Pricing is where one-person AI service businesses quietly fail. Not because the math is hard — because founders anchor to the wrong number. Once your production line is running, a monthly deliverable might take you forty-five minutes of hands-on work, and if you let that forty-five minutes anywhere near your price, you will charge $100 for work the client values at $1,000 and wonder why the business never compounds.
The rule that fixes it: price against the client's alternative, never against your effort. A professional buying a productized AI service is not buying your minutes. They are buying a result they were otherwise going to pay staff for — badly, late, and at loaded cost. This guide runs that math for the three markets the one-person AI service model targets, then covers the decision rules that keep the number defensible: retainer versus per-project, the written floor, raising prices with proof, and the compounding second engagement.
The anchor: staff hours × loaded cost, one market at a time
Loaded cost is salary plus benefits, payroll taxes, and overhead — typically 1.25–1.4× base salary. That is the number a firm actually pays for an hour of staff time, and it is the honest anchor for your fee. Work it for each market:
Legal — the regulatory alert. Producing a monthly client alert internally means an associate tracks the rule changes, drafts the summary, builds the affected-client matrix, and writes the sendable versions: call it six to ten hours a month. An associate's loaded cost runs $100+ per hour before you count the real number — the $300+/hr in billable time those hours displace. The firm's internal cost is $600–$1,000 in loaded hours plus displaced billings; a finished alert at $800/mo is priced under the client's cheapest alternative, which is exactly where you want to sit.
Accounting — the monthly client brief. Briefs for eight key accounts, at two to three staff-accountant hours each, is 16–24 hours a month; at a loaded $45–$55/hr that is $720–$1,300 before partner review time. And the staffing to do it may not exist at any price: accounting bachelor's completions fell 7.8% in the 2021–22 academic year (AICPA 2023 Trends), so the junior who would produce those briefs is the hire the firm cannot make. $1,200/mo for finished, on-time briefs undercuts the internal cost and dissolves the hiring problem.
Real estate — listing copy and the market update. The agent's alternative is their own selling time: three-plus hours a month wrestling copy and updates, at whatever an hour of prospecting and showing is worth to a producing agent — more than $500 a month by any reasonable estimate, before counting what polished marketing is worth in a post-settlement market where agents must justify their value in writing (NAR practice changes, effective August 17, 2024). $500/mo is an easy yes against that arithmetic.
Present the anchor in the sales conversation exactly this way. Not "this takes me an hour" — never that — but "your associate spends eight hours on this; here it is finished, sourced, on the fifth of every month, for less than those hours cost you."
Retainer versus per-project: the decision rule
The rule is one sentence: if the deliverable recurs on the client's calendar, it is a retainer; if it is a one-off diagnostic, it is per-project — and per-project exists mainly to become a retainer.
Retainers are the business. Rules change monthly, books close monthly, listings turn continuously — so the same production line runs every month at near-zero marginal effort while the fee recurs. A retainer also buys the client the thing they value most, reliability, and buys you the thing you need most, revenue you can plan on.
Per-project pricing has exactly two legitimate jobs. First, the paid sample or one-off audit at $150–$500 — the wedge engagement that lets a cautious buyer transact small before committing monthly. Second, genuine one-time builds (an outreach system, a farm campaign), which should still carry a monthly run-and-refresh fee attached. If you find yourself quoting per-project prices for recurring work, you are converting an annuity into a lump sum at a discount — stop.
| Offer | Market | Price | Structure | Anchor: what it replaces |
|---|---|---|---|---|
| Monthly Regulatory Alert Service | Legal | $800/mo | Retainer | 6–10 associate hours/mo at loaded cost, plus displaced billables |
| Client-Ready Monthly Brief | Accounting | $1,200/mo | Retainer | 16–24 junior hours/mo the firm cannot hire for |
| Listing Copy + Market Update | Real estate | $500/mo | Retainer | 3+ hours/mo of the agent's own selling time |
| First paid sample / one-off audit | Any | $150–$500 | Per-project | The wedge that becomes the retainer |
These prices are floors and starting points, not ceilings — the standard offers from the AI Service Business Kit's playbook, which also packages the positioning and outreach around them.
The floor rule
Decide, in writing, before any sales conversation, the number below which an engagement is not worth your time — then never negotiate beneath it live on a call. The floor exists because discounting pressure always arrives mid-conversation, when you are most motivated to say yes, and a number you set in advance is the only one that holds. If a prospect cannot meet the floor, offer the smaller per-project wedge instead of a smaller retainer: shrink the scope, never the rate. A discounted retainer misprices you for the entire life of the engagement and — worse — becomes the reference price for every referral that client sends.
Raising prices with proof
Your first client prices you as an unknown. Every delivered month changes that, and your pricing should track it: re-price new clients 20–40% higher once you have a stack of real delivered samples and a testimonial or two. Grandfather early clients at their rate — they took the risk — or raise them modestly at renewal with notice. The mechanism is simple and honest: a prospect who can hold six months of real, sourced, on-time deliverables is buying certainty, and certainty commands a premium the first client never paid for. If you have raised prices twice and close rates have not moved, you are still underpriced.
The second-engagement compounding rule
The most reliable revenue increase in this business is not a new client — it is the second service sold to an existing one. A client who has trusted one monthly deliverable for two or three months buys the second with almost no sales effort: the alert firm adds an intake-response service, the CPA firm adds tax-planning memos (demand structurally dated by OBBBA's July 2025 §174 amendment window), the agent adds a farm campaign. Two services per client roughly doubles the retainer at zero acquisition cost — that is how an $800/mo client becomes a $2,000/mo client. Deliver flawlessly for two months, then propose the add-on; the menu of what to propose is in the nine productized services guide.
What NOT to do
Hourly billing. Hourly pricing punishes your efficiency: every improvement to your production line cuts your invoice. It also invites the client to audit your minutes instead of judging your deliverable. You are selling the alert, not the hour.
Per-token or cost-plus pricing. Pricing off your AI bill anchors the client to your input costs — pennies — and starts a conversation about margins you never need to have. The client's alternative costs hundreds of dollars in staff hours; that is the only cost that belongs in the conversation.
Unlimited revisions. "Until you're happy" is an unbounded liability on a fixed fee. Cap it: one revision round per deliverable, with a defined quality bar — sourced facts, delivery-day promise kept — doing the work revisions otherwise absorb.
Negotiating scope and price at once. When a prospect pushes, hold rate and move scope: fewer accounts in the brief, a quarterly instead of monthly cadence, the wedge project first. One variable per negotiation.
When $300/mo is the right price
There is one phase where a low anchor is correct, and pretending otherwise would be dishonest: the proof phase. Your first one or two clients in a new niche are buying an unproven service from an unknown provider — and you are buying something from them too: the delivered samples, the testimonial, and the referral that reprice everything after. A $300–$500/mo opening retainer for an early realtor client, or a deliberately modest first-month rate for a first CPA engagement, is not underpricing — it is paying for proof with margin you will recover at the next signing, on schedule, per the 20–40% rule above.
Two boundaries keep the proof phase from becoming a trap. It applies to your first two clients per niche, not your first ten. And the price rises at a named milestone you set in writing on day one — two delivered months, or the testimonial in hand — not "someday."
Get the sequence right and pricing takes care of itself: pick a niche where demand carries a date — the free Sub-Niche Opportunity Report scores 25 of them across legal, accounting, and real estate with the fee range printed in every row — anchor the fee to the staff hours replaced, floor it in writing, prove it cheap once, and raise it on evidence forever after.