Research · 2 October 2026

Evidence behind the report.

Detailed notes from current official provider pages, live public configurators and a website audit. These are point-in-time observations with purchase, delivery and availability limits made explicit. Customer source emails, addresses and full recipe PDFs are excluded.

For the audited ingredient reconstruction, use the main report’s full ledger and downloadable ingredient data. For the analysis and editable financial assumptions, return to the interactive report.

Source notes · Portugal competitors

Portugal competitor evidence — 2 October 2026

Research scope: Natalie's Healthy Kitchen, a fortnightly ingredient/recipe bag containing 3 recipes for 4 people (12 advertised portions). All commercial prices below were observed on official provider pages on 2026-10-02. No introductory discount is applied. These are consumer price observations, not a purchase or a guarantee of service to a specific postcode. No provider was contacted.

Findings that materially change the pricing discussion

1. Natalie's bag must be compared per 12 portions, not as a full week or fortnight of food. At €71/€76 this is €5.92/€6.33 per advertised portion. At €90 it would be €7.50; at €110, €9.17. Fortnightly billing does not mean 14 dinners are supplied.
2. The closest local ingredient-kit comparator, Mesa Fresca, publicly lists a 12-portion pickup box at €84, or €7 per portion, before any home delivery charge. Its 15-portion box is €109. The recipes-per-box and people-per-recipe differ from Natalie, so “12 adult portions” is the useful common denominator.
3. Prepared meals are a substitution benchmark, not evidence that a raw ingredient bag must have the same price. They remove the cooking task. Several prepared meal providers sell 250–350 g portions; others sell 450–550 g. Serving count alone hides a meaningful difference.
4. An accessible price ladder is already visible in the market: pickup, quantity discounts, small trial packs, optional extras, a paid subscription advantage, and commitment that makes production predictable. It is possible to serve several budgets without making every delivered bag cheap.
5. Public headline minimums can mislead. Dona Marmita's lowest price refers to a 270 g low-carb serving. Farm to Home's €7.50 minimum is a soup, while the visible main dishes are €12.50–€17.50. Food4Fitness packs sometimes combine main dishes and lighter meals; do not divide by the total number of items and call every item dinner.

Direct meal-kit comparison: Mesa Fresca

The live storefront also has 2-recipe boxes (€54 for 3 people; €89 for 5), even though parts of its FAQ still describe only 3- or 5-recipe options. Use the live product offer for price, and flag the FAQ discrepancy. Taxes are stated as included. Pickup is Monday 14:00–19:00 at Rua Silva Carvalho 50C, Campo de Ourique. The official service page confirms Greater Lisbon, Cascais and Costa da Caparica; Sintra-specific fulfilment has not been checked. These are raw ingredients to cook in about 30 minutes. Eggs, salt and olive oil are assumed pantry items rather than included ingredients. No subscription is required.

Sources: Official storefront and prices; official service explanation.

Inference: €90 for Natalie's delivered 12-portion bag would be close to Mesa Fresca's €84 pickup price plus delivery, if delivery quality, ingredient quality and portion adequacy are comparable. €110 requires a more distinct premium proposition and a paid willingness-to-pay test; it is not validated just by one enthusiastic customer. Mesa Fresca's larger box reaches €5.96 per portion, illustrating the need to separate “small box convenience” from “large household economy.”

Prepared-meal comparison — use portions and weights together

“12-equivalent” and “15-equivalent” are analytic scaling of a quoted unit price where the provider does not sell exactly that quantity. They are not checkout quotes. Actual purchasable packs are explicitly listed. Delivery is included only when the official page says so. Tax inclusion is explicit for Mesa Fresca and Zero Stress; other prices are displayed retail prices, with no separate VAT uplift applied, but explicit VAT wording was not verified.

Dona Marmita: accessible headline, materially different serving sizes

The official size selector shows the least expensive 12-pack at €75.27 for 270 g low-carb; Fit 340 g is €77.87, XL 550 g €103.82. The 15-pack equivalents are €94.09, €97.34 and €129.78. Comparing Natalie's family dinner to the 270 g minimum without a size caveat would be unfair. The provider offers nationwide mainland delivery in refrigerated vans, included in the price, with Tuesday and Thursday schedules outside Coimbra. Its preparation guide gives 7 days refrigerated or 3 months frozen storage and reheating in 2–3 minutes. This is a broad convenience/healthy meal option with a clear quantity ladder.

Sources: 12-pack size prices; 15-pack size prices; delivery policy; storage and preparation.

Food4Fitness: a good ordinary adult benchmark is L, not the smallest M

The clean main-meal comparator is 14 L meals, 350 g each, €125. Other regular offers are 10 M meals at €71, 10 L at €91 and 40 M at €265. M is 250 g, L 350 g and XL 450 g; this materially affects apparent affordability. No introductory coupon is applied. Subscription discounts are advertised as up to 10%, so they should be a separate committed-customer comparison, not assumed on a one-off order. Delivery costs are postcode-dependent below €90 and free above it. Orders arrive fresh/refrigerated, with 3–9 days remaining shelf life. The provider sells high-protein, low-carb and plant-based options and accepts eligible meal cards via supported payment routes. It has a Lisbon physical shop, adding a useful pickup precedent.

Sources: official regular packs; sizes visible on official product listing; delivery/storage/payment FAQ; store and positioning.

Prontíssima: explicit 300 g / 450 g distinction

Its exact 12-pack costs €89.90 at 300 g or €123.90 at 450 g. Both include delivery in mainland Portugal. The larger quantity option is 16 meals for €117.90 / €157.90 respectively. A first-order 10% reduction exists and is excluded here. Meals arrive refrigerated rather than frozen, with 5-day refrigerator guidance and the option to freeze upon receipt. The menu spans low-carb, balanced and traditional Portuguese food. There is no subscription commitment. Its family tray offer is approximately 2 kg for 4–5 people at €43.90, a distinct product that may be useful when assessing family-scale prepared alternatives; do not blend its serving estimate with the individual packs.

Source: official homepage, price ladder, family format and FAQ.

A Cozinha Verde: locally relevant vegan service and paid customization

The weekly offer is €46 for five vegan prepared meals, €82 for ten. A four-week ten-meal-per-week plan is €312 for forty (€7.80 per meal). Its five-per-week monthly option is €162 for twenty, which calculates to €8.10 despite the site's €8.00 label: use arithmetic, not the label. Lisbon delivery is free, Amadora/Loures/Odivelas €3, and Sintra/Cascais/Oeiras and listed south-bank zones €4.50 per delivery. Meals arrive in reusable boxes suited to refrigerator or freezer storage; weights are not published on this page. Optional swaps cost €1 and desserts €3 per portion. This demonstrates charging separately for extra operational complexity and optional treats. The page contains inconsistent Friday/Saturday ordering cutoffs; exact cutoff requires confirmation.

Source: official pack prices, fees and service terms.

A Marmita: prepared weekday lunch substitution

The official menu retrieved on 2 October already shows the 5–9 October week, demonstrating an actively rotating offer. Listed 400 g examples include vegetable curry and meat chili at €7.70, beef bolognese at €7.99, chicken tikka masala at €9.90 and cuttlefish feijoada at €9.99. These are examples, not a claim that every dish lies in this range. Its homepage states meals are quickly chilled to 3°C after cooking. The company's own LinkedIn advertises free delivery in Greater Lisbon, but the ordering system's route boundary and Sintra postcode were not checked. It provides business-specific terms, which suggests pooled office delivery as a useful route-density precedent. No bulk discount is assumed.

Sources: official current menu; official format description; company profile delivery claim; official business arrangements.

Farm to Home: premium food plus membership precedent

The current official page lists a €59 weekly subscription containing a loaf, two soups and two main dishes; it should not be treated as four interchangeable dinner mains. It includes mainland delivery and a continuing 25% discount on additional à-la-carte items. Standalone visible main dishes are €12.50–€17.50, generally around 500 g; the €7.50 headline minimum is a soup. Standalone orders require at least three meals; shipping is €20 below six and free from six. The subscription allows pauses/cancellation and offers farm experiences and loyalty rewards. It is a premium positioning example, not a mainstream low-price competitor. Its website describes regenerative farming and seasonal production; those are provider claims, not audited findings. Introductory 25% discounts are excluded.

Source: official current offer and FAQ.

Optional seventh prepared competitor: Zero Stress

This Amadora-based provider advertises frozen, individually customized meals up to 350 g. Regular one-off prices are €69/5 (€13.80 each), €135/10 (€13.50) and €260/20 (€13). Monthly advance-payment tiers are €240/20 (€12), €456/40 (€11.40) and €616/56 (€11). These are VAT-inclusive. Pickup at Rua de Angola 2A, Amadora is free; free delivery is described for Lisbon/Amadora/Odivelas at 20+ meals, with shipping quoted separately for other regions. Some plan cards instead promise a 10 km free-delivery radius, so Sintra/Cascais eligibility must be checked. The page is a live advertised offer; provider scale, testimonials and nutritional benefits have not been independently verified. Use only as a small-business premium reference, not a representative market median.

Source: official Zero Stress pricing and delivery details.

Supermarket anchors: accessibility cannot be judged only against premium delivery

These observations are listed product prices with no home-delivery fee included. They represent cheaper customer substitutes; flavour, ingredient quality, portion adequacy and variety have not been taste-tested. The provider says product selection can change with the delivery address, so availability at a specific store or postcode remains unverified.

The family pie gives approximately 417 g per advertised person, and family duck rice approximately 333 g. The frozen €2.99 duck rice is a separate product from Cozinha Continente's €3.99 ready meal. Keep the two distinct. The chicken/quinoa Equilíbrio product is not vegetarian. No introductory discount is shown on these product prices. A balanced meal may require a salad or other additions; those are not included in this table.

Sources: spaghetti bolognese product; 300 g duck rice; chicken curry/quinoa; family duck rice; family pie; frozen duck rice; official prepared-meal category.

Inference: A €90–€110 ingredient bag is accessible to a convenience-oriented segment, not automatically to an ordinary low-income household. The supermarket prepared-meal anchors already reach roughly €48–€62 for twelve small individual servings before shopping/delivery. An affordable digital tier is therefore more defensible than claiming the premium physical bag is affordable for everyone. The comparison also shows why Natalie's product needs to articulate fresh ingredients, variety, trusted curation and family cooking, rather than win only on immediate calories per euro.

Exclusions and research limits

• EatTasty: historical price articles are widespread, but the official site could not be fetched and secondary customer discussions report a closure. Do not list it as an active competitor or reuse its 2018–2021 prices.
• Go Natural: official restaurants and delivery channels were verified, but a directly comparable regular home meal-pack price was not established. Relevant to eating out, less relevant to a 12-portion ingredient bag.
• Eat Fit: an official pack page describes frozen high-protein meals, but its current site returns 403 and no reliable current regular pack price was obtained. Avoid filling the gap with an old press article.
• Availability, final tax treatment and checkout charges have not been tested using a private address. Retail displays are point-in-time observations. Site menus may change faster than published explanatory text.
• Family portions and 300 g diet meals should remain visually distinct in the interactive report. A comparison calculator can offer portion-count and portion-weight filters, but it should never imply nutritional equivalence.

Business implications — original analysis, not provider facts

• €90 is a credible test point: Natalie at €7.50 per advertised portion falls near local pickup ingredient-kit pricing after adding a reasonable delivery charge. That supports a test, not a guaranteed demand prediction.
• €110 can be a premium option: at €9.17 it enters the territory of larger prepared meals. Premium delivery must justify cooking still required through ingredient quality, curation, household fit, personal service, flexible skipping, recipe access and optional value. Avoid promising new labour-heavy customization to justify a price rise.
• Access should come from a simpler product: shopping-list/recipe access can address budgets without the ingredient sourcing, packing and delivery burden. Pickup can reduce fulfilment cost but saves only the avoidable route cost, not all ingredient-kit labour.
• Predictability is monetizable: a fixed ordering deadline, paid reservation, limited recipe changes and planned route windows improve costs. Farm to Home and Cozinha Verde show how member benefits and customization fees make that visible to customers.
• Supporter tiers need cost control: recipe archive access and a thank-you/community role are cheap marginal benefits; extra physical treats create ingredient and labour costs. Model supporter income separately from food gross margin.
• Do not infer Natalie's margin from competitor prices: competitor purchasing scale, meal weights, tax mix, staffing, spoilage, route density and marketing costs are not known. Price, subscriber retention and true unit costs belong in the sliders as editable assumptions.

Source notes · International competitors

International meal-kit benchmarks — checked 2 October 2026

Decision-useful finding

Natalie's public €71 vegetarian / €76 mixed bag for 3 dinners × 4 people is close to the delivered recurring price of large European meal-kit brands. Those prices do not, by themselves, prove €70 is unusually cheap. A higher boutique price needs a clear promise: carefully developed seasonal food, personal service, local sourcing, reliable delivery and fair compensation for the founder. A €90 or €110 bag may be sustainable and appealing to some existing customers, but willingness to pay must be tested across the customer base. One enthusiastic customer's €110 ceiling is evidence to interview further, not a market-wide demand estimate.

Compare one 12-portion order, not one week of food. Natalie's bag arrives fortnightly; the brands below normally offer weekly schedules that can be skipped. A €76 fortnightly bag is 26 × €76 / 12 = approximately €164.67 per average calendar month, if all 26 deliveries are taken. It still provides only three family dinners per fortnight, not fourteen.

Like-for-like recurring prices

All configurations below are 3 recipes × 4 people = 12 portions. Introductory discounts, optional premium recipes and groceries outside the kit are excluded. Europe-to-euro conversion uses the ECB 2 October 2026 reference rate: EUR 1 = SEK 11.2900; GBP 0.85033. Consumer prices are used without adding or removing tax.

Important source conflict resolved: Quitoque's indexed marketing pages give €5.99/portion for 3 × 4. The live selector on 2 October instead displays €6.25/portion and €74.99 per order. Use the live €74.99, not 12 × an older rounded advertising number.

Shipping cautions: HelloFresh Sweden's standard SEK 49 is what the public configurator showed; its welcome page says some high-demand delivery windows incur an additional cost. Quitoque's widest 10-hour home window is €2.99; a 5-hour window is €3.99–€4.99 depending on location/time, 4-hour and 3-hour windows €5.99, 2-hour €7.99. Remote areas can incur an additional fee. Linas' terms specifically say its delivery fee depends on area and option. Do not silently add the often-quoted SEK 39 from comparison sites; it was not independently verified here.

Exact source ledger and observation details

Sweden — HelloFresh

• Price: official plans — checked live UI, final result URL step 14. Entered the public central-Stockholm postcode 11122 solely to check coverage; no personal address, account or order was submitted. Final selected values: 4 portions per dinner, 3 dinners per week. Displayed food SEK 786, shipping SEK 49, total SEK 835. Displayed SEK 65.50/portion is before delivery; delivered is SEK 69.58/portion.
• Tax: terms §7.1 states prices include VAT; the live selector separately itemises shipping.
• Service: how it works and homepage describe pre-portioned ingredients and recipes, delivery, skip/pause/cancel, 75 recipes per week and digital recipe cards. These are uncooked ingredient kits.
• Delivery windows: welcome page explains optional extra cost for some delivery times.
• Public recipe library: recipes. An archive alone therefore faces free alternatives.
• Intro offer: homepage currently promotes savings up to SEK 1,039 plus snacks for three months; excluded from the benchmark. No promotion was applied to the observed plan.

Sweden — Linas Matkasse

• Price: official registration selector, reached from its own homepage's order link. Selected 4 portions and 3 dinners. UI displayed SEK 637 intro / SEK 910 original recurring price, with a 30% promotion. SEK 53.08 advertised portion price is promotional, not the recurring SEK 75.83 before shipping.
• Shipping and flexibility: terms, last updated 15 July 2026; sections 4.4 and 6.7 state current fee varies by delivery area / option. Registration requires account creation before the address stage, so the fee was not checked. No account was created.
• Homepage offers 150 recipes each week and add-on groceries, fruit, breakfasts and snacks, with no fixed-term commitment. Some older FAQ copy on that same page still says 80 dishes; the more prominent current headline says 150. Avoid relying on menu-count detail for the price comparison.
• Public recipe bank provides a free archive. Terms §3.1 say paused customers retain order history and recipes.
• VAT: amounts are consumer-facing displayed prices; explicit VAT wording was not found in the current public terms. Do not claim a specific Swedish VAT rate from this research.

UK — Gousto

• Official price summary, updated 10 July 2026, explicitly lists full standard prices with delivery / introductory discounts excluded. 4 people × 3 recipes: £47.25 food, £3.94 rounded food-only per serving, plus £3.99 shipping. Delivered: £51.24 / 12 = £4.27/portion. The official price list supports the price directly.
• Provider's published comparison observed 3 September 2026 corroborates Gousto £51.24 including £3.99 shipping. It reports other providers' prices, but those rival figures are competitor claims, not our independently verified quotes; they are not in the main table.
• Published standard table also shows single-person boxes, scale economies as portions / recipes increase, and skip/pause/cancel flexibility. Current new-customer advertising is 50% first order, 40% second, 20% following six; excluded from recurring prices.
• Premium surcharges: the provider's comparison explicitly excludes premium-recipe surcharges. Benchmark is standard recipes only. No unverified fixed premium surcharge is claimed.
• VAT: public consumer box prices are used as shown; no tax-exclusive conversion or additional VAT estimate is applied.

Germany — HelloFresh

• Price: official plans, live final result step 7. Public Berlin postcode 10115, 4 portions × 3 recipes, no promotion applied: €66.50 food, €5.99 delivery, €72.49 total. The displayed €5.54/portion excludes shipping. Delivered €6.04/portion.
• Terms §8, updated 21 July 2026, explicitly include applicable VAT. Extra meals and add-ons can be purchased without permanently changing subsequent orders (§5).
• The website advertises savings up to €120 plus a knife; no such introductory offer is included in the comparison.
• Business vouchers / staff meals: gift vouchers and one-off bulk deliveries to an employer site or individual homes show a B2B / group-order channel. Do not infer Natalie's demand for this without a pilot.

France — Quitoque

• Price: official box selector, live 4 people / 3 meals = €6.25 rounded per portion; €74.99 weekly food price. Compared before any voucher.
• Shipping: official explanation of prices includes an image fee table; inspected full-size official tariff graphic. Home delivery €2.99–€7.99 by window, partner pickup €1.99, with geographic availability and remote-area caveats. The official FAQ says exact fee depends on address / delivery day.
• Tax: conditions of sale §3.3, updated 21 July 2026, specifies tax-inclusive euro prices.
• Service: official box page offers 44 recipes each week, seasonal produce, home or pickup service, pause/cancel and extra market products. This is an uncooked ingredient kit.
• Intro discount: 3 × €25 is marketed subject to offer conditions / minimum order. Separate from recurring price; not used above.

Foreign exchange

• ECB daily reference XML fetched directly on 2 October 2026: dated 2026-10-02, GBP 0.85033, SEK 11.2900, USD 1.1225 per EUR 1. Conversion rule: foreign-currency price / reference rate = euro comparison price.
• The search-indexed HTML displayed older 30 September data and the PDF 1 October data, so neither was used. Current XML is the authoritative live observation.
• ECB methodology says reference rates are for information and generally updated on working days; real card conversions can differ.

A digital budget tier has a real precedent

Sorted Sidekick official pricing FAQ states UK pricing for new customers after 13 May 2025 is £6 monthly or £60 annually (approximately €7.06/month or €70.56/year at today's ECB rate). Outside-UK prices are localised, so do not present the UK euro conversion as a Portugal checkout quote. Its service description sells coordinated three-recipe packs, one combined shopping list, portion scaling from 1 to 10, cooking guidance and household sharing. Customers buy food separately and choose their own store. This is strong evidence that Natalie's proposed shopping-list tier is an established product category.

For a second example, eMeals quotes its US dinner planning product from $4.99/month when purchasing 12 months, or $35.99 for three months, with weekly recipes and grocery lists. Groceries are separate. The page contains inconsistent trial durations (7 and 14 days), so do not cite a trial length. eMeals' US grocery integrations are a model to study, not an immediately available Portuguese integration.

Suggested hypothesis for Natalie, not an existing competitor fact: test €5–9/month for a seasonal three-dinner plan, a consolidated shopping list organised by store aisle, Portuguese supermarket substitutions, portion adjustment, and a searchable archive. A grocery-cost target and pantry checklist would make this stronger than merely charging for PDFs. Existing delivery customers can receive it as part of their bag price; digital members can upgrade for convenience.

Practical mechanisms to borrow

1. Price convenience separately. Quitoque proves that pickup and wider delivery windows can be less expensive than narrow home-delivery windows. A local partner pickup point can work if the saving actually exceeds partner fees, staffing, storage, customer collection failures and added travel. Test one point / one collection window before multiplying locations.
2. Digital access keeps a lower-spend customer in the business. Sell planning and confidence; the customer does the buying. Track digital support time and cancellation just as carefully as food contribution. New recurring revenue is not automatically effortless or free.
3. Offer portion / frequency flexibility, with operational limits. Big brands make skipping / pausing normal. For a founder-led service, a deadline and one standard bag reduce waste and procurement uncertainty; unlimited customisation could erase the margin gain.
4. Attach low-complexity add-ons. Linas sells breakfast, snacks, fruit and groceries; Quitoque has a market. Natalie should start with one shelf-stable or predictable product only if supplier margin and packing time are positive. A wider catalogue adds spoilage and admin costs.
5. Supporters can pay for belonging and continuity. Optional membership could offer behind-the-scenes updates, recipe voting or a periodic cooking session. Its price is voluntary and additional; it should not be counted as guaranteed food margin or promise unlimited founder attention. A clear recurring service proposition is stronger than a permanent rescue campaign.
6. Archives are a bundle benefit, not sufficient differentiation. HelloFresh and Linas already expose recipes publicly, and Linas preserves history for paused accounts. The valuable digital layer is coordinated menus, local substitutions, portion arithmetic and shopping lists.
7. Investigate group / employer orders. One collection drop for several households may improve route economics and acquisition costs. An employer or coworking-space pilot could pay for a wellbeing benefit or sponsor affordable subscriptions. Demand remains unvalidated; do not build a large B2B operation before testing.

Cross-country affordability must remain a separate question

Nominal euro prices standardise currency and servings; they do not standardise wages, rent, taxes, household size, food preferences or service expectations. Eurostat's 2025 AIC per-capita PPS table puts Portugal at 86, Sweden 103, France 103, and Germany 120 (EU = 100). AIC measures consumption volume after price-level adjustment, including services provided by government and nonprofits. It is useful background showing different household material conditions, not a salary or household disposable-income measure. Do not mechanically rescale a Swedish meal-kit price by these indices.

An accessible offer means a lower cash commitment and meaningful customer choice. Test a digital plan, fewer dinners / less frequent bags, pickup and a small sponsor-funded quota. Do not assume the premium delivered bag has to fit every household budget, and do not subsidise all bags through unpaid founder labour.

Source notes · Website and SEO

Natalie’s Healthy Kitchen — public website and SEO audit

Research date: 2 October 2026, Europe/Lisbon. This document uses public website evidence only. No private customer email, account, home address, confidential financial information, or unpublished recipe is included. Her website was inspected, not changed; no message, subscription, checkout or order was submitted.

Decision

Fix the existing customer journey before buying traffic. There are three verified broken links into ordering, and the site sends conflicting discount messages. Those repairs are likely to produce value sooner than a broad SEO campaign. The medium-term opportunity is unusually clear: a local, dietitian-led, mostly vegetarian meal-kit service for families in Lisbon and Cascais, with Portuguese pages opening a second audience. A small public recipe collection can introduce Natalie’s expertise while a paid archive remains useful to members.

SEO supports a business with positive contribution per order. It cannot repair a bag that loses money. Optimise for customers inside profitable delivery routes or choosing pickup; measure margin and retention alongside search visits.

Verified current public offer

The shop offers three recipes for four people: 12 portions per delivery, every other Monday. Vegetarian means three vegetarian dishes; mixed means two vegetarian plus one fish/seafood dish. The live product pages show:

These are the live public prices and 3% recurring discount observed on 2 October. Older search-tool product snapshots still show earlier prices; live HTML, product JSON-LD and rendered product controls support the figures above. They do not establish any grandfathered customer price or final checkout adjustment.

The FAQ describes recipes supplied as a PDF, a fixed menu, typical delivery between 15:30 and 18:30, and longer seasonal breaks. These are ingredients to cook, not finished meals. The website announcement observed in the browser names 12 October as the next delivery. Its current skipping instructions require cancellation and restarting rather than a simple skip action; product copy instead asks customers to contact the business. Clarify the real workflow.

The delivery map visibly covers the Lisbon–Oeiras–Cascais corridor with some inland coverage around Amadora, Cacém/Rio de Mouro, Odivelas and Loures. The polygon is illustrative rather than a postcode guarantee. It does not justify advertising delivery throughout Sintra municipality or Ericeira. Publish the actual served postcodes in text and confirm border addresses before producing new locality pages.

What was actually measured

• Direct HTTPS retrieval of all 19 sitemap URLs; extracted current titles, descriptions, H1/H2, canonicals, language tags, JSON-LD, links and HTTP status.
• Retrieved robots.txt and sitemap.xml. Tested observed legacy order links and the previously public menu URL.
• Inspected rendered delivery page, homepage, mobile navigation, shop and vegetarian product in a browser at 390 × 844. Restored the viewport afterwards.
• Looked up nonbrand local searches and public brand/review queries. Search-tool results are evidence of competing pages, not a measured Google ranking, local pack position or keyword volume.
• Attempted the public PageSpeed API; it returned HTTP 429 quota exhaustion. No Lighthouse score, LCP, INP, CLS, field performance pass/fail, traffic, backlink count, Google Business Profile review count or revenue attribution has been measured. No Search Console, GA4 or merchant account was accessed.

Raw crawl evidence is retained locally in work/natalie-seo-crawl.json and work/natalie-seo-extract.json, with downloaded public HTML and the API response. These technical work files should not be copied wholesale into a public report.

Prioritised findings

P0 — stop losing people who already want to order

The last point is both an SEO and conversion improvement: a genuine menu offers fresh, useful content and removes pre-order uncertainty. Keep one permanent menu URL, show the current delivery date, and retain dated archive previews only where they are useful. Do not present three dinners as two weeks of complete food.

The footer email visually separates the final letter onto adjacent text, although its mailto destination is correct. Repair its presentation. Source HTML also contains a newsletter storage-configuration warning, but it was absent from the rendered visitor accessibility tree; an administrator should verify the signup storage and a test opt-in. This is not evidence that a real customer signup failed. GA4 and a Facebook pixel appear in source; their presence does not establish correct consent handling, one pageview per visit or working purchase attribution.

P1 — improve commercial search relevance

The site has valuable foundations: public content is server-readable; key pages return 200; canonical links are present; HTTP and non-www visits reach the preferred HTTPS www host. Product structured data exists and current single-order prices agree with the page. This is a refinement project, not evidence that the whole site is invisible.

Example homepage title: Meal Kits in Lisbon & Cascais | Natalie’s Healthy Kitchen.

Example homepage description: Three dietitian-planned recipes for four people, with fresh ingredients delivered every other Monday in Lisbon and Cascais. Choose vegetarian or fish.

Example product H1: Vegetarian meal kit — 3 dinners for 4 people. Put its current price alongside it, with 12 portions and what pantry staples customers supply. These are proposed copy examples, not text already on her website.

P1 — serve Portuguese customers as deliberately as English-speaking residents

Every sampled HTML page declares en-US; no hreflang alternatives were found. English suits part of the existing audience. A well-written European Portuguese version of home, shop, delivery, how-it-works and FAQ can reach local households too. Translate the whole useful page, prices and ordering instructions; do not merely translate its navigation or add an automated language overlay.

Use /pt/ pages with correct pt-PT language, self-canonicals and reciprocal hreflang with English equivalents. Preserve existing English URLs rather than force a migration just for tidy paths. A visible language choice should work without a geolocation redirect. Google’s localized-page guidance supports reciprocal language alternatives.

Priority search intents to test in Search Console, not claimed volumes:

“Refeições prontas” is a different promise: customers usually expect precooked food. A clear meal-kits-versus-ready-meals guide can address the comparison, but the product page must not claim precooked delivery. Ericeira can be relevant to a digital shopping-list service even if not served by the delivery route. Do not create an Ericeira delivery page before real service exists. Avoid interchangeable pages for dozens of suburbs.

P2 — clean up index signals and structured data

The live sitemap contains 19 URLs, including four tag archives showing the same one-article collection, /home, and /404-error-page-not-found. The named error page returns 200 and has no observed meta noindex. Actual missing routes return 404, which is appropriate. Remove the named error-page URL from the sitemap and exclude it from indexing through supported Squarespace settings. Do not block crawling first if Google needs to see noindex.

/home returns 200 with canonical to the root, a useful existing consolidation signal. Prefer a redirect if supported without damaging the configured homepage; at least submit and link the canonical root. Tag archives have self-canonicals but little distinctive content: keep only useful collections, or use supported noindex settings for thin archives. This small site does not have a proven crawl-budget crisis.

JSON-LD already includes WebSite, Organization and LocalBusiness; product pages additionally have Product/Offer. Refine existing output rather than add a competing second block. The business address is a generic free-text Lisbon label and the Organization telephone has an extra zero versus the visible number. Correct the public business identity and contact fields, use appropriate country/locality and honest area served, and validate. Never publish a founder’s private home address just to satisfy a schema warning.

Add Recipe schema only to real public recipe pages, with the completed-dish image and visible matching recipe information. Use truthful preparation/cooking time, yield, ingredients and instructions; nutrition values only when calculated. No invented ratings. Recipe enhancements are eligibility, not a promise of rich results. Google Recipe documentation.

The homepage’s raw HTML includes 28 image elements, 14 without nonempty alt; several are decorative/tracking elements, so this count is not 14 accessibility violations. The delivery map and meaningful food imagery deserve useful descriptions. Decorative shapes should retain empty alt. Squarespace already serves responsive image candidates; inspect the actual images downloaded on mobile before claiming all need recompression.

P2 — mobile and performance: verified basics, unresolved speed

At 390 × 844 the homepage settled into a readable image-led hero, hamburger menu opened, shop navigation worked and the vegetarian product adapted to the viewport. Initial fade-in produced a transient dim state before content settled; that is not a measured loading failure. The product screenshot places a large photograph before price and purchase controls. Test a more concise mobile product header so a visitor sees portion count, current price and a purchase choice sooner. No order was placed.

Raw downloaded homepage HTML was approximately 403 KB; /home was approximately 803 KB in its sampled response; the shop approximately 425 KB. These are response-body sizes from this fetch, not total page weight or compressed transfer size. They warrant inspection, not a conclusion about slow user experience. First check hero asset sizes, animation delay, fonts and Instagram embed cost. Reduce unnecessary repeated blocks/widgets before considering a platform change.

Obtain a real mobile PageSpeed/Lighthouse test and Search Console Core Web Vitals. Targets are LCP ≤2.5 seconds, INP ≤200 ms, CLS ≤0.1 in suitable real-user data; insufficient traffic may leave field data unavailable. A lab score is diagnostic and does not prove field performance or a search ranking outcome. Google Core Web Vitals guidance.

Search competition and public reputation

The best direct SEO benchmark found is Mesa Fresca’s English Lisbon page and its Portuguese counterpart. They plainly explain the meal-kit category, geography, weekly timing and a pickup option. Natalie can compete on her specific dietitian expertise and mostly vegetarian family proposition, then match the practical clarity. This observation does not establish which brand ranks first.

Prepared-food substitutes compete for broader convenience searches: Origem describes takeaway serving Lisbon, Oeiras, Cascais and Sintra; Quase Quase explains regional frozen-food delivery and fees; Cozinha Por Mim offers ready food from Cascais. These are search competitors and alternatives to cooking, not equivalent meal-kit products. Natalie should explain that she removes planning and shopping while customers still cook.

Brand searches surfaced the website and its Squarespace subdomain. The subdomain sample correctly canonicals to the main domain, so this is not evidence of missing canonical configuration. No authoritative Natalie Google Business Profile link or independent review count was verified in this pass. This means unverified, not absent. The site has customer testimonials, but these are not an independently verified public review corpus.

If Natalie has an eligible business delivering directly to customers, confirm or claim one Google Business Profile, keep service areas truthful, add real photos, current ordering link, hours and consistent phone/name. If customers do not visit her business premises, use a service-area profile and hide the residential address. Google service-area guidance, address guidance.

Ask real customers for honest reviews through the verified profile link after successful deliveries; no incentives, fake reviews or selectively asking only happy customers. Named testimonials require permission. Do not add self-serving LocalBusiness review-star markup from her own testimonials; Google restricts that feature. Google review-snippet rules.

Public discovery and paid archive can coexist

Choose public recipes deliberately rather than publishing every historic PDF. Give enough to cook a complete sample and understand her taste. The paid proposition is an organised, regularly useful planning service; a generic recipe collection competes with abundant free recipes. A sample recipe should link to delivery when within coverage and to the digital tier for everyone else.

For a simple first release, keep the full archive authenticated and out of the public sitemap; do not send protected recipe text to the browser and merely hide it with CSS. A noindex tag is not access control. Public teaser pages should have meaningful original descriptions and a clear membership link. If Natalie later intentionally wants paywalled articles indexed, implement Google’s supported paywall markup and test that the reader and crawler experience is honest. Google paywall guidance. Ordinary public Recipe schema should describe the complete public sample, not promise a recipe whose instructions only exist behind login.

Lean 30 / 60 / 90-day programme

First 30 days — make existing demand work. Repair the three broken ordering links and 301 old routes; choose one current discount; clarify recurring price, next delivery and skip process; restore a public dated menu. Improve home/shop/product metadata, main headings and delivery text. Remove the named error page from indexable content/sitemap. Confirm existing analytics and newsletter setup; establish Search Console baseline and conversion events. Confirm GBP status privately with Natalie. Rough effort allowance: one 8–12-hour technical/content sprint, depending on Squarespace permissions and editing constraints.

Days 31–60 — widen the reachable audience. Publish properly reviewed Portuguese equivalents of the five purchase/help pages; add reciprocal hreflang. Improve and link existing Lisbon and HelloFresh comparison pages. Add one substantial Cascais/Oeiras page only after coverage confirmation. Publish four complete public sample recipes and one shopping-list example using already tested recipes. Validate product/recipe markup. Offer customers the digital/pickup/delivery choices the business can actually fulfil.

Days 61–90 — invest only in what is working. Add another four to eight carefully selected sample recipes if effort and demand justify it, repurposing material rather than requiring new recipe creation. Improve pages receiving qualified impressions but few clicks or orders. Test concise mobile price/portion presentation. Seek a few relevant local editorial/community links through genuine collaborations if Natalie approves. Compare customer acquisition and retention by tier and delivery area. Pause broad content production if the revenue it creates does not justify Natalie’s time.

A sensible maintenance allowance after the initial sprint is 1–2 hours per week plus periodic customer service; this is an effort estimate, not a vendor quote. Avoid a large agency retainer, mass backlink purchases or thousands of AI-generated recipes while the service economics are being repaired.

Measurement and realistic targets

No baseline traffic was supplied, so growth targets are provisional operating targets, not forecasts. Collect 28 days of baseline in Search Console and consent-aware analytics; compare rolling 28-day periods and allow for holidays, pricing changes and seasonality.

Track view_item, add_to_cart, begin_checkout, purchase, menu views, successful delivery-eligibility checks and actual newsletter opt-ins. Verify the purchase event fires once and reconcile with commerce orders. Do not submit test purchases without authorization. Tag GBP and partner links. Separate branded traffic (people already seeking Natalie) from nonbrand acquisition; separate recipe readers outside Portugal from route-eligible customers. A digital-tier conversion can be valuable outside the route, but a distant recipe visit alone is not delivery demand.

For a customer-acquisition ceiling, use expected retained contribution, not lifetime gross sales: cost to acquire one customer should be less than the contribution from their expected paid deliveries over the chosen payback window, after service costs. The economic simulator can display this alongside subscription count and churn.

Before doing paid SEO/advertising, Natalie would need to supply privately: Search Console/analytics read access or exports; exact served postcodes, route capacity, orders and retained-customer counts; permission to use recipes/photos; current pricing/discount/skip rules; verified Google profile URL. Financial documents and customer data need not become public to answer those questions.

Public source ledger

All retrieved 2 October 2026 unless the linked page states otherwise. Public live HTTP/HTML and rendered observations take precedence over stale search snapshots for current price.

1. https://www.natalieshealthykitchen.com/
2. https://www.natalieshealthykitchen.com/our-meal-kits
3. https://www.natalieshealthykitchen.com/our-meal-kits/p/vegetarian-bag
4. https://www.natalieshealthykitchen.com/our-meal-kits/p/mixed-bag
5. https://www.natalieshealthykitchen.com/faq
6. https://www.natalieshealthykitchen.com/delivery-area
7. https://www.natalieshealthykitchen.com/our-story
8. https://www.natalieshealthykitchen.com/how-meal-kits-work
9. https://www.natalieshealthykitchen.com/healthy-home-delivered-meal-kits-in-lisbon-fresh-ingredients-balanced-recipes
10. https://www.natalieshealthykitchen.com/hello-fresh-portugal
11. https://www.natalieshealthykitchen.com/news/how-natalies-healthy-kitchen-compares-to-hello-fresh
12. https://www.natalieshealthykitchen.com/robots.txt
13. https://www.natalieshealthykitchen.com/sitemap.xml
14. https://www.natalieshealthykitchen.com/404-error-page-not-found
15. https://www.mesafresca.pt/pages/healthy-meal-kits-lisbon
16. https://www.mesafresca.pt/pt/pages/healthy-meal-kits-lisbon
17. https://origem.org/refeicoes-para-entrega/
18. https://www.quasequase.pt/servicos
19. https://cozinhapormim.pt/
20. https://developers.google.com/search/docs/specialty/international/localized-versions
21. https://developers.google.com/search/docs/appearance/structured-data/recipe
22. https://developers.google.com/search/docs/appearance/structured-data/paywalled-content
23. https://developers.google.com/search/docs/appearance/core-web-vitals
24. https://developers.google.com/search/docs/appearance/structured-data/review-snippet
25. https://support.google.com/business/answer/9157481?hl=en
26. https://support.google.com/business/answer/2853879?hl=en

The old /menu, /our-bags and two /our-bags/p/... URLs were directly tested and returned 404. Guessed /delivery-map, /how-it-works and /next-weeks-menu probes were not treated as broken internal links. The evidence of actual broken links is the observed source page linking to the obsolete destination.