Privacy

Page updated on 17 October 2026. Short, because there is little to tell: no advertising, no tracking cookies, no data sold. We do measure aggregate site visits — see below.

Who is responsible

The data controller is 15953456 Canada Inc., 40 Ed Clark Gardens, Th09, Toronto, Ontario M6N 0B5, Canada. Privacy contact: privacy@demandlens.app — the privacy officer answers there.

Representative in the European Union

We have not designated a representative in the Union, and we rely on the exemption in Article 27(2)(a) GDPR: our processing is occasional, does not involve special categories of data on a large scale or data relating to criminal convictions, and is unlikely to result in a risk to the rights and freedoms of individuals.

This is a position, not a permanent state. We re-examine it as volume grows, and the day the exemption no longer holds we designate a representative and name them here. Saying nothing at all would leave you unable to tell an exemption from an oversight.

What we collect

• The idea text you submit — needed to run the scan. Treat it as you would any text sent to an online service; see the sub-processor list below.
• Your email address — collected at checkout, used to deliver the report and to let you retrieve it later.
• Payment data — handled by Stripe. We receive confirmation of payment and never see your card number.
• Scans and reports — the artefacts produced, so you can retrieve them.
• Connection metadata — IP address and user agent, in server logs and for rate limiting.
• Your email address, if you join a waitlist or ask for the newsletter — kept for that purpose alone, never merged with the reports you bought, and unsubscribing removes it. We send one message to confirm the address is yours: without that confirmation it is never added to the list, and nothing else is sent.

Aggregate benchmarks

If we ever publish figures about the ideas we have measured — for example, how often each opinion is returned among software ideas — those figures are computed on de-identified aggregates above a disclosed threshold: a cell too small to hide an individual is not published at all.

Your idea text never feeds any of it. An idea text cannot be anonymised, so it is excluded by rule rather than by care.

What we do not collect

No advertising pixels, no third-party trackers that follow you across other sites, no behavioural profiling, no cookie or device fingerprint used to identify you individually. Fonts are self-hosted, so your browser makes no request to a font provider when you visit.

Audience measurement

We use Cloudflare Web Analytics to count aggregate visits and measure page performance. It sets no cookie and stores no persistent identifier on your device — Cloudflare’s own description is that it counts page views, not people, and treats browser fingerprinting as more intrusive than cookies rather than using it as a substitute. No individually identifiable data about you is collected through this measurement.

When you arrive through a link we published — a launch post, a comparison page, a shared report — that link may carry a channel name. Your browser keeps it in session storage for the life of the tab, and we count in aggregate how many scans, purchases and sign-ups each channel led to. It names where a link was published, never who followed it; it is not stored per visitor and it is never shown on an invoice.

Cookies and local storage

We use one strictly necessary cookie: a session cookie set after you verify a one-time retrieval code, so your recovered reports stay accessible. It is required for that feature to function and carries no tracking.

Your browser also keeps one access key per report in local storage, so a report you paid for stays readable when you come back. It never travels in a URL. Nothing else is written to your device.

Because we set no advertising cookies and our audience-measurement tool sets none either, no consent banner is required.

Who else processes your data

We use the following sub-processors, each for a stated purpose. The rule we apply: a party is listed when it receives data derived from your idea text, even if it never sees the text itself.

The public sources we query — Hacker News, Stack Exchange, GitHub — receive only a search term, never your idea text, your email, or any identifier. They are not sub-processors of your personal data, and we name them here so the distinction is yours to check rather than ours to assert.

• Stripe — payment processing: email, payment details (we never see or store card data).
• DeepSeek — structuring of the submitted idea into a brief: the idea text, filtered before sending (see International transfers). DeepSeek processes this in China; according to its privacy policy, submitted data may be used to train its models.
• DeepSeek — pre-scan detection and classification of the submitted idea text: the idea text (for pre-scan detection, at most the first 600 characters; filtered before sending, see International transfers), keywords and excerpts from public threads. DeepSeek processes this in China; according to its privacy policy, submitted data may be used to train its models.
• DeepSeek — expansion of search queries from the submitted idea: the idea text filtered before sending (see International transfers) and keywords; known competitors are also sent for query expansion. DeepSeek processes this in China; according to its privacy policy, submitted data may be used to train its models.
• DeepSeek — brand name suggestions for the submitted idea: the idea text filtered before sending (see International transfers). DeepSeek processes this in China; according to its privacy policy, submitted data may be used to train its models.
• DeepSeek — pre-mortem — the reasons your idea could fail: the idea text filtered before sending (see International transfers), its sector, and excerpts from public threads. DeepSeek processes this in China; according to its privacy policy, submitted data may be used to train its models.
• TypeSafe — language model — reading of the idea text (language, intelligibility, whether it contains several ideas), choosing which sources to query, sorting and assessing the collected public threads, choosing and checking the citations drawn from them, and tying each pre-mortem reason to an excerpt that supports it; when TypeSafe does not answer, each of these steps falls back to the earlier rule-based treatment, with no fallback request sent to DeepSeek (DeepSeek’s own roles are listed on its line): the idea text you submit, excerpts from the public threads collected (Hacker News, Stack Exchange, GitHub) and the citations drawn from them, filtered before sending (same categories as for DeepSeek), sent during either a free or a paid scan; we do not intentionally include your email address, your order or session identifier or your payment data in the data sent. TypeSafe processes this data in the United States. Retention at TypeSafe: for as long as necessary taking into account the purpose of the processing, according to its data processing agreement. Standard contractual clauses: the data processing agreement offered by TypeSafe includes the European Commission’s standard contractual clauses (Implementing Decision (EU) 2021/914). TypeSafe’s own sub-processors: hosting and computing by infrastructure providers (list published by TypeSafe: trust.typesafe.ai/subprocessors). Training on submitted data: TypeSafe states in its privacy policy that it does not train or fine-tune models on this data.
• Supabase — database — entitlements and orders: email, order and entitlement records.
• Railway — API hosting: all data in transit and the scan/report file store.
• Vercel — front-end hosting: connection metadata (IP, user agent).
• Resend — email delivery — your report, and the codes that restore access to it: email, the report link and its order reference, and the single-use recovery code.
• DataForSEO — search-volume measurement: a keyword derived from the idea text you submit.
• Porkbun — domain availability and pricing: candidate domain names derived from the idea text.
• rdap.org — domain availability check — rdap.org, then the registry of the top-level domain concerned: candidate domain names derived from the idea, without the idea text.

Why we are allowed to process it

Performing our contract with you (delivering and retrieving the report), complying with legal obligations (tax and accounting records), and our legitimate interest in keeping the service available and unabused (rate limiting, logs).

How long we keep it

Orders and entitlements are kept as long as needed to let you retrieve your reports and to meet accounting obligations. Server logs are short-lived. You may ask us to delete your data at any time, subject to records we are legally required to keep.

International transfers

We are established in Canada. For TypeSafe (United States), the data processing agreement offered by TypeSafe includes the European Commission’s standard contractual clauses (Implementing Decision (EU) 2021/914).

Other sub-processors on the list also process data outside the EU/EEA, notably DeepSeek (China). For DeepSeek (China), there is no European Commission adequacy decision, and no standard contractual clauses or other specific safeguard have been concluded with it. Data processed in China may be accessible to the authorities of that country under its law. Before your idea text is sent to DeepSeek (China), a filter masks email addresses, phone numbers, postal addresses and links to personal profiles. We also ask you not to enter personal data in the idea field. The filter does not recognise names or other free-form personal details: if you enter them, they may be transmitted. The filter recognises postal addresses in French, English and Spanish formats, and French, Spanish in their usual written form, North American and international phone numbers. It does not recognise other formats, for example German or Italian addresses, Canadian postal codes or UK mobile numbers in national format, and it may miss other unusual formats. The same filter applies to what is sent to TypeSafe (United States): the text of your idea and the public-thread excerpts sent with it are filtered before sending.

Your rights

Depending on where you live, you may have the right to access, correct, delete, or port your data, to object to or restrict processing, and to withdraw consent.

Write to privacy@demandlens.app and we will answer within the time limit that applies to you.

Complaints

If you are in the EU/EEA you may complain to your national supervisory authority. In Canada you may complain to the Office of the Privacy Commissioner of Canada, and residents of Québec to the Commission d’accès à l’information.