far it can grow, how hard it is to copy, and what the advertising rules allow. Each criterion returns a plain-English reason next to its number, so a shortlist can be argued with rather than merely accepted.
A channel your sector’s advertising rules prohibit is excluded outright rather than scored down, because a weighted average is exactly how a prohibited channel ends up recommended when everything else about it looks good.
Founder hours are priced in. A channel that looks free because it only costs sixteen hours a month is not free.
Read this before relying on the regulatory part. The sector advertising rules that ship — for NDIS, the trades, and AHPRA-regulated allied health — are drafts awaiting legal sign-off, and the product labels them that way on screen rather than in a footnote. Treat them as a prompt to check, not as advice.