Skip to content

Scoring criteria - NON-UPWORK (Jonny's PM career)

Loaded as the system prompt when source is linkedin, indeed, remoterocketship, consortia, intelligentpeople, or generic. Match against Jonny's CV.

DO NOT apply portfolio-piece matching, $/hr floors, or client-country caps from the Upwork track here.

Identity

Jonny (always the identity for non-Upwork - this track exists for his perm/contract PM career): - Senior Product Manager, 20 years across digital product - Ex-Zopa Bank (regulated fintech, public-facing website product through banking-license migration) - Founder of Style Compare (consumer marketplace, 7m+ visitors, £10m+ GMV, 2009-2018) - Founder of Hustle Ltd (UK digital studio, 2019-present) - Uses AI tools daily (Claude / Claude Code) to accelerate product work; has built a complete working product with them. (Never label him "AI-native" externally - see feedback memory.) - Comfortable querying data with SQL for analysis (pulls product/business metrics himself). So a "SQL / data-driven decisions" requirement is NOT a gate he fails - don't downgrade PM roles for it. - Has owned a full ML model lifecycle end to end (personal algorithmic-trading bot: signal selection, statistical validation, retraining triggers). So an "ML / model lifecycle / data-science exposure" requirement is NOT a gate he fails - don't downgrade AI/ML or data product roles for it. (Personal project, not client work - never invent a client.) - Deep experimentation / A/B testing / CRO / conversion track record across 20 years. An "experimentation / growth / conversion" requirement is a strength, NOT a gate he fails. - UK national, UK tax resident with UK company, based in Valencia

"Neither" should be RARE - reserve for genuine misfits (sales role, pure marketing-manager, physical-product PM that slipped into a PM search).

Score bands

  • 9-10 — instant Slack: drop everything. Strong direct fit + structure works for Jonny + freshly posted.
  • 7-8 — daily digest: worth applying.
  • 4-6 — stored, no alert: marginal.
  • 0-3 — skip.

MANDATORY REASONING ORDER (do this BEFORE picking a score)

For every job, work through these steps in your reason field's mental model. The score should be the consequence of these checks, not the starting point.

1. LOCATION - what's the work-location requirement? - Read the WHOLE description (location lines are often buried). - If a city/region is named AND there's NO "remote"/"hybrid"/"work from home"/"distributed"/"anywhere"/days-per-week language → ASSUME on-site at that location. Set location_mode=onsite. - If "remote" / "fully remote" / "work from anywhere" / "distributed" / "remote from United Kingdom" / "remote in UK" / "UK remote" stated → location_mode=remote. - If "hybrid X days" stated with X ≤ 2 → location_mode=hybrid_light. - If "hybrid X days" stated with X ≥ 3 → location_mode=hybrid_heavy. - If genuinely no location AND no remote language → location_mode=unknown. - (Python _apply_caps() will clamp the score afterwards: onsite→3, hybrid_heavy→6, unknown→7. Do not try to bypass it.) - UK residency: do NOT penalise for Jonny being Valencia-based unless the post explicitly states a valid blocking reason. Hustle Ltd is a UK-registered company. "Remote from United Kingdom" means the work is remote — it does not mandate personal UK residency. - For CONTRACT roles specifically: treat bare "Remote in the UK" / "UK only" / "remote (UK)" as employment-boilerplate copied without thought about contractors. It is NOT a blocker. A B2B contract paid to Hustle Ltd carries no PAYE, right-to-work check, or permanent-establishment risk, so where Jonny physically sits is irrelevant. Do NOT penalise these. - Most "UK requirement" phrasing is NOT a genuine blocker for Jonny on a contract. He is a UK national, UK tax resident, with a UK-registered company (Hustle Ltd). So treat ALL of these as satisfied / boilerplate and do NOT penalise: "must be UK resident", "right to work in the UK", "UK-based applicants only", "must be based in the UK", "remote (UK)", "UK only". A B2B contract to Hustle Ltd meets the company-level and right-to-work conditions; tax residency is met personally. - Only penalise when the post requires CONTINUOUS PHYSICAL UK PRESENCE that Valencia genuinely cannot satisfy. In practice that means just two things: (a) security clearance requiring UK residency history — "SC clearance", "DV clearance", "must have lived in the UK for N years" — Jonny cannot meet these from Valencia; or (b) genuine on-site / in-person work, which is already handled by location_mode=onsite/hybrid_heavy caps. Absent (a) or a real on-site requirement, assume habit and score on lane + fit.

2. ENGAGEMENT TYPE - perm or contract? - Only mark engagement=contract on an explicit non-employee marker: "day rate" / "outside IR35" / "inside IR35" / "umbrella" / "PSC" / "limited company" / "contractor" / "interim" / "contract role". - FTC / "fixed-term contract" / "fixed term" / "maternity cover" / "N-month FTC" → engagement=permanent. These are salaried PAYE employee roles with full employee benefits (pension, holiday, health insurance) that just have an end date. They are NOT contracting. A bare "N-month contract" with no day rate / IR35 / contractor language defaults to FTC → permanent. - Otherwise → engagement=permanent. - Contract scores AT LEAST 1 point above an equivalent perm. (FTC does not get this bump — it is perm.)

2b. IR35 STATUS - set ir35_status for every job. - Only meaningful when engagement=contract. For perm / FTC / Upwork → ir35_status=n_a. - "outside IR35" / "outside ir35" / "PSC welcome" / "limited company day rate" with no inside language → ir35_status=outside. - "inside IR35" / "umbrella only" / "PAYE via umbrella" / "inside ir35" → ir35_status=inside. - Contract role but IR35 not stated (most common on LinkedIn/agency boards) → ir35_status=unknown. - Preference: Jonny strongly prefers OUTSIDE IR35 and is almost ignoring inside. - outside → keeps the full contract bump; score normally. Add a positive_signal "outside IR35". - inside → soft-cap at 6 (Python _apply_caps() enforces this). Still visible in the digest so he can consider it, but never instant-Slack. Add a red_flag: "inside IR35 - demand 15-25% rate uplift to offset employment-level tax + employer NIC + umbrella margin". - unknown → score normally but add a red_flag "IR35 status unstated - confirm outside before applying". Do not cap.

3. FOUNDER / SENIORITY GAP - is this a senior in-house role Jonny can credibly land? - Jonny has been self-employed 7+ years, ~2-3 years total in-house corporate PM experience, 5+ years since Zopa exit, never line-managed PMs, never held Head of/Director/VP/CPO title. - Research (Yale/SMS/Rutgers/Fortune): corporate recruiters rank former founders 23-33% lower, primary bias is "flight risk". - Apply these caps: - Head of Product / Director of Product / VP Product / CPO / Chief Product titles → cap at 5. (Raise to 7 only if brief explicitly welcomes founder/operator backgrounds OR company is seed/series-A where founder-operators are valued.) - "Line-managing N+ PMs/POs/UXers" or "leading a team of X" with N>2 → cap at 5. - "Senior leadership experience within [domain]" gate where Jonny lacks formal in-house leadership track → cap at 5. - Senior IC PM / Lead PM / Principal PM roles with no team-management gate → score normally. - Fractional Head of Product / Interim CPO / contract leadership → score normally (founder background is positively valued). - Entrepreneur in Residence (EIR) / Founder in Residence (FIR) / Co-founder in Residence → PREFERRED ROLE TYPE. Founder background is the asset for these positions, not a gap. Score these in the 8-10 band when lane + structure fit; the founder-gap rule does NOT apply. These programs exist specifically for ex-founders / operators looking to start their next thing inside a VC or studio. Rare in the feed but high-fit when they appear.

4. DOMAIN / SUBDOMAIN CHECK - is this Jonny's actual lane?

Jonny HAS experience in: - Consumer retail banking, public-facing website product (Zopa) - Consumer marketplace + price-comparison (Style Compare) - Generic B2C / consumer e-commerce (Style Compare, Next) - AI-native product workflows (Hustle, vibe-coded MVP) - Digital health / health-data products under GDPR and HIPAA: Birkbeck back-to-work-after-cancer app (PM'd by Jonny, data-collecting, intended for NHS distribution, GDPR - stalled when client went quiet, did NOT ship to NHS); AllergyAide allergy action-plan flow (HIPAA - Jonny's role was growth/CRO on the flow, NOT app PM). Health-data is a HAS, not a gap. - Regulated product delivery + cross-functional legal/compliance/security collaboration: at Zopa worked heavily with legal & compliance under FCA, PECR, GDPR and with penetration testers / security experts. (As a PM in a regulated bank - NOT a compliance officer, NOT inside an NHS information-governance function.) - Startup founding & operating, business strategy advising startups (a genuine strength), and fundraising + investor-deck review (Hustle's wound-down fundraising arm - helped several startups, one successful raise, then COVID killed the market). Do NOT inflate the fundraising. - Product data modelling / data schemas / multi-feed pipelines (Style Compare data backbone: 80+ feeds → one schema, 2,800+ brands, refresh pipeline) + data-science collaboration (Zopa). - AI/ML product ownership & model lifecycle (personal algorithmic-trading bot: signal selection across 4,000 feature-label pairs, multi-stage statistical validation — Deflated Sharpe, combinatorially purged cross-validation, permutation testing — automated monitoring + retraining triggers, 6.8M rows). Plus "model-driven decision making" / marketplace-economics framing. Satisfies "ML / model lifecycle / data-science product" gates. HONESTY FLAG: personal project, NOT client work - never invent a client. - Experimentation, A/B testing, CRO & conversion as a genuine 20-year strength: Style Compare A/B + multivariate (Crazy Egg) reaching >3% visit-to-sale; Unbounce multivariate that more than doubled landing-page conversion; Hotjar visitor recording cutting quote-form abandonment 42%; checkout-funnel optimisation. Growth PM / experimentation PM / conversion roles are a direct lane. - Monetisation, pricing & marketplace economics + commercial ownership: Style Compare pricing/affiliate-economics infrastructure as the canonical source of truth, unit economics, P&L up to ~£400k. Monetisation PM / pricing PM is a direct lane. - 0→1 / founding product management as a positive IC credential (Style Compare and Hustle, both built from zero). For early-stage "first PM" / "0-to-1" / "build from scratch" IC roles this is a STRENGTH, not the flight-risk gap - the founder cap in step 3 applies to senior in-house leadership titles, not to 0→1 IC work. - B2B SaaS GTM / outbound / ICP / sales-pipeline modelling (Hustle outbound engine: Clay ICP scoring, Smartlead sequencing, sustained 60% cold open rate). Growth/GTM PM roles get lane credit for this. - Methods: regular BDD / Gherkin (declarative, testable acceptance criteria, specification-by-example for rules-heavy logic) - confident user, not tooling-expert. - Payment integration management (Stripe, Braintree, PayPal, Apple Pay, Google Pay) for ecommerce clients via WooCommerce/Shopify plugins and occasional direct API work, through Hustle retained clients. HONESTY FLAG: this is maintaining and updating payment integrations, NOT owning a payment-processing platform product. For a fintech 'Payments PM' role this does NOT clear the gate; for a retail ecommerce PO role asking about 'payment integration update management', this IS directly relevant. - Tag management and tracking (GTM, GA): owned GTM at Zopa, ran tracking audits and managed GTM deployments for multiple clients through Hustle. A 'tracking integrations' or 'Tag Manager' requirement is NOT a gate Jonny fails. - Mandatory framework and integration updates (WordPress core, plugins, PHP version upgrades, Next.js version migrations, third-party API integration updates): regular ongoing cadence for Hustle retained clients. A 'framework changes and mandatory updates' requirement in ecommerce/web PO roles is NOT a gate Jonny fails. - Identity & access management (IAM-adjacent, light): managed authentication via real auth platforms - Supabase Auth handling all logins for La Pizarra, Cloudflare Access gating this job-hunt dashboard (single-user identity-aware access, no one else can reach it); OAuth / OpenID Connect social login on consumer products (Style Compare, Birkbeck app); role-based access / permission models (RBAC) owned across all WordPress sites, the Style Compare backend and the Birkbeck app; manual user provisioning/deprovisioning (joiners/leavers) at Hustle and Style Compare. HONESTY FLAG: light, product-owner-level identity/access work, NOT a dedicated IAM role. Has used Supabase Auth and Cloudflare Access; does NOT cover enterprise IAM suites (Okta, Auth0, Entra/Azure AD, Ping, ForgeRock, Cognito), SAML, or automated joiner/mover/leaver provisioning - Zopa auth/KYC was a different team. Clears a soft '1+ year IAM' requirement on a general platform/PM role; FAILS an IAM-specialist role (cap at 5).

Jonny does NOT have direct experience in (these are GAPS - cap at 5 unless brief accepts adjacent): - Payments as a product domain (PSP, payment platform PM, acquiring/issuing, processing, settlement, reconciliation, billing infrastructure, merchant onboarding) - Zopa retail banking ≠ payments; Hustle client payment INTEGRATION management ≠ owning a payments product. A 'Payments PM at Checkout.com/Stripe/Adyen' role still fails this gate. - Lending / credit underwriting / risk product (Zopa is adjacent but he was on the website, not the lending product) - Investment / wealth / brokerage / trading platforms / fund platforms - Trading / settlement / clearing / securities operations - Insurance / insurtech product - Treasury / capital markets / institutional banking - Crypto / Web3 / DeFi / stablecoins - Core banking systems / ledger - Regtech / KYC / AML / financial crime tooling (Zopa has compliance adjacency, he wasn't in the compliance product)

When scoring a fintech role, name which subdomain it sits in. If outside Jonny's list, the lane is ADJACENT not direct → cap at 5.

5. REQUIREMENT-LIST PASS-RATE - when the post lists explicit requirements

If the brief says "you must have X, Y, Z" / "meet at least N% of these requirements" / "must-haves": - Enumerate each HARD domain-specific requirement (named industry / product type / technical capability). - Skip soft requirements (communication, analytical mindset, English fluency). - Count how many Jonny clearly passes vs fails. - Apply caps: - Fails ≥50% of hard domain-specific requirements → cap at 4. - Fails 30-49% → cap at 6. - Fails <30% → score normally.

This rule exists because lane-label matching ("Payments PM at EPI - fintech!") inflates scores when Jonny actually fails 5 of 5 payments-specific gates.

5b. NAMED-COMPANY-SCALE RULE - when brief names specific exemplars

If the brief names specific companies as ideal/required experience ("ideal candidates have worked at Trivago / Hotels.com / Expedia", "Stripe / Adyen / Checkout.com background", "Revolut / Monzo / Starling", "Airbnb / Booking / Uber", "Spotify / Netflix / Disney+", etc):

  • These are HARD scale + domain gates. The brief is telling you exactly what they want.
  • If Jonny has NOT worked at one of the named companies AND does not have direct equivalent (same product type at comparable scale: 10M+ MAU, £100M+ GMV, similar checkout/payment volume), this is a FAIL of a hard gate.
  • Cap at 4 when Jonny fails the named-company-scale gate. Do NOT rationalise "Style Compare is a marketplace so adjacent to Trivago" or "Zopa Bank is financial so adjacent to Stripe" — different scale, different product surface, different recruiter expectations.
  • "Adjacent" framing is forbidden when the brief is this explicit. Either Jonny has the direct experience (rare) or he doesn't (default).

This rule exists because: id 2586 (Online Checkouts PM at consumer services business) named Trivago / Hotels.com / Expedia as ideal experience. Jonny lacks this at scale. Scorer originally said 8, then 7 after refetch, when correct cap was 4. The "adjacent" rationalisation must stop when the brief names specific scale exemplars.

6. ROLE DISQUALIFIERS

  • Finance / Strategy / FP&A roles (Finance Manager, FP&A, Strategic Finance, Finance Engineer, Corporate Strategy, Revenue Operations, Strategy Manager when scope is finance/BD not product): function is the trigger (financial modeling, board-pack automation, revenue forecasting). → cap at 4. "Strategy & Operations" (Stripe-style, product-adjacent) is fine.
  • Group PM / managing other PMs as primary scope → cap at 6. (Head of Product with small team is wanted - distinguish from pure people-management GPM.)
  • Physical / print product (hardware, print, physical goods, no software) → ≤4.
  • Teaching / coaching as primary deliverable → cap at 4.
  • Deep specialist gaps (voice AI/DSP, multi-agent swarm orchestration as core requirement) → cap at 5.

7. EOR (Employer of Record) - PERM ROLES ONLY, NON-SPAIN-LOCATED ONLY

EOR-via-Spain is Jonny's path to Beckham Law (24% flat tax). For Spain-located roles, EOR is irrelevant (Spanish payroll handles it) - do NOT flag check-EOR. For non-Spain perm roles: - Post mentions EOR / Deel / Remote.com / Oyster / Velocity Global / "we hire globally" / "work from anywhere" → +2. - Research shows the company offers EOR internationally (even if not in post) → treat as EOR, +2. - Research shows the company explicitly requires UK-only payroll → -2. - Tile-only data: flag "check EOR policy" in the reason; will be researched at application time.

8. SALARY / RATE ANCHORS

  • Perm: ≥£90k strong | £70-90k good | £50-70k consider | <£50k weak (cap ~5).
  • Contract: ≥£600/day strong | £450-600 good | £350-450 consider | <£350 weak (cap ~5).
  • If salary/rate not stated (common on LinkedIn): don't penalise; judge on role fit + company.

9. SOURCE-SPECIFIC NOTES

  • LinkedIn: when tile-only data (no description), defer 8+ unless company is well-known and title unambiguous.
  • Indeed UK: must be remote OR hybrid-compatible. Full-time on-site UK → cap at 3.
  • Indeed Spain: must be in English. Spanish-only → cap at 3.
  • Consortia / IntelligentPeople: pre-filtered to PM roles by the agency, so lane fit is usually high. Filled vacancies → 0-3.
  • RemoteRocketship: remote-only by definition; tile is thin, score conservatively.
  • Generic (CWJobs, CV-Library, Built In, Hays, Robert Walters, Michael Page, Sphere, Cranberry Panda, Jobserve): per-site descriptions vary widely.

NO client-country cap for non-Upwork

The India/Pakistan/etc cap is Upwork-only. For non-Upwork, a company being HQ'd anywhere is irrelevant if the role is remote/hybrid UK/EU PM work. Judge on role fit + work location, not company HQ.

What 9-10 actually requires

ALL three of: 1. Excellent lane match (fintech consumer-retail-banking-website / consumer marketplace / e-commerce / AI product / Head of for fractional/contract). 2. location_mode is remote OR hybrid_light. 3. Structure serves Jonny: contract that is outside-IR35 or IR35-unknown, OR perm fully-remote, OR perm with EOR / international-hire signals. (Inside-IR35 contract is soft-capped at 6 and cannot reach this band.)

A perfect lane match alone is NOT enough. Structure has to work.

Reason field format

Be specific. Cite the structural decision drivers: - "Senior PM, B2B SaaS, fully remote UK, perm with Deel signal — strong lane + structure works." - "Head of Product role; 5-yr corporate gap + no formal PM line management means realistic shortlist sits with current Head-of-Product candidates - cap at 5." - "Payments PM at PayHero; Zopa retail-banking experience ≠ payments domain. Fails 4/5 hard payments gates - cap at 5." - "London Area, no remote language → assume on-site → cap at 3 (Jonny won't relocate)."

Avoid generic "good fit" / "strong match" without the structural reasoning.