Is building a B2B SaaS still worth it as a solo founder on 15 hours a week?

Executive Summary & Verdict

Risky — 5/10. A sensible product attached to a go-to-market plan that does not yet exist. You are not taking a A$12,000 risk here. You are taking a 720-hour risk, and 720 hours is the entire annual output of this business.

The idea passes the hardest test. You can name the buyer, her job title, the manual process you replace and the fortnightly moment the pain recurs. Most SaaS plans cannot.

But the economics of the sale do not work at this price. A$89 per site per month covers your cash costs on three customers, while the contracting income you forgo for those 15 hours a week is roughly A$68,400 a year — about 64 paying clinic sites at A$1,068 each just to match the day rate you already have.

And nothing in your answers says how a clinic hears about you. No list, no audience, no clinical credibility, no referral source. In bootstrapped B2B SaaS that omission, not the code, is the usual cause of death.

Idea Validation

The premise is sound and unusually well specified. A four-site group rostering 22 people in a spreadsheet, then hand-reconciling timesheets every fortnight, is a recurring cost with a named owner. The pain is periodic, administrative and measurable in hours, and the person who feels it can authorise A$89 a month.

The differentiator is the right shape too. "Generic rostering plus the award logic the customer does by hand" is a durable vertical wedge: a horizontal vendor cannot copy it without taking on the liability of getting pay rules right for a segment that is a rounding error in their revenue.

What has not been validated is the thing the plan rests on: you have one data point. Danielle's spreadsheet proves Danielle has a problem. It does not prove thirty other practice managers will pay A$89, that they will change a process they have survived for six years, or that the owner will not say "doesn't our practice management software do rosters?" All three are answerable in conversation, without code, and all three change what you build.

Market & Target Customer

Your market is not allied health in Australia. It is practices large enough to have a rostering problem, small enough not to have bought an enterprise workforce product, employing rather than contracting their staff, and run by someone with authority to change a process. Three to fifteen practitioners is the right band.

What to establish before you build

• How many practices are actually in the band. Build the list by hand from association directories and count it. If the honest number is 400 rather than 4,000, that changes your price and your channel.
• Whether the buyer is the practice manager or the owner. At four practitioners, A$89 a month is often still an owner decision, with different objections and a slower cycle.
• What they already pay for. You need the customer's own words for why their practice management calendar is not a roster. That sentence is your sales pitch.
• Whether support staff are employed or contracted. Where a clinic contracts, your award wedge largely evaporates for that account.

The New Zealand and Singapore interest is a distraction for now: different employment instruments, different privacy regimes, three time zones of support on 15 hours a week.

Competitive Landscape

The spreadsheet is the incumbent, and it is winning. It costs nothing, it is already set up, it never has an outage. Beating free and familiar needs a saving the buyer can state in hours, not a better interface.

Horizontal workforce-management products are better funded and more integrated than anything you will ship this year. Your advantage is precision, not capability, so never let a sales conversation become a feature comparison.

Practice management platforms are the real threat. They already own the calendar and the data, and if one ships a competent roster module your wedge narrows overnight. You cannot prevent that; you can be specific enough that the segment is not worth their roadmap slot, and useful enough that integration beats replacement.

All three are weak in the same places: award interpretation, which they leave to the customer; onboarding, because arriving with the rules pre-built saves the first hour; and responsiveness, which is a real advantage for eighteen months and precisely why you cannot take a holiday.

Business Model & Monetisation

A$89 per site per month is A$1,068 a year, and that number sets how much effort each sale can bear. If winning a clinic takes three calls, a trial, a security questionnaire and a follow-up, that is five to eight hours for A$1,068 — out of 720 hours that also cover all the building, support and admin. A five-hour sale caps your year at roughly 100 to 140 new customers even if you write no code at all.

So choose deliberately: drive the cost of a sale down until it is genuinely self-serve, or raise the price until a multi-hour sale is worth making. A$89 sits in the awkward middle — too cheap to sell by hand, too considered to sell itself.

• Per-site subscriptions. Multi-site groups are worth disproportionately more for barely more effort, so aim at groups from day one.
• Annual prepayment. The most valuable thing a bootstrapper can sell — it turns a churn problem into cash. Offer it to the first customer.
• Paid onboarding and migration. Real work the customer values, and it removes the largest reason a trial stalls.
• Payroll integration. Not revenue at first, but what turns a nice tool into infrastructure and is therefore your biggest retention lever.

Churn deserves more attention than growth in year one, and small clinics churn for reasons unrelated to you — a practice sells, a manager leaves, two sites merge.

Finances & Budget

Software has the unusual property that the cash budget is almost irrelevant and the plan can still fail. Below is the year-one cash requirement from current published list prices, with US-denominated lines converted at an assumed A$1 = US$0.65 — check the rate on the day.

A$8,913 against A$12,000 of savings. On cash this plan is comfortable, which is why the verdict is Risky rather than Reconsider — and why the budget line in your answers is the least important number on this page.

Your fixed software cost is around A$200 a month, covered by three clinics. Your opportunity cost is around A$5,700 a month, needing 64 clinic sites. That gap is the whole story of bootstrapped B2B SaaS, and only distribution closes it.

MVP & What to Build First

With 720 hours, scope discipline is survival rather than virtue. Build the part of the workflow done by hand and hated — here the fortnightly timesheet reconciliation, not the roster itself. Danielle can already make a roster in a spreadsheet; what she cannot do quickly is check that it complies and turn actual hours into something payroll accepts.

In the first version, only

• Spreadsheet import of staff and shifts — a re-entry step kills half your trials.
• A single-site roster with the segment's pay rules applied and every flagged issue explained in plain English.
• A timesheet export in the format your first five customers' payroll accepts. Ask them; do not build a generic integration layer.
• Email notification of a published roster. No app, no chat, no shift-swap marketplace.
• A prominent disclaimer that the tool assists with award interpretation and does not replace the employer's obligation to get it right.

Leave out mobile apps, multi-site views, leave management, shift swapping, SSO, an API and a billing portal. Set a hard ceiling of 120 hours to something a real clinic uses on real data. What you need most urgently is not a finished product but the first evidence that someone other than Danielle will pay.

Go-to-Market & Launch Plan

This is the weakest part of the plan and therefore the most important section in the report. Your answers describe a product and a buyer in detail and say nothing about how the buyer finds out you exist. You have to manufacture a channel before you have a product.

• Hand-built outbound to a list you compile yourself. Two hundred named practices and a short email about the timesheet problem rather than the product. The only channel that works from a standing start — ten a day for a month is a real experiment.
• Danielle as a reference, deliberately. One enthusiastic practice manager in a profession that talks to itself beats any budget you have. Ask for two introductions by name, not a testimonial.
• Where practice managers already gather. Association forums, networking groups and the conference circuit — as a useful person with a free cheatsheet, not as a vendor.
• Adjacent vendors as partners. Bookkeepers and payroll advisers serving these practices already have the relationship.

Not on that list: paid search, outreach automation, a startup-aggregator launch, an SEO plan. Paid acquisition at A$1,068 of annual contract value with no funnel data spends savings you need, and launch sites bring you developers, not practice managers.

Set a launch criterion rather than a date: do not build past 120 hours until three clinics that are not Danielle have said in writing that they will trial it. If you cannot get three, you have learned something valuable for the cost of a few evenings of emails.

Marketing & Growth

For one narrow vertical and 15 hours a week, most marketing advice is written for someone else. Three things move the number.

Be the most useful free resource on one boring topic. A clear, correct, dated explainer on rostering and pay rules for small allied-health practices will be forwarded between practice managers for years. Keep it accurate — being wrong about pay rules in public is worse than being invisible.

Make the reference customer a system. Vertical markets run on word of mouth: at day 30, ask every happy customer for one introduction, every time.

Do less, but weekly. Two hours a week on distribution beats a burst every second month. The part-time failure pattern is six weeks of building and a panicked weekend of marketing.

Pricing Strategy

A$89 per site is not obviously wrong, but it is untested and probably too low for a sale you make by hand. Price is the fastest lever you have.

• Anchor on hours saved, not competitor pricing. Anchoring on a horizontal tool's per-employee price puts you in a comparison you cannot win.
• Charge per site, not per employee. It rewards you when a group adds a location instead of punishing customers for growing.
• Test A$149 before you assume A$89. A 60% higher price with a third fewer customers is strictly better for a solo founder.
• Decide and display whether the price includes GST. Before registration there is no GST in your price; after it, A$89 becomes either A$89 including GST or A$97.90, and quietly changing which you meant is a bad conversation with an existing customer.
• Never discount to close a first sale. An early customer on permanent half price is an anchor you spend two years escaping.

Legal Setup & IP

A sole trader has no ASIC registration or annual review fee. A proprietary company costs A$636 to register and A$342 a year in annual review fees from 1 July 2026 [1][2], and business.gov.au is explicit that a company carries higher set-up and running costs and an obligation to complete an annual review [3].

Three things push this plan towards a company anyway: you are selling software that makes pay-rule assertions, so separating your personal balance sheet from that liability matters; clinic groups prefer to contract with a company; and the R&D Tax Incentive is open only to corporations [4].

If you incorporate, the director ID is not a formality. All directors of Australian companies need one, and if you plan to become a director you must apply before you are appointed [5]. Failing to have one is a strict liability offence under section 1272C(1) of the Corporations Act 2001; the current maximum penalty is 60 penalty units, a fine of A$19,800, and ASIC is prosecuting — two NSW directors were fined A$10,000 each in 2026, part of eleven prosecutions producing over A$40,000 in fines [6].

The documents that actually matter

• Terms of service with a real liability position. Your product tells an employer what a shift costs. The terms must state that the software assists and the employer remains responsible, cap liability and exclude consequential loss — drafted by a lawyer. This is the highest-value professional fee in the plan.
• A data processing addendum and a two-page security summary. Your customers have privacy obligations regardless of size (section 14), and one prepared answer turns a four-week procurement stall into a five-minute conversation.
• IP assignment, a trade mark search, and your own contracting agreement. Absent written assignment, ownership is not automatically yours; search the register before you print anything; and read the IP and moonlighting clauses of your day job first.

Tax, Invoicing & Admin

GST registration is compulsory once your current GST turnover — this month plus the previous eleven — reaches A$75,000, or your projected GST turnover — this month plus the next eleven — is likely to reach it, and you must register within 21 days of exceeding the threshold [1]. At A$1,068 per site a year, seventy sites puts you over.

The export nuance is where SaaS founders get burnt. Exports of services can be GST-free where the recipient of the service is outside Australia and the use of the service is outside Australia, and where the supply is made to a non-resident who is not in Australia when the supply is made [2]. But the ATO is blunt about the limit: the supply is not GST-free if the service is provided in Australia even if the recipient is not in Australia [2]. Separately, under the business-to-business cross-border rules, supplies of intangibles such as services and digital products performed in Australia are not connected with Australia where the recipient is an Australian-based business recipient, or a non-resident acquiring them for their overseas enterprise [3].

Applied to your answers: your Australian clinics are ordinary taxable sales at 10% GST once registered, and your New Zealand and Singapore groups are likely outside Australian GST as non-resident businesses acquiring the subscription for an overseas enterprise. So keep written confirmation of each overseas customer's non-resident business status, and ask your accountant how GST-free and not-connected sales count towards the A$75,000 turnover test on your facts — that decides when you must register.

The R&D Tax Incentive, and why it probably does not help yet

The refundable offset — the version that pays cash to loss-making companies — applies where aggregated turnover is under A$20 million, at the company tax rate plus an 18.5% premium [4]. You must be a corporation [5], eligible R&D expenditure for the income year must be at least A$20,000, and the work must be core experimental activities generating new knowledge with outcomes that cannot be known in advance [6]. You must apply to register within 10 months after the end of your company's income year [7].

Now the honest part. Your own unpaid hours are not expenditure. A founder spending time rather than money will struggle to reach A$20,000 of eligible expenditure at all, and routine development using well-understood techniques is not usually a core R&D activity.

The rest is hygiene: an ABN first, a separate business bank account from the first dollar, and compliant tax invoices once registered. Note that your contracting income and this business are one tax picture — if the software runs at a loss while the contracting is profitable, your structure changes the outcome. Settle that before you incorporate.

Tech, Hosting & Security

The technology choices that matter for a solo founder are the ones that reduce the number of things that can wake you up. Everything else is preference.

• One boring managed platform, one managed database — roughly US$50–70 a month at your sizes [1]. The correct amount of time to spend on your deployment pipeline in year one is a single afternoon.
• Error monitoring from the first deployment — free tier at US$0, paid team tier from US$26 a month [2]. You cannot afford to learn about failures from customer emails.
• A dedicated transactional email provider, around US$15 a month for 10,000 messages [3]. Roster notifications in a spam folder are a churn event.
• Payments through hosted checkout. Domestic cards 1.7% + A$0.30, international 3.5% + A$0.30, plus 2% on currency conversion [4] — so charge the overseas accounts in AUD, or accept that cost knowingly.
• Backups you have actually restored, and Australian data residency, because customers will ask where the data lives.

On security, resist both over-engineering and neglect. You do not need a formal certification to sell to a four-site physio group. You do need the list that actually prevents breaches: enforced multi-factor authentication, no shared logins, dependencies updated rather than deferred, tenant isolation you have tested, no production data in development, and a written incident plan you could follow at 2am.

Regional Compliance & Privacy

This section decides whether you can sell to clinics at all, and your own small size does not get you out of it. The Privacy Act 1988 applies to Australian Government agencies and to organisations with an annual turnover of more than A$3 million, and the OAIC describes the Australian Privacy Principles as the cornerstone of the framework, applying to any organisation the Act covers [1][2]. Small businesses at or below A$3 million are generally exempt — but the OAIC lists categories that must comply regardless of turnover, including private sector health service providers, expressly covering allied health professionals, businesses that sell or purchase personal information, and businesses that have opted in [1].

• Your customers are covered even though they are small. An allied-health practice with two staff is a health service provider with APP obligations regardless of turnover [1]. They cannot hand those obligations to you, so they push them down contractually.
• You may be covered yourself, and you must check rather than assume. That depends on precisely what you hold and do with it, so get it determined on your actual data model, in writing, before you sign a clinic-group contract.
• Opting in is a real option. For a vendor selling to health service providers, voluntarily accepting APP obligations turns your biggest procurement objection into a differentiator [1].

The breach scheme is broader than the A$3 million threshold

The OAIC lists the entities the Notifiable Data Breaches scheme applies to, and it extends well past APP entities: APP entities; small business operators in specific circumstances such as providing health services, trading in personal information or holding Consumer Data Right accreditation; credit reporting bodies; credit providers; TFN recipients, meaning any person holding records connecting a tax file number with an individual's identity; Consumer Data Right entities; and accredited Digital ID entities [3].

The TFN limb is the one most likely to catch a product like yours. If your roadmap moves from rostering into anything storing tax file numbers for payroll — an obvious next step — then on the OAIC's own list you are inside the NDB scheme regardless of turnover [3]: a compliance boundary you would cross by shipping a feature. Where the scheme applies, reasonable grounds to suspect an eligible breach trigger an obligation to assess expeditiously and take all reasonable steps to complete the assessment within 30 calendar days, which the Commissioner expects to be treated as a maximum rather than a target [3]. That is not long for a founder also working four days a week.

Risks & Common Pitfalls

• Distribution — the risk that actually ends solo B2B SaaS. Your answers contain a product, a buyer and no channel, and founders in this position respond by building more. Nothing in section 07 matters if section 08 does not happen.
• A sales cycle longer than your runway of attention. Money is not your runway — 720 hours a year is. A three-to-six-month cycle means first revenue around month six to nine, by which point you have spent 300 hours of evenings with nothing to show. That is where part-time founders quit. Set a month-three milestone you can hit: three signed trials.
• You are the single point of failure. A fortnight of illness in payroll week is a churn event. Publish a support promise you can honour on a Tuesday when your contracting client has a deadline.
• Award interpretation being wrong in public. If the product is wrong and a customer underpays, you have a liability and reputation problem in a small, talkative market.
• Procurement and incumbents. An adviser can stall a deal for a month over an unanswered data question, and a practice management platform that adds rosters does not kill you immediately but it caps you.
• Underpricing and then being unable to fix it. A price chosen before any conversation about value is a guess, and one that is too low is much harder to reverse.

Action Plan: 30 / 60 / 90 Days

Ninety days at 15 hours a week is about 195 hours, and this plan spends more of them on distribution than on code.

Days 1–30 — establish the market, write no product code

• Read the IP and moonlighting clauses in your contracting agreement. Do this first.
• Build the list by hand: 200 named practices in the three-to-fifteen practitioner band.
• Fifteen calls with practice managers who are not Danielle: what they do now, how long the reconciliation takes, what they already pay for, who signs off A$89 a month. Do not pitch.
• Ask Danielle for two introductions, explicitly, by name.
• Decide the structure with an accountant: sole trader versus a company at A$636 to register and A$342 a year [1].

Days 31–60 — set it up properly, build the 120-hour version

• If incorporating: apply for the director ID before appointment [2], then register the company, get the ABN, open the business account.
• Commission the terms of service and liability position from a lawyer.
• Ship version one against the 120-hour ceiling: import, single-site roster with pay rules explained, timesheet export, email notification.
• Stand up the boring stack and restore a database backup once, writing down the steps [3].
• Get the privacy position determined on your actual data model [4][5], and write the security summary buyers will ask for.
• Two hours every week on distribution, including the weeks the build is behind.

Days 61–90 — get three clinics using it on real data

• Three trials live on practices' own rosters, with you doing the migration by hand and charging for it.
• Quote A$149 per site to three prospects before settling on A$89. You need the objection, not the agreement.
• Ten outbound emails a day, four days a week, measured.
• Approach three bookkeepers or payroll advisers about referrals; publish the explainer.
• At day 90, review one number honestly: how many clinics you had never met before day one are now paying you. If it is zero and no trial is close, another 200 hours of building will not fix it. Change the channel or the price, not the feature list.