# Ofigo: business and launch guide

Working location: **Figueiró dos Vinhos, Leiria, Portugal**. Prepared 26 October 2026. Ofigo and ofigo.pt were selected by the owner on 28 October 2026. Selection does not establish domain registration, a registered company or trademark clearance. Amounts are euros. Timelines are relative to a staffed project start, not promised launch dates.

## 1. The decision

Start with a **scheduled, neighbourhood food delivery service**, initially serving a verified, compact area around the town centre on Friday–Sunday evenings. Use existing licensed restaurant kitchens, a responsive website and a human dispatcher. Begin with three contracted restaurants, one named local delivery operator and a trained backup. A single order contains food from one restaurant.

The service promise is: **local meals, a clear total price, a realistic delivery window, and someone accountable when something goes wrong**. The product should feel familiar to customers of Uber Eats, Deliveroo and Glovo; execution should fit a small town.

Do not lease premises, buy a fleet, hold grocery stock, build native mobile apps, or promise town-wide instant delivery before demand and route economics are measured. The first business milestone is 50 genuine paid orders with known costs and reliable fulfilment. That is a learning milestone, not proof of success. Success means sustained customer use, safe delivery, partner retention and sufficient cash contribution to pay for labour, management, overhead and taxes.

**Current state:** a privately published software prototype, research and execution documents. Example restaurants, prices and allergens are fictional. No commercial partners, riders, incorporation, bank account, payment processing or real deliveries have been established by this project. No outreach has been sent and no budget committed.

## 2. Evidence, assumptions and unknowns

| Finding | Evidence and limitation | Consequence |
| --- | --- | --- |
| The municipality had 5,281 residents and covers 173.44 km² | DGAL municipality record, population and area dated 2021 [S1]. This is the municipality, not the proposed delivery area or a current customer count. | Start with a compact route; measure reachable households. |
| Population is older and geographically dispersed | INE municipal profile reports 5,210 residents in 2022, including 1,850 aged 65+ [S2]. Different date from the census; do not combine these into a growth forecast. | Simple Portuguese interface, accessible text and later an assisted phone route. |
| Absence of delivery competition is unproven | An Uber Eats listing gives a Figueiró dos Vinhos address [S3]; a local directory advertises home delivery for Pizzaria Bar Manjericão [S4]. Listings may be stale or location labels inaccurate. | Test actual checkout availability at specific addresses and times. |
| Existing restaurants may already deliver or offer takeaway | Municipal material refers to takeaway/home delivery during 2020–2021 [S5]. Historical, not evidence of today's service. | Interview restaurants about current delivery and offer incremental orders, not an assumed monopoly. |
| Demand, willingness to pay and driver supply remain unknown | No interviews, paid orders or driving trials have taken place. | Treat all commercial figures below as hypotheses. |

Research was checked on 26 October 2026. No claim of exclusivity, guaranteed market size, current platform coverage or secured funding is justified.

### Address-level competition audit

With permission from address owners, test five representative locations: town centre, nearby residential streets, a peripheral address, and two outlying settlements. Try Uber Eats, Glovo, direct restaurant websites and telephone ordering on a weekday lunch, Friday dinner and Sunday dinner. Record date/time, address area, available restaurants, minimum basket, delivery charge, quoted time and whether checkout is actually possible. Do not place paid competitor orders without a defined research allowance. Search results alone cannot settle coverage.

## 3. Customers and the first offer

Prioritise households ordering dinner for two or more people. A family basket supports delivery economics better than a single low-priced meal. Secondary segments are workers arriving home late and visitors in accommodation near the route. Older residents and carers may benefit, but do not present this as clinical, social-care or emergency meal provision.

Initial offer to test:

- One restaurant per order; short, verified menu of dishes that travel well.
- Food minimum **€15**; initial delivery price hypothesis **€4.90 including applicable VAT**. No separate service fee. Packaging and any mandatory charges included in the displayed food prices.
- Friday, Saturday and Sunday evening windows: **18:30–19:15, 19:15–20:00, 20:00–20:45 and 20:45–21:30**, Europe/Lisbon.
- Order cutoff 90 minutes before the window. Start with two deliveries per window, shared across restaurants. Actual service requires route-tested capacity; a slot is not an assertion that every address can be reached.
- Proposed coverage: a route-tested central polygon, initially aiming for journeys of roughly ten road minutes from collection. That is a design hypothesis, not current coverage. A postcode or straight-line radius is insufficient.
- Restaurant commission hypothesis: **15% of the consumer food basket, excluding the commission's VAT; VAT is added to the restaurant's platform invoice**. Negotiate this explicitly and show settlement examples. Food remains the restaurant's sale; the platform's precise legal role requires advice.

Avoid permanent discounts, subscription promises, cash collections, alcohol and regulated goods in the first pilot. Online payment must use a regulated provider approved for the final commercial structure. Start paid delivery only once the flow is integrated and reconciles correctly.

## 4. Business model and financial discipline

All inputs are planning assumptions, not supplier quotes or verified local costs. Consult a Portuguese certified accountant before finalising invoicing, VAT treatment, CAE activity codes or legal form. The standard mainland VAT rate is 23% [S8]; whether each service attracts it depends on the final arrangement. Food, packaging and drinks can have different treatment [S9]. Do not apply a single VAT rate to the whole customer order by default.

### Per-order model

Let B be the food basket including its taxes, F the consumer delivery fee, C the commission percentage quoted before commission VAT, H fully loaded driver cost per paid hour, and D delivered orders per **paid** hour, including waiting, positioning and quiet periods.

```
platform revenue excluding VAT = B × C + F ÷ 1.23
variable costs = H ÷ D + 0.60 vehicle + 0.70 payments + 0.02 × B incidents + 0.35 support
contribution = platform revenue excluding VAT − variable costs
monthly result before income tax = orders × contribution − fixed costs
break-even orders = ceil(fixed costs ÷ contribution), only if contribution > 0
```

Driver labour and vehicle costs are separate assumptions. Do not count mileage twice if an operator's hourly quote already includes the vehicle. The payment allowance must be replaced by a provider quote including marketplace, refund, dispute and payout fees. An incident allowance is an expected cost, not a guarantee that cash refunds are capped.

For a **€26 basket**, **€4.90 delivery**, **15% commission**, **€14 fully loaded driver hour**, and **€850 monthly overhead and management allowance**:

| Deliveries per paid hour | Platform revenue/order | Variable cost/order | Contribution/order | Implication |
| --- | ---: | ---: | ---: | --- |
| 2 | €7.88 | €9.17 | −€1.29 | More volume at this productivity loses more money. |
| 3 | €7.88 | €6.84 | €1.05 | Roughly 812 orders/month to cover €850; verify exact model output. |
| 4 | €7.88 | €5.67 | €2.21 | Roughly 384 orders/month; high route density is still required. |

The €850 fixed-cost assumption comprises €500 owner/management time, €100 accounting, €80 insurance, €70 software, €20 phone and €80 other overhead. These are allowances, not quotes.

The full model and exact rounded outputs are generated in `docs/financial-scenarios.csv` by `scripts/financial-model.ts`. The website includes an interactive version. A €26 basket plus delivery produces €30.90 of customer spending; **€30.90 is not platform revenue**. The commission VAT is a tax liability, not extra profit. The restaurant's settlement depends on the contracted invoice/payout structure and provider fees, not merely B minus 15%.

### Capacity is a constraint, not a spreadsheet assumption

The initial cap is **8 orders/evening**. At 12 evenings per four-week planning month that is **96 orders**, before cancellations. A paid 18:00–22:00 shift means 48 driver hours and only **2 deliveries per paid hour** at full capacity. It is inconsistent to assume three or four per hour while keeping this exact staffing and capacity unchanged.

At 96 orders: net platform revenue is about €756.84; driver coverage €672; other variable allowances €208.32; contribution after driver coverage about **−€123.48**; less €850 overhead/management gives about **−€973.48** before income tax. At 50 orders with the same coverage, the loss is greater. Owner driving is still economic labour even when no cash wage is drawn.

Therefore the initial pilot cannot be sold as a self-supporting launch. It tests whether short shifts, grouped collections, larger family baskets, realistic delivery fees or contracted restaurant delivery capacity can produce at least **€2 contribution per order** and sufficient volume. Any increase from two to three orders per window requires measured safe travel/preparation times and revised promises. Do not assume four deliveries per hour without route evidence.

### Proposed learning budget and cash control

Proposed total pilot exposure ceiling: **€3,500**, requiring funded cash before commitments. No spending has been authorised with a payment source or executed.

| Planning allocation | Amount | Basis |
| --- | ---: | --- |
| Company registration | €220 | Published minimum for standard online incorporation [S6]; legal form not yet chosen. |
| Accountant setup | €250 | Assumed, obtain written quote. |
| Legal/contract review | €400 | Assumed, obtain scope and quote. |
| Insurance initial provision | €200 | Assumed; cover and premiums unverified. |
| Thermal bags, seals and hygiene kit | €180 | Assumed; exclude vehicle purchase. |
| Website/tools and test payments | €100 | Allowance; check actual service pricing. |
| Pilot printing/materials | €50 | Small, tracked distribution. |
| Four-week operational loss allowance | €1,300 | Planning buffer; reconcile actual paid shifts and overhead. |
| Unallocated contingency | €800 | Only draw against documented need. |
| Total | **€3,500** | No debt or guaranteed funding assumed. |

Keep a weekly cash forecast: opening cash + platform receipts − restaurant liabilities/payouts − driver costs − provider fees − taxes reserved − overhead − refunds = closing available cash. Restaurant funds and VAT are not working capital to spend. Freeze new slots if cash cannot meet committed obligations and the next two weeks of operating costs. Use verified actual provider settlement delays and refund timing.

Funding priority: a strictly capped founder-funded experiment if affordable. Investigate local municipal business support and current IEFP/IAPMEI programmes as optional leads; do not assume eligibility, approval or grant income. No debt is needed to validate 50 orders.

## 5. Phases, sub-phases and release gates

### Phase 0 — Foundations and prototype: current increment

0.1 Document the hypothesis, sources, costs and unknowns. 0.2 Build a Portuguese mobile-friendly ordering experience. 0.3 Exercise the entire simulated order lifecycle and failure cases. 0.4 Publish privately for the owner to inspect.

Deliverables: this guide; detailed backlog; source code; server-backed simulated orders; test evidence. Exit: owner can select a menu, create a demo order, reload it, simulate each valid status and inspect its history. This does not permit real commerce.

### Phase 1 — Customer and supply validation: proposed weeks 1–2

1.1 Complete the address-level competition audit. 1.2 Interview **30 residents**, including at least ten households with children, ten workers/commuters and ten older residents or carers; these are recruitment quotas, not a representative statistical sample. Ask about their last actual takeaway, how they got it, total cost, delivery problems and preferred windows. Test a specific all-in price; avoid only asking “would you use an app?”. Obtain separate permission for future contact.

1.3 Interview **eight restaurants**; record travelling menu, packaging, peak capacity, acceptance method, allergens, commission tolerance, current direct delivery and willingness to sign. 1.4 Drive candidate routes, measure collection waiting, parking and door handover, and map the initial polygon. 1.5 Obtain staffing, insurance, accounting and payment quotes.

Gate: at least **three documented partner expressions of interest** and **20 reachable households willing to test at the displayed prices** in the same service area; a viable operator and backup; draft route/cost evidence. These are lead indicators; only paid orders validate demand. If the fee is unacceptable or coverage already meets the need, revise the offer before further spending.

### Phase 2 — Commercial and operational readiness: proposed weeks 3–4

2A Business: choose legal form and CAEs with an accountant; identify owners/directors; perform name/trademark clearance; establish company/bank/accounting arrangements, required registrations and insurance. Any incorporation or bank onboarding needs actual identity, signatures and funds; software cannot substitute for these.

2B Supply: sign restaurant agreements; verify business details and operating authority; obtain approved menu/ingredient/allergen records and photographs; agree prices, packaging, acceptance timeout, refund allocation and payout schedule. Sign an appropriate operator agreement with real employment status assessment.

2C Product: separate customer/restaurant/dispatcher/courier permissions; implement real address coverage, merchant acceptance, hosted payments, webhook processing, refunds, invoices, notifications and stock/availability controls. Add a global ordering pause, timeouts, backup/restore drill and reconciliation reports. Keep payments and live mode disabled until every launch gate is met.

2D Rehearsal: **ten complete dry runs**, including merchant rejection, exhausted window, unavailable driver, wrong address, lost internet, duplicate payment event and food-safety incident. Record timings. Validate the operating capacity; lower it if needed.

Gate: signed partners and operator, working support contact, verified prices/allergens, appropriate legal documents and insurance, payment and accounting reconciliation, tested safety procedures, ten dry runs, documented owner go/no-go. All open blockers in `docs/RELEASE_CHECKLIST.md` must be closed with evidence.

### Phase 3 — Paid, invitation-only pilot: proposed weeks 5–6, extend up to four weeks if needed

3.1 Invite the initial consenting households in batches. 3.2 Run three evenings/week inside the polygon, staffed throughout ordering and delivery. 3.3 Start at two deliveries per window and close windows when capacity is reached. 3.4 Review every shift and contact customers through the agreed support process when delayed. 3.5 Track 50 paid, genuinely fulfilled orders; separate free/internal test orders.

Gate: at least 50 paid orders, **≥90% delivered within the booked window**, **≥95% complete/correct**, no unresolved serious safety issue, all settlements reconciled, and actual cost data for every paid shift. Report denominators: with 50 orders, one failure changes a metric by two percentage points. A 50-order sample cannot prove long-term safety or retention.

### Phase 4 — Repeatability and sustainable economics: proposed weeks 7–10

4.1 Measure 30-day repeat purchase among customers whose first order is at least 30 days old; report cohort size. Target ≥30%, with no claim of significance for a small cohort. 4.2 Seek **≥€2 contribution/order**, including paid idle time, mileage, support, payment fees, refunds and discounts. 4.3 Improve collection coordination and basket value; test one pricing or window change at a time. 4.4 Reconcile actual P&L and cash every week.

Gate: four consecutive weeks meeting operational targets, positive measured contribution, a credible funded path to full overhead/management break-even, and at least two backup fulfilment options. If contribution stays negative after two bounded interventions or customers will not pay sustainable prices, pause expansion and decide whether to pivot or close.

### Phase 5 — Controlled public release and expansion: week 11 onward only if earned

5.1 Open publicly only after commercial readiness and repeatability gates. 5.2 Add **one** additional service window or contiguous zone; fund and staff it first. 5.3 Review each change for four weeks. 5.4 Add merchant self-service and courier tooling only when manual operation becomes the measured bottleneck. 5.5 Consider adjacent towns, groceries or corporate lunches only with separate demand/capacity evidence.

Success threshold: cover all labour, management, overhead, taxes and contingencies with positive operating cash for at least three months, while meeting service standards and retaining partners. Do not use GMV or app downloads as substitutes for profit or service quality.

## 6. Product and user scenarios

**Core functionality:** schedule and follow a single-restaurant meal order. The complete first increment includes browsing, cart, transparent quote, capacity-limited slot, validated demo checkout, persistent order, guarded state transitions and audit history. User scenarios and acceptance criteria are in `docs/USER_SCENARIOS.md`.

A household selects two meals, sees all allergens and the final price, chooses an available window, submits once, and receives a durable reference. Repeating after a lost response returns the same order. A dispatcher must confirm the restaurant and rider before accepting a real request. A customer sees recorded states, never fabricated GPS progress. A rejected/unfulfillable real request releases capacity and follows a tested void/refund policy.

Current prototype: three fictional menus, €15 minimum, €4.90 example delivery, one demo zone, account-scoped data, manual simulated operations. No live charge, messaging, merchant registration, real dispatch or live GPS. “Delivered” in the prototype means a user simulated that state. Production customer cancellation rules are a separate legal/product decision; demo users can cancel their own nonterminal orders for testing.

Technical approach: React/TypeScript via the supplied Vinext/Sites runtime, Cloudflare D1, schema migrations, server validation and integer-cent prices. Cart preferences are browser-local; orders are durable. Authentication uses the private host's identity. Final public customer login should be appropriate for local residents; do not assume a ChatGPT account is the eventual checkout requirement. Every real operational role needs explicit server-side authorisation.

## 7. Operating model

The founder is accountable for business commitments. AI can research, prepare plans, build software and assist analysis; a named person must handle food, driving, signatures, customer support and professional sign-off.

| Role | Accountable work before pilot | Operating cadence |
| --- | --- | --- |
| Owner/CEO | Funding, legal entity, go/no-go, partner terms | Weekly P&L, runway and service review |
| Local operations lead | Routes, safe collection/delivery, support and backup | Before, during and after every shift |
| Restaurant | Verified menu/allergens, preparation, sealed food, food invoice | Confirm availability before each shift |
| Delivery operator | Hygiene, lawful vehicle/work arrangements, pickup/handover evidence | Every order and paid shift |
| Certified accountant | CAE, VAT/invoicing, payroll, settlements and filing calendar | Setup, weekly reconciliation, statutory deadlines |
| Portuguese legal adviser | Contracts, platform/worker status, consumer/privacy/payment role | Before live launch and material changes |
| Product/engineering | Tested flow, access control, monitoring, recovery | Each release and incident |
| Sales/marketing | Partner onboarding and permission-based acquisition | Small experiments with cost and conversion measured |

The shift SOP and exception playbook are in `docs/OPERATIONS.md`. Dispatch records customer promise, restaurant acceptance, preparation estimate, courier availability and any changed ETA. A server receipt is not merchant acceptance. For live service, propose a three-minute merchant acceptance target and a five-minute rejection timeout; test them with partners and implement timers before launch.

## 8. Sales and marketing

Restaurant pitch: incremental local dinner orders with clear limits, a small verified menu and transparent settlements. Discuss economics rather than promising volume. Prefer a four-week non-exclusive pilot, no permanent fee waiver and no listing of names/logos before agreement. Potential leads from research include Pizzaria Bar Manjericão and the restaurant appearing in the Uber Eats listing; current details and willingness to partner are unverified.

Customer message to test: “O jantar da sua vila, à hora combinada. Veja o preço completo antes de pedir.” Explain the actual coverage and window. Do not advertise “only delivery service”, “always available”, or a 20-minute promise without evidence.

Acquisition sequence: partner-counter QR leaflets with consent, accommodation partners within the route, and existing local community channels where posting is permitted. No unsolicited bulk messaging or scraping of residents. Prepare posts/outreach as drafts until actual recipients, identity and channel permissions exist. The user's broad autonomy does not create restaurant contracts or access to local community groups.

Measurement: source → menu visit → checkout → paid fulfilled order → 30-day repeat. Customer acquisition cost equals attributable spend divided by first-time paid customers, including discounts. Compare it with realised cohort contribution, not the whole food bill. Propose a €50 initial print test; increase only after conversion and payback are measured. No paid advertisements or external messages have been sent.

## 9. Legal, tax, trust and accounting workstream

This section is an implementation brief for qualified local advisers, not a declaration that the business is compliant. Official sources are dated or evolving and must be checked against the final operating model.

1. **Entity and activity.** Compare sole trader vs unipessoal Lda with an accountant; a limited-liability company is the working candidate if the marketplace/fulfilment arrangement warrants it. Confirm current CAE classification, start-of-activity obligations, registrations, RCBE where applicable and business insurance. Online company creation requires authentic signatures/identity [S6]. Commercial companies require organised accounting [S7].
2. **Commercial role and funds.** Decide who sells food, who sells delivery, who invoices whom and who receives each payment. Use a licensed payment provider configured for the agreed marketplace/agency structure; do not casually collect and redistribute merchants' money from an ordinary bank account. Assess payment-service implications and provider onboarding with counsel. Implement provider-signed, idempotent webhooks, refund reconciliation and clear liability allocation before accepting funds.
3. **Food safety.** Verify the applicable food-business/transport obligations with ASAE and advisers. Adopt documented hygiene/HACCP-based procedures appropriate to handling and transport, clean insulated carriers, hot/cold separation, tamper-evident packaging, time/temperature limits approved by a food-safety professional and traceability. Restaurant food preparation compliance alone does not discharge the delivery operator's duties. No arbitrary universal temperature guarantee is encoded in the prototype.
4. **Allergens and product information.** Obtain accurate ingredient and allergen records from each restaurant; display mandatory information before purchase and provide it at delivery. ASAE explains distance-sale information requirements [S10]. Review cross-contact wording; a free-text “allergy note” cannot make an unsafe order safe. The prototype's fictional allergens cannot be reused commercially.
5. **Consumer terms.** Identify the legal trader, address, NIF, contact route, total prices, delivery conditions, complaint process, cancellation/refund rules and applicable dispute-resolution information. Determine electronic complaints-book registration and display duties [S11]. Perishable/prepared-food exceptions need accurate drafting; do not promise a blanket 14-day return period or eliminate remedies for missing/unsafe goods. Check current ADR and platform disclosure requirements; do not add an obsolete EU ODR link by rote.
6. **Workers and vehicles.** Assess actual working arrangements under Portuguese law, including platform-work presumptions. Article 12-A-related case law demonstrates that calling someone a contractor does not resolve classification [S12]. Confirm accidents-at-work cover, social security/payroll, vehicle insurance covering business delivery, vehicle documentation and lawful hours. Do not model unpaid labour as profitability.
7. **Privacy/security.** Document controller/processor roles, lawful bases, data minimisation, service-provider agreements, access roles, retention, deletion and rights handling [S13]. Use contract necessity for order fulfilment where appropriate; separate optional marketing consent. Avoid collecting health information for the first pilot. Restrict addresses to the order's assigned parties; do not expose them in logs, broad dashboards or marketing. Confirm transfers/subprocessors, incident response and notification obligations with counsel. The prototype is private and requests fictitious data; it is not approved for real customer data.
8. **Accounting.** Separate food GMV from earned revenue; reserve VAT and restaurant liabilities; reconcile daily provider receipts, merchant invoices, platform invoices, refunds, fees and payouts. Use accounting-approved compliant invoicing software and fiscal document workflows, including required reporting/identifiers as applicable. Maintain an accountant-owned deadline calendar. Do not label an order confirmation as a fiscal invoice.
9. **Brand/IP/accessibility.** Search RNPC/INPI and relevant EU marks before using Ofigo commercially; check domain availability and rights to food photos, logos and menus. Obtain permissions. Test keyboard navigation, mobile forms, contrast, low-bandwidth use and screen-reader labels. Determine applicable accessibility obligations for the final business and service.

## 10. Decisions and risk register

| Risk | Early signal | Response/decision |
| --- | --- | --- |
| Demand too low or too dispersed | Few paid repeat orders inside one route | Shrink windows, group preorders; stop expansion. |
| Driver coverage costs exceed contribution | Low orders per paid hour | Measure full shifts, change fee/basket/window; do not omit idle time. |
| Restaurants slow or unavailable | Rejections/waiting spill into booked window | Pause partner, reduce slots, agree short menu/cutoffs. |
| Food safety/allergen incident | Seal/temperature/information failure or complaint | Stop affected orders/partner, preserve traceability, follow professional response plan. |
| No rider/vehicle backup | A single absence causes cancellations | Named replacement and ordering pause before every shift. |
| Duplicate orders/payment events | Retries, timeouts, webhook replays | Idempotency, durable references and provider reconciliation. |
| Cash/settlement mismatch | Provider payout delayed; merchant funds spent | Segregated ledger, refund reserve and cash forecast. |
| New or existing competition | Address audit shows satisfactory availability | Differentiate on confirmed local gaps, do not assert monopoly. |
| Public prototype mistaken for a live service | Someone expects actual fulfilment | Private audience and visible simulation labelling; no live switch. |

If sustainable delivery is not possible, bounded alternatives are restaurant-owned delivery coordination, scheduled family-meal drops, or a takeaway preorder tool with merchant subscription. Each needs fresh price/demand validation; do not quietly change customer promises.

## 11. Next concrete work

1. Test the private product end to end using fictitious information.
2. Run the address audit, 30 resident interviews and eight restaurant interviews; log findings with permission.
3. Secure three serious partner leads and a named operator/backup; collect route and cost evidence.
4. Obtain accountant/legal/insurance/payment-provider quotes and decide whether the capped pilot can be funded.
5. Implement the live-readiness backlog and close every release blocker before inviting paid orders.

No ongoing autonomous monitoring or future execution is scheduled by this guide. The repository and backlog preserve work for the next iteration.

## Sources

These are evidence sources, not partner endorsements. Research accessed 26 October 2026.

- **[S1]** DGAL, [Figueiró dos Vinhos municipality](https://portalautarquico.dgal.gov.pt/pt-PT/entidades-locais/concelhos/figueiro-dos-vinhos/), population and area dated 2021.
- **[S2]** INE, [municipal socioeconomic profile, December 2023](https://www.ine.pt/documentos/municipios/1008_2023.pdf), population data for 2022.
- **[S3]** Uber Eats, [Churrasqueira e Pizzaria listing](https://www.ubereats.com/pt/store/churrasqueira-e-pizzaria/klRRo_6DU9iz9CaRdWnclw). Listed address does not verify deliverability.
- **[S4]** Omnia/VIP360, [Pizzaria Bar Manjericão](https://vip360.pt/figueirodosvinhos/scene_87_pt.html/scene_127_pt.html). Directory advertises home delivery; confirm directly.
- **[S5]** Municipality, [historical business support measures](https://www.cm-figueirodosvinhos.pt/index.php/medidas-municipais-de-apoio). 2020–2021 context only.
- **[S6]** Justiça, [online company creation guide](https://justica.gov.pt/Guias/Como-criar-uma-empresa-online) and [published costs](https://justica.gov.pt/Como-criar-uma-empresa-online/Quanto-custa).
- **[S7]** Justiça, [certified accountant and organised accounting](https://justica.gov.pt/Como-criar-uma-empresa-online/Escolher-o-contabilista-certificado).
- **[S8]** Tax Authority, [CIVA Article 18](https://info.portaldasfinancas.gov.pt/pt/informacao_fiscal/codigos_tributarios/civa_rep/ra/Pages/ivara18_062022.aspx).
- **[S9]** Tax Authority, [VAT questions including takeaway packaging](https://info.portaldasfinancas.gov.pt/pt/apoio_contribuinte/questoes_frequentes/Pages/faqs-00930.aspx).
- **[S10]** ASAE, [foods sold online: distance-sale information](https://www.asae.gov.pt/ficheiros-externos-newsletter/asaenews-n-112-alimentos-vendidos-pela-internet-rv-pdf.aspx); DGAE, [restaurant guide, 2025 edition](https://areadocomerciante.dgae.gov.pt/do-seu-interesse/guia-da-restauracao.aspx). Assess transport/platform duties separately.
- **[S11]** ASAE, [Livro de Reclamações](https://www.asae.gov.pt/reclamacoes-e-denuncias/livro-de-reclamacoes.aspx).
- **[S12]** Diário da República, [Guimarães appeal decision of 17 October 2024, platform-work presumption](https://diariodarepublica.pt/dr/detalhe/acordao/2843-2024-930204075). Case-specific; assess actual facts and current law.
- **[S13]** European Commission, [data protection principles](https://commission.europa.eu/law/law-topic/data-protection/information-business-and-organisations/principles-gdpr_en); CNPD, [DPO obligations](https://www.cnpd.pt/organizacoes/outras-obrigacoes/encarregado-de-protecao-de-dados/).