LeaveVerdict

1. One-liner

Tells a 15-person New Jersey employer exactly how much protected leave a worker just earned, and proves it.

2. Trend signal — why now?

On 17 January 2026 the Governor signed Assembly Bill 3451. On 17 July 2026 the New Jersey Family Leave Act coverage threshold dropped from 30 employees to 15. It keeps dropping: 10 employees on 17 July 2027, 5 employees on 17 July 2028. Employee eligibility collapsed at the same time — from 12 months and 1,000 hours down to 3 months and 250 hours. The Governor’s office estimates this extends job-protected leave to 400,000 additional New Jersey workers.

Read that phase-in again. In two years a five-person business in New Jersey — a dental practice, a landscaping outfit, a two-location pizzeria — is administering job-protected family leave. These are companies whose entire HR function is the owner’s spouse and a QuickBooks login.

And the tooling floor sits nowhere near them. Tilt, the most SMB-friendly specialist, is “strongest for SMBs (51–200 employees)”. Cocoon charges “an upfront annual per employee per year fee based on company size plus a one-time implementation fee” and is “not currently accepting new business inquiries as we focus on our integration with TriNet.” A comparison of the major HRIS platforms states it plainly: “No HR platform in the comparison monitors employment law changes. A multi-state employer either buys an advisory subscription, retains counsel, or accepts that risk knowingly.”

The newly-covered population is entirely below the incumbent floor. Nobody is selling to 15 people. The law just walked 400,000 workers into the room.

Provenance:

• Signal 1 (demand): NJFLA threshold drops 30→15 employees 17 Jul 2026, then 10 (2027) and 5 (2028); eligibility cut to 3 months/250 hours; ~400,000 newly covered workers — https://www.duanemorris.com/alerts/amendments_new_jersey_family_leave_act_extend_coverage_400000_additional_workers_0226.html and https://www.bsk.com/news-events-videos/sweeping-changes-to-the-new-jersey-family-leave-act — 2026-08-31
• Signal 2 (feasibility): NJFLA regulations N.J.A.C. 13:14-1 are fully published, structured, and machine-readable — including the §13:14-1.14(b) written-guidance duty, the §13:14-1.10 certification rules, and the §13:14-1.6 concurrency worked examples — making the ruleset encodable without proprietary data — https://www.nj.gov/oag/dcr/downloads/FamilyLeaveAct-Regulations.pdf — 2026-08-31
• Signal 3 (economic): Substitutes are priced far above the segment — fractional HR runs “$75 to $250 per hour or $1,500 to $8,000 monthly retainers”, employment attorneys $200–$600/hour, and defending an employment suit costs “$10,000 to $250,000 or more”; meanwhile Tilt’s floor is 51 employees and Cocoon has closed to new business — https://www.hrbponline.com/post/how-much-does-fractional-hr-cost, https://www.novianlaw.com/the-average-cost-to-defend-an-employment-lawsuit/, https://www.hr.software/reviews/tilt, https://www.cocoon.com/pricing — 2026-08-31 Category: Regulatory arbitrage

3. The opportunity

Everyone selling into this space sells tracking. Nobody sells the answer.

The NJFLA does not punish you for bad spreadsheets. It punishes you for getting one specific decision wrong when an employee walks into your office and says “my mother has cancer” or “we’re having a baby in March.” At that moment the owner has to answer questions that are genuinely hard:

• Is this person eligible? (3 months and 250 base hours in the preceding 12 months — a part-timer may or may not clear it)
• Does the company even count as covered? (15 employees “anywhere”, not just in New Jersey)
• Does this run concurrently with FMLA, or on top of it?
• Can I refuse the intermittent schedule they asked for?
• What do I have to give them in writing, and when?

The concurrency question alone is a minefield, and the state has already made an example of someone. A North Bergen manufacturer told an employee who had taken FMLA after childbirth that she could not take another 12 weeks under NJFLA. Wrong. As the alert puts it: “The time permitted under the FMLA, to recover from childbirth, is not a leave that is covered under the NJFLA. Therefore, the use of FMLA time did not exhaust any of the 12 weeks available under NJFLA.” The employer threatened to treat non-return as resignation, later pointed security cameras at her workstation, excluded her from meetings, and fired her. The company was found liable and forced to rewrite its leave policies.

That employer did not need a PTO calendar. It needed someone to say “no — she gets 12 more weeks” in the thirty seconds before it replied to her.

The intermittent-leave trap is worse, because the burden is inverted. To deny an intermittent schedule the employer must show the leave would “cause the employer measurable harm, economic or otherwise, significantly greater than” a consecutive leave — and the regulation states flatly: “The burden of proof in these instances rests with the employer and will be determined by the Division on a case by case basis.” A 15-person employer denying a request without building that record is manufacturing the plaintiff’s exhibit.

The remedies make it expensive: pain and humiliation damages, restoration of benefits, out-of-pocket expenses, punitive damages up to $10,000, and reasonable attorney’s fees — plus all NJLAD remedies including compensatory damages and reinstatement with back pay.

The gap: incumbents sell an absence-tracking system to companies that already have an HR department to operate it. The newly-covered employer has no HR department and one leave request every eighteen months. They don’t want software. They want a verdict, in writing, that holds up.

4. Target market

• Primary customer: The owner, office manager, or bookkeeper at a New Jersey business with 15–50 employees and no dedicated HR person. Medical and dental practices, restaurant groups, contractors, landscapers, auto shops, day cares, small manufacturers, independent retail chains. Revenue roughly $2M–$15M. From July 2027 the same product sells to 10-employee firms; from July 2028, 5-employee firms.
• Why they buy: They just got a leave request and have no idea what they owe. Their alternatives are to guess, to call a lawyer at $200–$600/hour, or to retain fractional HR at $1,500–$8,000/month. The first is how North Bergen happened; the last two cost more per month than this product costs per year.
• Rough TAM reasoning: New Jersey has on the order of hundreds of thousands of employer establishments, and the state itself put the newly-covered worker count at ~400,000. Even a narrow read — tens of thousands of NJ employers in the 15–50 band, expanding sharply as the threshold walks down to 5 — supports a business many multiples past $5M ARR. This does not need market share; it needs a few thousand customers.
• Why now for them: The duty switched on 17 July 2026 and the population doubles twice more, in 2027 and 2028. Most of these employers have never administered job-protected leave in their lives, and the Division on Civil Rights guidance landed alongside the change.

5. Product sketch (MVP)

• Coverage check: answer six questions about headcount (including out-of-state staff) and get a dated determination of whether NJFLA applies to you this year — and which threshold year catches you if not.
• Eligibility verdict: enter a specific employee’s start date and hours; get a yes/no on the 3-month / 250-hour test with the arithmetic shown.
• Entitlement math: the answer that actually matters — how many weeks this person gets, whether FMLA runs concurrently or stacks on top, and how the 24-month window is being measured under the method your company elected.
• Written notice pack: the per-employee written guidance the regulation demands from employers without a handbook, plus a leave-response letter, a compliant certification form, and the current DCR posting — generated, dated, and stored.
• Intermittent-request record: walks the owner through the “measurable harm” test before they answer, and produces the contemporaneous memo that carries their burden of proof if they deny.
• Leave file: every determination, notice, certification, and date kept in one exportable file per employee — the thing you hand a lawyer on day one of a claim instead of reconstructing from memory.
• Threshold watch: tells you the month your growing headcount crosses 15, or the year the phase-in reaches down to you.

6. AI angle — what’s load-bearing

Remove the AI and this is a static calculator that breaks on the second real question.

The load-bearing work is turning a messy human situation into a statutory classification. The owner types “Maria’s been here since February doing about 22 hours a week, her dad’s in hospice and she wants Tuesdays and Thursdays off for a while.” The product has to determine: is hospice care a “serious health condition” under the regulation’s inpatient/continuing-treatment definition; is a father a covered “parent”; does 22 hours since February clear 250 base hours; is a two-days-a-week pattern an intermittent leave or a reduced leave schedule (the regulation defines them differently — intermittent periods must each be “at least one workweek”); and does the employer’s answer trigger the burden-of-proof memo.

That classification step — free-text situation to statutory category, with the citation attached — is the product. The arithmetic afterwards is trivial. The drafting of the notice, the certification form, and the harm memo in the employer’s own facts is the second AI job, and it is what converts a verdict into a defensible artifact.

Language grounding is straightforward here because the ruleset is public, finite, and cited — the regulation, the amended statute, and the DCR fact sheet. Every answer ships with the paragraph it came from, which is exactly what makes an owner trust it enough to act.

7. Localization angle (if any)

N/A as a country play, but this is aggressively state-localized and that’s the whole wedge. New Jersey’s stacking behaviour against FMLA and its TDI/FLI interaction are specific to New Jersey; a generic national leave tool gets the North Bergen question wrong. Start in one state where a dated statutory change just created the customers, then port the engine to the next state that expands coverage — the ruleset shape transfers even though the rules don’t.

8. Business model — path to $1M–$5M ARR

• Pricing: $149/mo for the 15–50 band (coverage + eligibility + notices + leave file). $349/mo for multi-entity or multi-state employers. One-off “single leave verdict” at $299 for the owner who shows up mid-crisis and won’t subscribe yet — this converts, because it lands during the panic.
• ACV: ~$2,000.
• Rough math to $1M ARR: 500 customers × $149/mo × 12 = $894K, plus one-off verdicts and a slice of $349 accounts clears $1M. 500 New Jersey employers out of tens of thousands in-band.
• Rough math to $5M ARR: ~2,300 subscribers, which realistically means riding the 2027 (10-employee) and 2028 (5-employee) expansions and porting to two or three more states with comparable leave regimes. New Jersey alone at $5M is possible but tight; the phase-in plus a second state is the honest route.
• Expansion path: per-employee leave files as headcount grows, additional states, then adjacent NJ duties the same buyer already fails at — earned sick leave interaction, the annual DCR posting and distribution requirement, and the handbook clause the regulation demands.

9. Go-to-market wedge — first 100 customers

• Rank for the panic query. The buyer’s first move is to search their exact situation: “does NJFLA apply to 20 employees”, “can I deny intermittent leave NJ”, “does FMLA count against NJ family leave”. Publish one page per question, each ending in the free coverage check. This is a small, high-intent, low-competition keyword set owned today by law-firm alerts that answer nothing and ask you to call.
• Free coverage check as the whole top of funnel. Six questions, instant dated determination, emailed as a PDF. Employers who come back “covered” and have no handbook get told, accurately, that they owe every employee written guidance right now — and the paid tier generates it.
• Sell through the people who already have the client list. New Jersey bookkeepers, payroll bureaus, and small CPA firms field this question and cannot answer it. Recruit 20 as referral partners at 20% recurring; they get to look competent instead of saying “ask a lawyer.”
• Target the associations by vertical. NJ dental, restaurant, landscaping, auto-repair, and child-care associations all publish member newsletters and run compliance webinars. Run the webinar for them for free — “the July 2026 change and what your 18-person practice now owes” — and close from the attendee list.
• Chase the phase-in twice more. Every June before 17 July 2027 and 2028, the newly-covered band is a fresh named list defined by headcount. Rerun the entire playbook against employers with 10–14, then 5–9 staff.

10. Build complexity — justification

Low. The ruleset is public, bounded, and already written down — the regulation is a few thousand words with worked examples. There is no integration requirement for v1: the owner types in a start date and hours rather than syncing an HRIS, which is fine because they only do this a couple of times a year. The real work is encoding the decision tree correctly and having a New Jersey employment lawyer review every determination path and template before launch. A solo builder ships this in 6–8 weeks; the legal review is the long pole, not the code.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Compliance tooling, not legal representation. Must carry clear “not legal advice” positioning and route genuine edge cases to counsel.
Ethical — no harm / dark patterns	✅	Helps employers grant leave correctly. The product’s default on a close call is to advise granting, not denying.
Market exists (evidence above)	✅	Enacted statute, dated phase-in, ~400,000 newly covered workers, priced-out substitutes.
1–5 person team can build this	✅	Solo builder plus a reviewing employment attorney.
Launchable with <$50K / ₹40L	✅	Legal review is the main cost; well under $50K.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	16/20	Severe when it hits — punitive damages, fees, NJLAD remedies, and a documented enforcement example. Docked because it hits episodically: a 15-person firm may see one leave request a year, which weakens the felt urgency between events.
Demand evidence	15	12/15	Enacted law, dated phase-in, state’s own 400,000 figure, incumbents whose floor sits above the whole segment. Docked for the absence of direct verbatim complaints from NJ small employers — the buyer’s pain is documented through lawyers and vendors, not in their own words.
Build feasibility	15	13/15	Public bounded ruleset, no integrations in v1, 6–8 weeks. Legal review adds calendar time.
Distribution clarity	15	12/15	Named channels — panic-query SEO, bookkeeper/payroll referrals, vertical associations, and a phase-in list that regenerates twice. Docked because SEO takes months to compound and association webinars move slowly.
Revenue mechanics	15	11/15	Pricing sits an order of magnitude below every substitute, so willingness-to-pay is safe. Docked because $1M in New Jersey alone is achievable but $5M requires the phase-in plus more states.
Time to first revenue	10	8/10	The one-off $299 verdict can sell during the first month; subscriptions follow.
Defensibility	10	4/10	Honestly weak. A well-encoded ruleset is copyable. The moat is being the first name on the panic query, the accumulated leave files customers won’t move, and the referral bench — a 6–12 month head start, not a wall.
Total	100	76/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · content-heavy

The content requirement is real: the distribution wedge is winning a specific set of search queries, and that means writing the definitive public answer to twenty questions New Jersey lawyers currently refuse to answer for free.

Key assumptions to validate (3–5)

1. Assumption: NJ employers in the 15–50 band will pay a subscription for something they use once or twice a year. How to test: offer 30 of them the $299 one-off verdict first; if one-offs sell but subscriptions don’t, the product is a per-event service and pricing must be rebuilt around that.
2. Assumption: the panic queries have enough volume to matter in a single state. How to test: pull keyword volume for the twenty target phrases; if combined monthly volume in NJ is trivial, SEO is a support channel and referrals become the primary one.
3. Assumption: bookkeepers and payroll bureaus will refer rather than ignore. How to test: pitch 20 NJ bookkeeping firms; look for 5 signed referral agreements.
4. Assumption: an owner will trust a software verdict enough to act on it without calling a lawyer. How to test: put the free coverage check in front of 50 owners and measure how many proceed to generate notices versus how many say “I’ll run it past my attorney.”

Risk flags

1. Regulatory risk (favourable direction, but real): the Division on Civil Rights is expected to issue updated guidance around the amendments. Good news — it creates content and urgency — but every determination path must be re-reviewed when it lands.
2. Platform-adjacent dependency: if Gusto, Rippling, or a PEO decides state leave determination is table stakes and bundles it free, the standalone tier compresses fast. Note that today none of them monitor employment law changes, which is the whole opening.
3. Episodic usage churn: a customer with no leave requests for twelve months questions the renewal. The leave file, threshold watch, and annual posting duty have to carry the between-events value or churn eats the model.
4. Single-state concentration: the entire business rests on one state’s statute until the second state ships. A New Jersey-specific legal reversal would be fatal in a way a multi-state product wouldn’t be.
5. Liability positioning: selling determinations to non-lawyers demands disciplined scoping, clear disclaimers, and an escalation path to counsel on genuine edge cases.

14. Structured verdict

Score:                  76/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical solo founder who will write the content themselves, with a NJ employment attorney on retainer for review
Time to revenue:        6–10 weeks
Capital to launch:      $8–15K (mostly legal review)
Top 3 assumptions to validate first:
  1. Subscription vs per-event willingness to pay — sell 30 one-off $299 verdicts before building the recurring tier
  2. Panic-query search volume in a single state — keyword pull across the 20 target phrases
  3. Bookkeeper/payroll referral channel converts — 20 pitches, target 5 signed
Kill criteria:
  - Abandon if fewer than 5 of 30 NJ employers offered the one-off verdict at $299 buy it
  - Abandon if a major HRIS or PEO ships free NJ state-leave determination before v1 launches
  - Abandon if 12-month subscription retention runs below 60% in the first cohort — that means it is a service, not a SaaS

15. Next step — 1-week validation sprint

• Day 1–2: Build the free coverage check only — six questions, dated determination, no account required. Ship it on a single page targeting “does NJ family leave apply to my business”.
• Day 3–4: Call 30 New Jersey employers in the 15–50 band, sourced from dental, restaurant, and landscaping association member directories. Ask two questions: have you had a leave request since July, and would you pay $299 to have someone tell you in writing what you owe. Offer it live on the call.
• Day 5: Pitch 10 NJ bookkeeping and payroll firms on the referral deal.

Go/no-go: ≥5 paid $299 verdicts sold from 30 calls, or ≥3 signed referral partners from 10 pitches. Anything less means the pain is real but not yet urgent enough to open a wallet, and the right move is to wait for the July 2027 threshold drop rather than build now.