Online Bookkeeping Service

The shortcut: Don't compete with Bench's $300/month — niche down to e-commerce or SaaS founders and price $600-$1,500/month. Generalists who try to win on price hit a $5K MRR ceiling fast and never get past it.

Industry: Finance & Insurance | Investment level: Micro — $500-$2,000 | Time to launch: 4-8 weeks (QuickBooks ProAdvisor cert + 2 sample client files + LLC + first paid client gate the launch) Best for: Former corporate accountants, AP/AR clerks, payroll processors, and detail-obsessed spreadsheet people who can categorize 400 transactions a month without zoning out and sit on a 30-minute Zoom with a panicked founder without losing the room. What you'll likely make: $800-$1,800 month 3, $3,000-$5,500 month 6, $7,000-$12,000 month 12. Math is in Section 4.

Market Opportunity

Walk into a small business owner's QuickBooks Online file two years in and you'll almost always find the same thing: a category called "Ask My Accountant" stuffed with hundreds of transactions nobody touched, a bank account last reconciled in a different fiscal year, and a Stripe integration that's been silently duplicating deposits since March. The founder knows it's a mess. They've just decided not to look. Then the seed investor asks for "a quick look at burn" before next Tuesday's call, and now they need someone yesterday. That person — the panicked founder with a broken file and a deadline — is your buyer.

The U.S. has roughly 33 million small businesses per the SBA Office of Advocacy, and bookkeeping software market spend is projected to hit roughly $6.9 billion by 2030 per Grand View Research. The buyer pool isn't shrinking — it's growing because every Stripe, Shopify, Amazon, and Gusto integration creates one more thing the founder doesn't want to learn.

The trap is competing with Bench Accounting at $299-$499/month or Pilot at $499-$849/month. Bench uses proprietary software with no QuickBooks Online (QBO) handoff — a deal-breaker for any founder who'll one day raise capital. Pilot is built for VC-backed startups. Your wedge is the independent specialist who picks one niche and runs $600-$1,500/month retainers with real human responsiveness, QBO or Xero data the client owns, and a monthly close the client can hand to a CPA at tax time.

Launch With AI

Pro section. AI doesn't replace the bookkeeping — it cuts the parts that drained you (writing discovery → SOW, drafting catch-up project quotes, generating monthly close emails, building CPA referral pitches). Spend the saved time on what AI can't do: walking the panicked founder through their 'Ask My Accountant' category at 4pm Wednesday, calling the bank when a Stripe duplicate hits, and being the bookkeeper the CPA actually trusts to deliver clean books at tax time.

The trap most first-year bookkeepers fall into: they think AI will categorize transactions for them. It will categorize correctly 80% of the time + miscategorize the 20% that matter — Stripe deposits as 'sales' when they're really 'sales + Stripe fees + refunds + chargebacks split.' AI for the writing tail (SOWs, monthly close emails, catch-up quotes, CPA outreach) AND for QBO bank rule generation is fine. Every reconciliation + every chart-of-accounts decision + every tax-relevant categorization gets your eyes on it before it goes to the client.

Important up-front: AI cannot reconcile the bank, sign tax returns (you can't either without an EA/CPA cred), or refuse the distressed-books client without a catch-up fee. It will also confidently generate "I save you 15 hours/month" claims you can't substantiate. You set the chart of accounts, the close procedure, the scope-of-practice line; AI scales the writing around them.

AI Tools You'll Use

Tool	Price	What it does
ChatGPT Plus	$20/mo	Discovery → SOW, monthly close emails, catch-up project quotes, CPA outreach
Claude Pro	$20/mo	Long-context QBO file review (paste 12 months of transactions → variance analysis)
QBO + Xero built-in AI	included with subscription	Auto-categorization, bank rule suggestions, anomaly detection, recurring transaction matching
Keeper AI	$16-$50/mo	Workflow automation, document chase, client portal, monthly close checklist
Loom AI	free	Monthly close video walkthroughs, catch-up project handoff videos

The Workflow

Discovery → fixed-fee SOW + catch-up quote (ChatGPT, ~30 min/prospect). Distressed-books prospects = the highest-margin work, but only with a catch-up fee. Paste:

"I just had a discovery call with a [Shopify e-commerce founder, 18 months behind on books, $1.2M revenue, 1 bank + 1 credit card + Stripe + Amazon]. They want monthly bookkeeping going forward. Build the proposal: (a) the catch-up project quote (fixed fee $2,500-$4,000 paid 50% up-front BEFORE I open the file — 18 months × ~$150/mo of cleanup work), (b) the ongoing retainer (Standard $700/mo or Growth $1,200/mo — I recommend Growth for e-com because of Shopify + Amazon + Stripe reconciliation), (c) the explicit out-of-scope (NEVER tax advice, NEVER signing returns, I refer to my CPA partner — name them), (d) the engagement letter (50% catch-up deposit + monthly retainer auto-debited via Stripe ACH at $5/transfer flat — saves $360/year vs cards on a $1,000 retainer), (e) the deliverables (clean monthly close by the 12th + P&L + balance sheet + 30-min review call + a SimplePractice/Keeper portal for document chase). Tone: senior bookkeeper. NEVER 'I'll get your books in shape' eager filler. Output as Notion-ready 1-pager."

Send within 4 hours of the discovery call. The catch-up + retainer combo is what makes distressed-books clients profitable — without the catch-up fee, the first 6 months lose money.
Long-context QBO file analysis (Claude Pro, ~60 min/new client). New clients drop a 200-page QBO export. Paste it into Claude:

"Below is the QBO export for [client] for the past 12 months. Sources: 1 bank, 1 credit card, Stripe (e-commerce), Amazon FBA. Generate the 1-page audit: (a) the 5 categorization errors that need cleanup (e.g., 'Stripe deposits booked to Sales — should be Sales + Stripe Fees + Refunds split via Stripe-to-QBO sync rules'), (b) the 3 unreconciled bank-feed gaps (specific months + the variance), (c) the 8 recurring transactions that need bank rules created (recurring SaaS subscriptions categorized inconsistently — Slack, Asana, Notion, etc.), (d) the suggested chart of accounts cleanup (the 4 duplicate accounts to merge, the 3 obsolete accounts to deactivate), (e) the 1-line CPA-ready summary ('books are 87% accurate by transaction count; 13% need cleanup before tax-ready'). NEVER hallucinate transaction amounts; flag any number for verification against the actual export. Tone: senior bookkeeper. Output as Notion-ready 1-pager."

Verify every cleanup item against the actual QBO before executing. Claude long-context turns a 6-hour QBO orientation into 60 minutes — that's billable hours back to monthly close work.
QBO + Xero bank rule generation (QBO/Xero AI + ChatGPT, ~30 min/client). Bank rules categorize 80% of transactions automatically. Most bookkeepers don't build them. For each new client, paste:

"I'm setting up bank rules in QBO for [Shopify e-com client]. Their recurring transactions: [paste — 30 most common transactions including amount, payee, current category]. Generate the bank rule recommendations: (a) for each recurring transaction, the rule (if amount = X AND payee contains 'Y' THEN categorize as Z), (b) the conditional logic for tricky ones (Stripe deposits = split between Sales / Stripe Fees / Refunds via the Stripe-to-QBO sync), (c) the recurring rules to set up (monthly SaaS like Slack $79.99 → Software Subscriptions, monthly Klaviyo $45 → Marketing Software), (d) the 'review monthly' rules (any transaction >$500 from new payee = flag for human review). Output as a QBO bank rule list I'll execute. NEVER auto-categorize anything labeled as 'review' without my eyes on it."

Bank rules cut monthly close time from 8 hours to 3 hours per client — at 8 retainer clients, that's 40 hours/month of recovered time.
Monthly close email + Loom video walkthrough (Loom AI + ChatGPT, ~20 min/client/month). Most bookkeepers send the P&L as a PDF attachment. The Loom walkthrough is what makes the client refer 2-3 friends. Record a 4-min Loom each month, then paste the auto-summary:

"I just finished the monthly close for [client] — P&L + balance sheet ready. Below is the Loom AI auto-summary of my walkthrough video. Generate the email: (a) the headline (the 1 number that matters this month — '$Y revenue, $X net profit, +12% MoM'), (b) the 3 things I noticed (the AR aging bump, the unusual Stripe refund pattern, the recurring SaaS bill that shifted), (c) the 1-line ask for the client ('please confirm the $4,200 transaction on March 15 was the new MacBook for Sarah — I categorized it as Computer Equipment'), (d) the Loom video link, (e) the next-month-close date. Tone: senior bookkeeper transparent + collaborative. NEVER 'books are closed!' filler. Output as paste-ready email."

Send the first business day of every month. Monthly Loom walkthroughs drive 25-40% of new-client referrals — clients tell their founder friends about the bookkeeper who actually explains things.
CPA referral partnership outreach (ChatGPT + Canva, ~90 min one-time). One CPA referral relationship = 3-6 client referrals/year. Paste:

"I'm a niche bookkeeper specializing in [Shopify e-commerce / SaaS founders]. Build the 1-page outreach to 5-10 local CPAs: (a) the email I'll send ('CPAs hate doing bookkeeping. I send you tax season clients with clean books, you send me their monthly bookkeeping work — let's grab coffee'), (b) the value proposition I bring (clean monthly closes by the 12th + QBO/Xero handoff + scope-of-practice respected — I never sign returns, I never give tax advice, I always route tax questions back to you), (c) the partnership terms (no referral fee in either direction — the relationship is based on mutual clean handoff), (d) the post-coffee follow-up email I send within 24 hours, (e) the 1-pager I leave (my niche + my packages + my CPA-ready monthly close deliverable). Tone: senior bookkeeper respectful of CPA authority. NEVER 'I'd love to help your clients' eager filler. Output as Canva-ready 1-pager + email templates."

Drop into Canva. Email 5-10 local CPAs. One active CPA relationship × 3-6 referrals/year × $700-$1,200/mo retainer = $25K-$86K of annual revenue per CPA partner.

Time Saved Per Week

Roughly 6-9 hours/week once your SOW template, bank rules, and monthly close workflow are built:

Discovery → SOW: 4 hours/prospect → 30 min (ChatGPT)
New-client QBO orientation: 6 hours/client → 60 min (Claude long-context)
Bank rule setup: 4 hours/client → 30 min (QBO AI + ChatGPT)
Monthly close emails + Loom: 60 min/client → 20 min (Loom AI + ChatGPT)
CPA referral outreach: 6 hours up-front → 90 min (template + Canva)

Trade that time for: 5 cold messages to founders in your niche, the second QBO ProAdvisor cert (Advanced), and the third CPA coffee meeting.

Total AI Stack Cost

Budget tier ($20/mo): ChatGPT Plus only. QBO + Xero built-in AI included with subscription; Loom AI free; Canva free covers signage. Right for first 90 days.
Full tier ($86/mo): ChatGPT Plus + Claude Pro + Keeper + Loom AI free. Worth it once you cross 4 retainer clients — Claude long-context on QBO file analysis alone saves 5 hours per new client.
Compare: A part-time bookkeeper doing categorization + monthly close + client comms runs $1,500-$3,500/mo. The full AI stack is one-thirtieth that cost — and the senior-bookkeeper voice your CPA partners trust stays yours.

NEVER use AI to give tax advice or sign returns. Stay inside scope-of-practice (bookkeeping only). Route all tax questions to your CPA partner.

Cancel anything you don't open in a 7-day window. Pick QBO OR Xero primary (not both — pick QBO for US, Xero for international/Shopify-native).

Your First Win

30 minutes from now your fixed-fee catch-up + ongoing retainer SOW template is built. Open ChatGPT (free tier works). Paste:

"I'm a niche bookkeeper specializing in [Shopify e-commerce founders / SaaS founders / professional services firms]. Build me the EXACT fixed-fee SOW template I'll send EVERY new prospect within 4 hours of discovery call. Cover: (a) the catch-up project quote (fixed fee based on months-behind × $100-$200/month + the 'paid 50% up-front BEFORE I open the file' clause — this is non-negotiable for distressed-books work), (b) the ongoing retainer (Basic $400/mo, Standard $700/mo, Growth $1,200/mo — recommend the right tier based on transaction volume + integration count), (c) the explicit out-of-scope (NEVER tax advice, NEVER signing returns, I refer to my CPA partner [name]), (d) the engagement letter (50% catch-up deposit via Stripe + monthly retainer auto-debited ACH $5 flat saves $360/year/client vs cards), (e) the deliverables (clean monthly close by the 12th + P&L + balance sheet + 30-min Loom walkthrough call + a Keeper portal for document chase). Tone: senior bookkeeper. NEVER 'I'll get your books in shape' eager filler. Output as Notion-ready 1-pager + paste-ready engagement letter clauses."

Use on every prospect going forward. The catch-up + retainer combo is what makes distressed-books clients profitable — at 50% deposit × $3,000 catch-up = $1,500 in the door BEFORE I touch the file + $700/mo retainer = $9,900 of year-1 revenue per distressed-books client. That single SOW is worth your first $50K of pipeline.

Product / Service Offering

You're selling three packages, all built on the same core deliverable — a clean monthly close — sized to transaction volume:

Basic monthly bookkeeping. 1 bank, 1 credit card, under 150 transactions/month. Monthly reconciliation, P&L by the 12th. $300-$500/month. 3-5 hours.
Standard package. 2-4 accounts, 150-400 transactions, P&L + balance sheet + 30-minute review call. $500-$900/month. 6-10 hours.
Growth package. Add A/R aging, A/P tracking, payroll coordination with Gusto, Shopify or Stripe reconciliation. $900-$1,500/month. 10-18 hours.
Catch-up project. Distressed books 6-24 months behind. Fixed fee $1,500-$4,000, paid 50% up front. Required before any new retainer starts.

Pick one niche and own it. The two best: e-commerce sellers (Shopify, Amazon FBA — inventory COGS, sales tax, platform-fee reconciliation) and SaaS founders (deferred revenue, MRR reporting, Stripe-to-QBO sync). Two tools you'll live in: QuickBooks Online (~80% US small-business market share) and Xero (strong with Shopify-native founders). Keeper at $16-$50/month replaces the email chaos most solos drown in by client number five.

Revenue Model

Unit economics for one solo bookkeeper working from a laptop, no employees, software subscriptions only:

Service	Price	Variable cost (tools + payment fees)	Your time	Take-home per engagement
Basic monthly bookkeeping	$400/mo	$20 (QBO Accountant prorated) + $5 (ACH)	4 hrs	~$375/mo
Standard monthly retainer	$700/mo	$25 + $5	8 hrs	~$670/mo
Growth retainer (e-com or SaaS)	$1,200/mo	$35 + $5	14 hrs	~$1,160/mo
Catch-up project (12 months behind)	$2,500 fixed	$40 + $73 (card)	25 hrs	~$2,385 one-time

Your first $1K month = two basic clients at $400 plus one $300 catch-up sprint, or one standard at $700 + one basic at $400. About 12-15 hours, doable alongside a day job.

Your first $3K month = three standard retainers at $700 + one growth client at $1,200, or four standards. Roughly 35-45 hours/month — the threshold where you should screen new prospects against your niche, not take anyone who'll pay.

The retainer is the engine. Six retainers averaging $800/month = $4,800 MRR before any catch-up revenue. By month 12, target 8-10 retainer clients plus 1-2 catch-up sprints per quarter — lands $7,000-$12,000/month gross. The signal to upsell: when a client asks "can you also handle the Stripe-to-QBO mess?" — move them from basic to growth, don't write a free task into the existing retainer.

Bill retainers via Stripe ACH at $5/transfer flat versus 2.9% + $0.30 on cards — a $1,000 retainer costs $5 via ACH, $30 on card, saving about $300/year per client Stripe pricing. Set up ACH on day one.

Startup Costs

QuickBooks Online Accountant + ProAdvisor program: free through the Intuit ProAdvisor program. Passing the Advanced ProAdvisor exam (10-20 hours, free practice tests) lists you in the public Find-a-ProAdvisor directory — real inbound with zero ad spend. Renewal is annual.
Xero Certified Advisor: free through Xero's partner program. 4-6 hours self-paced. Worth doing in month 2 if you're targeting Shopify or international founders.
LLC + EIN: $35-$500 LLC filing depending on state — LLC University 50-state table. Apply for your EIN free at IRS EIN Online — never pay a third party.
E&O + bookkeeper bond: $500-$1,500/year via Hiscox or Insureon for $100K-$300K coverage. Most bookkeepers bundle a $10K-$25K fidelity bond into the same policy for an extra $50-$150/year.
Workflow + document portal: Keeper ($16-$50/month) once you have 3+ clients. Skip in months 1-2 — Google Drive + a shared client folder is fine for two clients.
Sample client files: the highest-leverage 15-20 hours up front. Build 2 anonymized clean monthly closes (real or mock) you can show on a discovery call. Without them, founders won't trust you with their actual books.
Google Workspace: $7-$14/user/month. Almost all client document review happens here.

Realistic all-in: $500 if you self-source samples, defer Keeper to month 4, use library or trial QBO access, and bind E&O at the lower end. $2,000 if you bind E&O up front for $1M coverage, subscribe to Keeper from day one, pay an attorney to review your engagement letter template, and pre-pay a year of Xero certification CE.

Legal & Formation

Business entity. Single-member LLC the moment you sign your first paid client. Separates your personal accounts from a client who claims a missed sales tax filing "cost them" $8,000 in penalties. Sole prop is acceptable for the first 30 days — switch before client number two. Get your EIN free directly from the IRS; the $50-$300 "EIN filing services" are reselling a free five-minute form. The S-corp election is worth the math once net profit clears roughly $80K-$100K/year — file IRS Form 2553. Most niche bookkeepers hit that in year two or three.

Licenses, credentials, and sales tax. No federal license is required to be a bookkeeper, and the IRS does not require a PTIN (Preparer Tax Identification Number) unless you're preparing tax returns for compensation IRS PTIN FAQ. The credentials that matter: QBO ProAdvisor (Advanced), Xero Certified Advisor, and — once you have 2+ years of experience — the EA (Enrolled Agent) from the IRS if you want to expand into tax. EA requires passing all three SEE (Special Enrollment Examination) parts at $206/part plus 24 hours/year CE (2 hours ethics). State sales tax: most states exempt professional services, but Texas, Hawaii, New Mexico, South Dakota, and West Virginia tax most services broadly, and Hawaii's GET applies at 4-4.5% on gross receipts. Verify at Avalara's services tracker before setting retainer pricing — building tax in is easier than asking three months later.

Industry-specific risk. Three traps end bookkeeping practices early. First, scope-of-practice creep into tax filing. A bookkeeper without an EA, CPA, or attorney credential cannot sign tax returns for compensation, cannot perform audit or attest work, and cannot give tax advice. The moment a client asks "so can I deduct this?" you're at the line. The right answer is "I'll route that to your CPA" — get one before client number three. Crossing the line creates IRS Circular 230 exposure and personal liability E&O won't cover. Second, data security. You'll have read or write access to bank feeds, payroll, and sometimes bill-pay. Use a password manager (1Password Business, ~$8/user/month), enable two-factor authentication on every QBO and Xero account, and never store client tax IDs or bank credentials in email or unencrypted spreadsheets. A one-page written security policy plus encrypted file transfer answers 95% of any SOC 2 framework question from a buyer. Third, AICPA SSARS No. 21. If you issue financial statements handed to a bank, lender, or investor, SSARS No. 21 classifies that as a CPA "preparation engagement" — which non-CPAs cannot perform. Add a one-line disclaimer on every statement: "Prepared by a non-CPA bookkeeper. Not GAAP-compliant, no assurance provided." Washington State under RCW 18.04 effectively requires a CPA license for write-up work — check your state's CPA practice act before taking a client whose lender will see your reports.

Marketing & First Customers

Your first 5-10 clients come from three channels, and none is cold-emailing every nonprofit in your zip code:

QBO Find-a-ProAdvisor directory. This is the highest-converting passive channel in the niche. Complete the Advanced ProAdvisor certification, fill out your profile with city + niche + 3 services, and ask 5 existing or pilot clients for written reviews in the first 60 days. ProAdvisors with 5+ reviews and the Advanced badge appear above the fold for local searches. Expect 2-4 inbound inquiries/month in a metro of 500K+ once your profile is mature, with a 30-50% close rate on prospects who already use QBO. Directory link.
Indie Hackers, founder Slack groups, and SaaS or e-commerce communities. Pick one and post weekly — not pitches, but plain-English answers to bookkeeping questions. Indie Hackers plus 1-2 niche Slacks (e.g., SaaSHub, eCommerceFuel) at $99-$1,000/year. Expect 1-2 paying clients per quarter at $600-$1,200/month from one well-tended community presence after 60 days.
CPA referral partnerships. Tax-only CPAs hate doing bookkeeping. Bookkeepers shouldn't sign returns. Coffee with 5-10 local CPAs in months 2-3, with a clear pitch: "I send you tax season clients with clean books, you send me their monthly work." A single active CPA referral relationship typically generates 3-6 client referrals per year at higher-than-average price points because the trust transfer is automatic.

Direct outbound works once you have 2-3 case studies. Build a list of 100 niche-fit founders (Shopify shops doing $500K-$5M revenue via SimilarWeb or Apollo at $49+/month) and send a two-sentence cold email referencing one specific issue. Hit rate runs 2-5% with a sharp niche.

First 90 Days

Week 1-2. Form your LLC. Get your EIN free from the IRS. Open a business checking account the same week. Set up Stripe with ACH enabled.
Week 1-3. Start the QBO ProAdvisor Advanced certification. 10-20 hours of study. Pass the exam in week 3.
Week 2-4. Build 2 anonymized sample monthly closes — one from a real or mock e-commerce client, one from a SaaS or service client. These are your sales asset. Without them, founders will not hire you.
Week 3-5. Bind E&O + bookkeeper bond ($500-$1,500/year). Draft your engagement letter (scope, deliverables, due dates, payment terms, scope-creep clause). Have an attorney review before client number two.
Week 4-6. Publish your one-page services site with niche, pricing ranges, and the two case studies. Complete and publish your QBO ProAdvisor profile. Apply for Xero Certified Advisor if e-commerce is your niche.
Week 5-8. Coffee with 5-10 local CPAs. Join 1 niche community (Indie Hackers + one Slack). Post your first three educational threads.
Week 8-12. Sign your first 1-2 paying retainers. Target $500-$900/month each. Run the first monthly close on time, on the 12th, with a 30-minute review call. Ask for a written testimonial within 30 days of the first clean close.
End of day 90. 1-2 retainer clients live, ProAdvisor profile published, $700-$2,000 MRR, 1 catch-up project pitched. The catch-up project pitched is the leading indicator of month-12 income — it converts a one-time buyer into a 12-month retainer faster than any other tactic.

Common Pitfalls

Taking distressed books at flat retainer rates without a catch-up fee. A client 18 months behind can easily be 40+ hours of clean-up. Charging $400/month ongoing without a $2,000-$4,000 catch-up project fee paid up front means you lose money for the first 6 months and resent the client every month after. Quote the catch-up as a separate fixed-fee project, payable 50% before you open the file. The cost of skipping this rule: roughly $3,000-$5,000 in unbilled labor per distressed client — more than three months of normal retainer revenue.
Scope creep into tax advice. The moment you answer "so can I deduct this?" without an EA or CPA credential, one wrong call creates personal liability your E&O may not cover. A single bad deduction call that triggers an IRS audit can generate $2,000-$8,000 in penalties plus a CPA's $1,500-$3,000 cleanup fee — and the client will blame you. Build a CPA referral partner before client number three and route every tax question to them. The partnership generates inbound bookkeeping clients in return — the right answer professionally and commercially.
Skipping the engagement letter. A one-page engagement letter (scope, deliverables, due-by date, payment terms, what's NOT included) prevents the most common dispute in this niche: "I thought monthly bookkeeping included my taxes." Without one, the average disputed invoice runs $800-$2,400 and most solos eat it to avoid the fight. Less than half a Saturday to draft and send via Bonsai or DocuSign.
Skipping E&O before the first paid client. A bookkeeping error that triggers an IRS payroll-tax penalty for a client generates a claim equal to the penalty plus the CPA's hours to fix it. Average payroll-tax penalty per IRS small business research runs around $845/year per affected client, and the cleanup CPA bill on top is typically $1,500-$3,000. A $100K-$300K E&O policy via Hiscox or Insureon is $500-$1,500/year — less than the take-home from one standard retainer client's first month. Bind it before client number one.

Get your full launch plan — take the free 60-second quiz.