How Lemon Man improves the culture of employment in Singapore.
A worker-reliability platform aimed at the shift, contract and freelance sectors the MAS rolling-bad-apples regime never reached. The mechanism is private-sector tripartite infrastructure — but the outcomes are public-good: more responsible employees, more lifelong learners, more civic-minded workers, fewer rosters broken by serial no-shows.
The one-line frame. Lead every tripartite conversation with it.
Three forces, all government-aligned.
Reliability becomes economically rational. Training becomes worth doing. Civic-mindedness gets rewarded. All three are wired into the Lemon Score and reduce it; all three map to existing Singapore-government priorities.
No-shows, last-minute bails, malingering MCs, data-secrecy breaches stop being free externalities. The corroboration multiplier — quadratic — ensures one employer is a complaint, two is a pattern. Workers internalise that reliability follows them. Employers stop carrying the cost alone. Roster fraud — the same lemon rolling through eight agencies undetected — becomes self-correcting at the market level.
WSQ completion is a first-class Lemon-Score reducer. SkillsFuture credit utilisation rises mechanically because workers can see the score effect. WSG-funded programmes (CCP, PCP, JR+) become stickier — completion rates lift because the score reward depends on finishing. For the first time, WSG can attribute training-induced behaviour change to specific programmes at sector scale.
Certified volunteer hours at IPC-registered charities, and tax-deductible IPC donations, both reduce the Lemon Score directly. Lemon Man becomes a measurable funnel into the civic ecosystem from a population segment — shift workers, gig labour — that Singapore Cares / Forward Singapore / Community Chest normally struggle to reach. Civic engagement, by mechanism rather than exhortation.
One sentence each. Already in the language they use.
Every tripartite touchpoint gets a sentence that fits inside their existing mandate. No agency is asked to invent a new category — Lemon Man is the private-sector extension of work they are already doing.
| Stakeholder | The anchor |
|---|---|
| MOM | Non-FI extension of the mandatory reference-check regime in force since Dec 2023. Same architecture — fixed schema, 21-day response, bounded lookback, right of reply. Closes a regulatory gap; the FI version was a tripartite success, the rest of the economy still lacks it. |
| TAFEP & WFA | Bidirectional by design. Workers counter-review employers on the same evidence standard. Protected characteristics rejected at intake. Annual transparency report. Architecturally cannot be characterised as one-sided. |
| WSG & SkillsFuture | WSQ completion directly drops the Lemon Score. Verifiable, third-party-confirmed. Behaviour-change attribution at sector scale, for the first time — answers the question SkillsFuture has been asking for a decade. |
| NTUC & SNEF | Worker bill of rights baked into the architecture: notification before publication, 21-day reply, hidden until reviewed, free independent appeal, structured rehabilitation. Counter-record gives workers a real channel against wage-shorting and illegal deployment — a mechanism they currently lack. |
| PDPC | Singpass-MyInfo claimed profiles. Evaluative-purpose exception for employer lookups. External DPO from day one. Hard-coded record expiry. SG-only data residency at launch. PDPA-by-construction, not by audit. |
| IPC sector · Community Chest · NCSS | Volunteer hours and tax-deductible giving as explicit score reducers — a measurable, opt-in funnel into the civic ecosystem from a hard-to-reach population. Civic outcomes become a measurable KPI of a private platform, the first time at scale. |
Four questions every agency will ask. Answers ready.
No. Bidirectional, decayed, evidence-required, contestable, appealable, time-limited. Functionally closer to a credit bureau than to a blacklist; Credit Bureau Singapore is the live within-SG precedent — already government-approved, already governed by MAS-aligned consumer-data norms.
Built to the Ramesh Krishnan v AXA standard: facts only, evidence at intake, reasonable care, narrow circle of publication, real chance to respond. The 21-day reply window mirrors the MAS reference-check rule explicitly.
Protected-characteristic records rejected at intake. Due-inquiry evidence required for disciplinary records. Annual transparency report. Bidirectional employer-side records mean the platform cannot be weaponised one-way.
The architecture protects them. A good record now travels with the worker — a peach is identified, not just a lemon. Bad actors can no longer roll between agencies burning rosters undetected. The rehabilitation pathways (WSQ training, time decay, civic engagement) are calibrated for low-friction access. Free for workers, always. Appeals free for both sides. No worker pays to clear a record.
Five steps. Sequenced. No agency asked twice.
A consultative pre-engagement before MVP, then sandboxed validation, then code-of-practice formalisation. Mirrors how MAS rolled out the reference-check regime: from industry pilot → guidance → mandatory norm.
- Informal pre-consultation with PDPC, MOM and TAFEP before the MVP build. Validate the architecture; capture early concerns; revise before code rather than after.
- WSG endorsement that WSQ completion is a recognised Lemon-Score reducer; ditto SkillsFuture credit utilisation. Single-page MoU. Mutual KPI: training completion uplift.
- NTUC briefing of worker-side features — claimed profile, counter-record, free independent appeal, decay path. Sign-off on bidirectional design before sector pilots.
- Sandbox pilot under MOM observation in one contained sector — licensed security or F&B recommended. Quarterly review against four metrics: no-show rate, WSQ completion rate, certified volunteer hours, worker disputes resolved in worker's favour.
- Industry code of practice with Lemon Man as default infrastructure for shift/contract reliability data — analogous to how Credit Bureau Singapore became the de-facto reference for consumer credit. End-state: a tripartite norm, not just a startup.
Four converging windows.
What to hand over, in which order.
Once tripartite outreach begins, three artefacts get circulated in sequence. None of them require the platform to exist yet — they describe the architecture, the legal posture, and the policy alignment.
- This brief — the institutional-register write-up. Circulate first.
- The legal memo (confidential, by counsel) — Ramesh Krishnan analysis, PDPA evaluative-purpose mapping, WFA-compliance checklist. Circulate to legal contacts only.
- The full Intel Cabinet PDF (compiled from the other admin pages — competitors, market sizing, pricing, whitespace, design audit). Circulate to senior policy contacts who want the full picture.
Use this when the conversation drifts toward enforcement. Lemon Man is incentive design, not enforcement.