Netherlands DBA Freelancer Risk Radar

A EUR 9/month self-serve web app for Dutch zzp'ers (solo freelancers) juggling multiple clients that connects to their invoicing history, computes a live false-self-employment (schijnzelfstandigheid) risk score against the tax authority's own multi-factor test, most sharply the 70%-revenue-concentration danger line, and continuously assembles a timestamped 'genuine entrepreneur' evidence file, filling the gap between the government's one-shot, client-only assessment tool and the enterprise freelancer-management platforms built for companies, not for the freelancer's own protection.

研究阶段进度

① 需求扫描

② 市场调研

③ 可行性分析

分诊打分总分: 30/35

市场调研评估

6.2/10

评估阐述

Scale: 0-10, weighing demand-side strength (size, growth, buyer clarity, revealed willingness to pay) against competition-side pressure (density of existing players, incumbent strength, remaining differentiation room). Demand side is unusually strong for a niche this size: CBS's own labor force survey puts roughly 256,000 Dutch freelancers (about 21.5% of the ~1.19M main-income zzp population) in a client-concentration profile CBS itself flags, and hard behavioral data shows 59,000 people converting to payroll status in H1 2025 alone versus 32,000 a year earlier, real revealed action, not just complaints. The June 2026 Senate adoption of the EUR 38/hour rechtsvermoeden law and the Deliveroo/Uber Hoge Raad rulings (which explicitly name client-revenue-dependence as a weighed factor) further lock in the regulatory tailwind. Competition side is markedly more crowded than the initial scan assumed: direct verification found at least six live freelancer-facing or dual-sided schijnzelfstandigheid tools (ZZP Pulse at a near-identical EUR 9.99/month, Wet DBA Check, ZZP Compliance Check, ZZP Safecheck, ZZP CHECK, 0900-jurist), several offering the core check for free or under EUR 30 one-time, which undercuts recurring willingness-to-pay. One close analogue (ZZPShield, which appears to have marketed almost exactly this three-sided continuous-monitoring-plus-evidence-file thesis) is confirmed dead (hosting deployment not found), a caution flag whose root cause could not be determined. The surviving, genuinely open differentiation is narrow but real: no live competitor ties its score to actual invoicing data for automatic, continuous recomputation. Bottom-up sizing: SAM ~EUR 27.6M ARR (256,000 people x EUR 108/year), realistic 3-year SOM EUR 280,000-830,000 ARR at 1-3% penetration. No Wft or tax-advisor licensing barrier was found (belastingadviseur is not a protected title; Wft licensing targets financial products, not employment classification), so the regulated-advice constraint is a liability/disclaimer discipline rather than a hard blocker. Score held to 6.2 rather than higher because the competitive field is real, partially free, and already includes one shut-down attempt at nearly the same positioning; held above the midpoint because the demand evidence is exceptionally well sourced and the invoicing-integration gap is genuinely unclaimed.

可行性评估

淘汰

可行性评分3.2/10

评估阐述

Scale: 0-10 (0 = completely unviable, 10 = highly viable), weighing technical, financial, regulatory and competitive feasibility together. Verdict: INFEASIBLE.

Biggest killer: the only thing separating this from six live rivals is a revenue-share calculation over invoice data three larger companies already hold. Ranking a freelancer's clients by trailing-12-month revenue share is an aggregation over an invoice table that ZZP Pulse (EUR 9.99/mo, already ingests invoices), Moneybird (400,000 users) and e-Boekhouden (200,000+) already own. Days of work for them; months for a newcomer who must first build the dataset and persuade a stranger to hand over a full billing history to fill it.

Four HIGH risks, three with no mitigation on the table:

Differentiation is copyable by three parties better placed to ship it, and shipping first is what proves it worth copying.
The nearest predecessor died in this exact niche. Direct investigation (escalated item #1) established far more than the research stage could: ZZPShield marketed the same claims ("continu, niet eenmalig", doorlopende compliance-monitoring, timestamped bewijsdossier), at the same price band (EUR 7,95 Basis / EUR 14,95 Professional vs the EUR 9 hypothesis), with the same search-led funnel (~13 knowledge-base articles + Amsterdam/Rotterdam/Utrecht/Den Haag city pages + bouw/IT/transport/media sector pages + free scan), launched into the peak of the enforcement window. Archive index: first capture 2026-03-05, last live capture 2026-04-20 (HTTP 200), zero captures after. Dead as of 2026-07-16 (Vercel DEPLOYMENT_NOT_FOUND). Confirmed life under ~4.5 months. Cause NOT determined and not assumed benign: footer credited "SaasFactory" (KVK 99795078), but that entity could not be identified (no resolving domain; kvk.nl JS-gated, openkvk.nl and drimble.nl behind Cloudflare bot challenges, not circumvented per the legal-path rule). Priced as HIGH, not neutral, because the shutdown signature excludes the benign readings: the domain is still ACTIVE at SIDN (reseller TransIP) with DNS and MX still configured while the application was removed from underneath it, and there is no redirect, successor, notice or acquisition. Acquisitions redirect; pivots leave notices; rebrands forward the search equity of a 40-page indexed site. A paid-up domain pointing at a deleted app is what abandonment looks like.
No viable acquisition channel. Ceiling on CAC is EUR 56 at benchmark 5%/mo churn (LTV contribution EUR 167) and EUR 35 at the realistic 8%. Bottom-up paid search lands EUR 100-600, and the channel is structurally closed rather than merely expensive: the keyword cluster is bid by payrolling/detachering firms and employment lawyers earning 50-100x more per converted click than a EUR 108/yr product. Organic is the only channel left, and it is exactly what ZZPShield built before disappearing. Stress-tested generously (every NL zzp'er searches the topic annually, new site captures 5-10% of the national stream against the state, KVK, unions and six rivals), steady state tops out at EUR 17K-161K ARR against the EUR 280K-830K 3-year SOM. Every cell short; the most generous is 42% below the low end.
Price anchored at free. Six rivals confirmed still live July 2026; two give the core check away and charge EUR 24.95-29 once; the only recurring competitor offer is EUR 15/YEAR, one seventh of the EUR 108 proposed.

Escalated item #2 (standalone vs plugin): plugin fails on the same bottleneck and adds a landlord. Moneybird's API is free and open, but their published listing requirements (help centre, dated 28 May 2026) demand "minstens 25 actieve administraties en is al minstens 2 maanden actief" before listing, with no revenue share and no co-selling. The listing is a reward for traction, not a source of it. e-Boekhouden has no self-serve marketplace at all ("neem contact met ons op"), and no compliance category exists in its directory. The plugin route solves data access and the trust barrier; it leaves acquisition exactly where it was, worsens pricing power (an add-on beside a EUR 3-41 subscription that already does the whole job), and maximises absorption risk by demonstrating the feature on the incumbent's own data.

Why 3.2 and not lower: demand is among the best-evidenced in this library and behavioural rather than survey-based (CBS: 59,000 conversions to payroll in H1 2025, population down 62,000 in 2025, ~256,000 in the concentrated-client profile; the state's own tool says in writing it is "not for contractors"). No licensing barrier exists (tax-adviser title unprotected; Wft covers financial products, not employment classification). Build is cheap (EUR 10,700 cash self-built, or EUR 50,700 at market cost) and cash break-even is 24 subscribers. The invoicing gap is genuinely open today. A determined operator could ship something people would use.

Why not higher: viability is decided by whether the thing can be sold profitably and defended once sold, and this fails both. Honest break-even is not 24 customers but ~1,012, the point at which the business beats what the same person clears contracting at EUR 95/hr. Holding even the low SOM needs 128 new paying customers every month forever against 5% churn, in a market shrinking ~3%/yr whose highest-intent buyers resolve their anxiety by leaving self-employment altogether.

What would reverse this: proof the predecessor died for an unrelated reason with workable acquisition economics (cheapest open question left, worth an afternoon before any code); a validated accountant/bookkeeper referral channel; repricing at ~EUR 90/mo to clients and intermediaries who face the actual back-tax bill (a different product, and Lance Free is already there with venture backing); or an enforcement shock removing the soft-landing regime.

Netherlands DBA Freelancer Risk Radar

Track: Micro SaaS / API Wrapper / Bot | Market: overseas (Netherlands, solo zzp'ers / freelancers with multiple clients) | status: PENDING_RESEARCH | Created: 2026-07-17T00:00:00Z | Updated: 2026-07-17T00:00:00Z

Scout output, for downstream research/feasibility. Full metadata in meta.json in this directory.

One-liner

A EUR 9/month self-serve web app for Dutch zzp'ers (solo freelancers) juggling multiple clients that connects to their invoicing history, computes a live false-self-employment (schijnzelfstandigheid) risk score against the tax authority's own multi-factor test, most sharply the 70%-revenue-concentration danger line, and continuously assembles a timestamped "genuine entrepreneur" evidence file, filling the gap between the government's one-shot, client-only assessment tool and the enterprise freelancer-management platforms built for companies, not for the freelancer's own protection.

Opportunity source (how it was found)

Method: Trend Sniffer (regulatory-enforcement escalation plus a national statistics office publishing hard behavioral-shift numbers) combined with Pain-point Extractor (a documented gap in who existing tools serve), synthesized into an Idea Generator product.
Signal (Trend Sniffer): the Netherlands' Wet DBA (Deregulation of Assessment of Employment Relationships Act) moved from a warning-only enforcement posture to active enforcement on 1 January 2025, with the Tax Administration authorized to impose serious-fault penalties from 1 January 2026. CBS (the national statistics office) reports the zzp population fell by 62,000 in 2025, the first decline in years, from roughly 1.3 million to 1.23 million, with 59,000 freelancers switching to employee status in H1 2025 alone (nearly double H1 2024's 32,000), and the switch-out rate among freelancers active 1-5 years rising from 5.8% to 8.7% year over year. CBS directly attributes the shift to stricter DBA enforcement. CPB (the national economic policy bureau) forecasts a further 3% contraction in 2026. On top of this, the underlying rules kept moving mid-scan: on 6 March 2026 the minister scrapped part of the VBAR reform amid market unrest, detaching a proposed EUR 38/hour legal-presumption threshold into standalone legislation due by 31 August 2026, meaning the compliance target itself is still shifting under both freelancers and their clients.
Signal (Pain-point Extractor): the Dutch government's own WBA (Webmodule Beoordeling Arbeidsrelatie) assessment tool is confirmed, via direct fetch of its official description page, to be (1) a one-time questionnaire that cannot save progress between questions, (2) explicitly non-binding ("only an indication, not a legal decision"), and (3) built exclusively for the hiring party: "The online tool is not for contractors or sector representatives, unless they act in the role of commissioning agent." The freelancer bearing the actual financial risk (loss of the self-employed tax deduction, back-payment liability) is structurally excluded from the government's own assessment tool. Among paid tools, Lance Free and MaximeWorks both build DBA-compliance dossiers, but for companies managing a freelancer network (client side), not for the individual contractor. The one freelancer-facing app found, ZZP Pulse, is a general hours/trips/receipts/VAT bookkeeping tool with a single free static "hour criterion" checker, not a continuous, multi-factor risk score built around the specific danger metric advisors keep repeating: over 70% of a freelancer's revenue coming from one client draws direct scrutiny.
Idea Generator synthesis: combine the freelancer's own invoicing data (which no existing product uses for this purpose) with the tax authority's published multi-factor test to produce a running risk score, rather than a one-off checklist the freelancer has to remember to rerun. The product answers a question none of the existing tools answer for the freelancer directly: "based on how I have actually been working and billing this quarter, and not what my contract says on paper, how exposed am I right now, and what evidence do I have if the Belastingdienst asks?"

Demand detail

Who wants this: individual Dutch zzp'ers, roughly 1.2 million people as of end-2025 per CBS, who work for more than one client and want to keep their self-employed tax status (zelfstandigenaftrek, MKB-winstvrijstelling) intact. The sharpest wedge within that population is freelancers whose work naturally concentrates around one or two larger clients for stretches of the year (a common pattern in IT contracting, interim management, and creative freelancing), the exact profile advisors flag as highest-risk under the 70%-revenue-concentration guidance.

What they are expressing: not "what is schijnzelfstandigheid," a question every Dutch business-advice site, KVK, and accounting blog has already answered, but "based on my actual client mix and hours this year, am I currently inside or outside the danger zone, and can I prove it if asked." The CBS numbers are the strongest evidence available that this is not abstract anxiety: 59,000 people made the costly, irreversible decision to become employees rather than risk staying self-employed in the first half of 2025 alone, and short-tenure freelancers (the group least equipped to absorb a back-tax assessment) are leaving self-employment at a faster and rising rate.

Strength and breadth of pull: enforcement is not a single deadline that resolves, it is an escalating ladder (warnings through 2025, serious-fault fines from 2026, a still-undetermined further legal-presumption threshold due by August 2026), which means the underlying anxiety compounds rather than settling. The government's own tool structurally cannot serve the freelancer directly, and no competitor identified in this scan closes that gap from the freelancer's side with continuous, evidence-based monitoring.

7-dimension triage score (detail in meta.json.triage)

Demand pull 5 / Acquisition feasibility 4 / Agent advantage 4 / Low-volume economics 4 / Operator hand lightness 4 / Market trend 5 / Policy redline 4 -> Total 30/35

Rationale summary:

Demand pull (5): multiple independently corroborated, directly fetched primary sources, including a national statistics office (CBS) publishing hard behavioral-shift figures (62,000 net zzp decline, 59,000 employee-status switches in H1 2025 alone) that prove people are taking costly, irreversible action out of DBA fear, not merely complaining online. The government's own tool confirmed, via direct fetch, to exclude the freelancer as an intended user is rare, decisive negative-space evidence.
Acquisition feasibility (4): reachable channels exist (ZZP Nederland, PZO, FNV Zelfstandigen, KVK starter programming, Dutch accounting/bookkeeping software marketplaces such as Moneybird and e-Boekhouden, and an actively spiking SEO topic around "schijnzelfstandigheid" and "DBA 2026"). Not a 5 because incumbent freelancer bookkeeping apps (ZZP Pulse and similar) could plausibly bolt on a shallow revenue-concentration flag at low cost, the same platform-absorption risk logged for other compliance-adjacent lanes in this library (Ireland auto-enrolment, Australia Payday Super).
Agent advantage (4): continuously recomputing per-client revenue concentration from invoicing data, mapping observed working patterns (contract duration, schedule fixity, tooling ownership) against the tax authority's published multi-factor test, and auto-compiling a timestamped evidence file is a structured, recurring, rules-based task suited to automation, though it depends on accounting-API integrations that must be built and kept current as rules shift.
Low-volume economics (4): near-zero marginal cost per additional freelancer once the scoring engine and evidence-file generator exist; the price point mirrors the EUR 7-19/month sweet spot already validated for individual-freelancer compliance and finance tools elsewhere in this portfolio (e.g., US freelancer tax estimator, Singapore self-employed finance OS).
Operator hand lightness (4): the core scoring and evidence-compilation loop is fully automatable; the recurring manual task is watching for further rule changes, since the underlying legal test itself is still being amended mid-2026 (the March 2026 VBAR scrapping and the pending August 2026 threshold legislation).
Market trend (5): enforcement is actively escalating on a public timetable (warnings through 2025, serious-fault fines from 2026, a further legal-presumption threshold due by August 2026), CPB forecasts continued zzp-population contraction into 2026 attributed to this exact enforcement, and expert commentary as recent as April 2026 confirms the confusion has not settled even under a new cabinet. This is a rising, unresolved window, not a saturating one.
Policy redline (4, not 5): the product sits directly adjacent to personalized tax and legal advice, the same licensed-professional-adjacency redline already logged in this library for Ireland (pension advice), Canada (immigration advice), and Singapore (financial advisory). It must stay strictly informational and administrative, a risk score and evidence organizer, never a personalized determination of employment status or a substitute for a belastingadviseur or advocaat, with a clear disclaimer to that effect.

Notes for downstream stages

Key assumption to stress-test first: whether an individual freelancer will pay a dedicated subscription for a risk score and evidence file on top of (or instead of) their existing bookkeeping app, versus treating this as a feature they hope their invoicing tool (Moneybird, e-Boekhouden, Jortt, or ZZP Pulse) adds for free. Downstream research should look for direct evidence of willingness to pay, for example paid legal-templates or DBA-specific consultation products already selling to individual zzp'ers, before committing to a standalone-tool thesis over an "API/plugin sold into existing bookkeeping platforms" thesis.
Competitor/comparable leads for research: Lance Free and MaximeWorks (client-side freelance management systems with DBA dossier features, confirmed via direct fetch to be built for companies managing freelancer networks, not individual freelancers); ZZP Pulse (freelancer-facing general bookkeeping app with a single free static hour-criterion checker, confirmed via direct fetch); the government's WBA tool (confirmed via direct fetch to explicitly exclude contractors as intended users). No freelancer-facing continuous DBA risk-score product was identified; downstream research should re-verify this negative finding, and also check higherlevel.nl (a well-known Dutch zzp community forum) and Dutch accounting-software marketplaces for any adjacent tool missed in this pass.
Data source note: the specific 70%-revenue-concentration threshold is repeated consistently across multiple advisory sources (Rijksoverheid.nl-adjacent guidance, sector Q&A pages) but is described as one factor among nine reviewed together, not a standalone legal bright line; downstream research and any eventual product copy must frame it as a strong risk indicator, not a certainty threshold, to avoid overstating legal precision. zzpshield.nl (referenced in search results as a possible client-side compliance product) could not be reached on direct fetch (HTTP 404 on both the specific page and the root domain) and should be retried at the research stage.
Redline/compliance notes: never frame output as a legal determination of employment status, never claim to replace a belastingadviseur or advocaat, and carry a prominent "indicative only, verify with a qualified adviser" disclaimer, mirroring the disclaimer pattern already required across this library's Singapore, Ireland, and Canada compliance-adjacent lanes. Track the pending August 2026 legal-presumption threshold legislation as a required content update once published.
Geography note for the coordinator: this is the first Netherlands-market lane in this library. Overseas_lanes previously covered Singapore (heavily), the US, the UK, Australia, the EU as a bloc, Ireland, Canada, New Zealand, and Latin America, but had zero continental-Europe-specific lanes outside Ireland and the EU-wide AI Act lane. This diversifies both geography (a new country) and buyer persona (an individual freelancer protecting personal tax status, closer in shape to the FIRE/retirement consumer lanes and the gig-driver defense lane than to any SMB B2B compliance lane).

assets/ evidence list

evidence.md: full raw source list with direct quotes and URLs covering (1) the Wet DBA enforcement escalation timeline and CBS's directly-fetched hard behavioral-shift statistics (62,000 zzp decline, 59,000 employee-status switches), (2) the government WBA tool's confirmed one-time, client-only, non-binding nature, (3) a competitive sweep confirming Lance Free and MaximeWorks are client-side tools and ZZP Pulse is a general bookkeeping app without a continuous DBA risk score, and (4) explicitly flagged items not obtained (zzpshield.nl 404, maximeworks.com DNS failure, one government-commentary article blocked, Google Trends chart data not retrievable) excluded from claims rather than guessed at.