ZinUp — Website Feature & Story Brief

For the website-design session. Self-contained: the new strategy story + every feature we're building (tagged live vs coming) + how to tell it. Plain language, no internal jargon. Written 2026-06-08. Supersedes the shorter content/website-direction-brief.md.

Status key: ✅ live now · 🔨 building now · 🗓️ planned (roadmap). Honesty rule: present 🗓️ items as "coming soon / roadmap," never as available today.

1. The one-line story

ZinUp turns a magazine into a beautiful, modern mobile experience — whether you start from a PDF, build it from scratch, or do both.

Elevator version (for the hero):

Most "digital magazine" tools just make your PDF flip on a screen. ZinUp is different: a real, swipeable, mobile-native magazine — convert your print PDFs, create new editions from scratch with no PDF at all, or weave the two together. Plus podcasts, translation, search-your-whole-archive, and an installable app — all without a developer.

Brand mission line (from the Oracle — carry it on the site):

"ZinUp is a perfect re-rendering of editorial intent — with a mobile visual rhythm." (Originally framed around print; the three-mode model below extends it to digital-first too.)

2. The strategy: Convert · Create · Weave

ZinUp is no longer "PDF-to-mobile" only. It's a publishing platform with three ways in, and the website must tell all three:

Mode	What it is	Who it's for
Convert	Upload a print/PDF magazine → automatic mobile-native reader	Publishers with print/PDF archives & current issues
Create	Build a digital magazine from scratch, no PDF	Digital-only / born-digital publishers
Weave	Run a PDF edition and add digital-only content alongside it — or switch the PDF off entirely	Hybrid publishers transitioning from print to digital

The three customer journeys to show on the site:

"I have print PDFs" → Convert them into a modern reader.
"I have no PDF, I just want to publish digitally" → Create from scratch.
"I have a PDF but want more" → Weave digital-only articles in (web exclusives, sponsored pieces, updates), or go fully digital over time.

Why this is the headline: no competitor in our market does all three, and none lets one publisher be print, digital, or both in a single property. That hybrid is ZinUp's unique position — see §4.

3. The full feature set (group these on the site)

A. Convert — from PDF to mobile

✅ Automatic PDF → mobile reader. Upload a magazine PDF; ZinUp extracts the articles and rebuilds them as a swipeable, TikTok-style card feed (not a page-flip flipbook).
✅ Discover tab — browse the issue by section, plus an ad gallery.
✅ "Original pages" view — readers can still see the true print layout if they want.
✅ Bulk archive upload — drop in a whole back-catalogue at once, each issue auto-numbered.
✅ In-app editor — fix/polish any article (headline, text, images) before publishing.

B. Create — build from scratch (no PDF)

✅ Author a digital issue from scratch — write articles (headline, standfirst, byline, body), add images, set a cover, save as draft, publish. (Tier-1 live in preview.)
🔨 Rich content blocks — pull-quotes, stat highlights, galleries, video & audio, buttons.
🔨 Advert cards — publishers place display ads in the feed.
🗓️ AI writing assist — draft, rewrite, suggest headlines, generate a starter issue. (Our standout feature.)
🗓️ Shoppable / product cards — tag products in the reading experience.
🗓️ Schedule / publish-later and starter templates.

C. Weave — print + digital together

🗓️ Add digital-only articles into an existing converted issue (web exclusives, sponsored, updates).
🗓️ "Switch the PDF off" — turn a converted title into a digital-only edition once content has moved across.

D. Read & engage (the reader experience)

✅ Mobile-native card reader — swipeable feed + Discover + saved, built for phones.
✅ Installable app (PWA) — readers add the publication to their home screen, with the publisher's own app icon. No app store needed.
✅ Podcasts — pull the publisher's existing podcast feed in; readers switch between read and listen.
✅ Read-aloud audio — any article can be listened to (text-to-speech).
✅ Translation — publish in multiple languages with a reader language switcher.
✅ Ask the catalogue (Q&A) — readers ask questions across the publisher's whole back-catalogue.

E. Grow & monetise

✅ SEO — automatic, Google-friendly meta for every article (native apps get none of this).
✅ Analytics — visit & engagement stats with a downloadable report.
✅ Custom domain — each publication runs on the publisher's own web address.
🔨 Advertising — display advert cards.
🗓️ Reader subscriptions / paywall — let publishers charge readers (Stripe).
🗓️ Shoppable commerce — product cards in the reading experience.

F. Manage (the publisher Studio)

✅ Self-serve Studio — create publications, add issues, manage everything, no developer.
✅ Draft vs publish controls — keep an issue private until it's ready; lock/hide; delete with confirmation.
✅ Per-publication settings — podcasts, translation, Q&A, audio, SEO, domain & app icon.

4. Why we win (the competitive angle)

Backed by a competitor study (research/competitor-analysis-digital-publishing-2026-06.md):

The market is old flipbook tech. The big names (FlippingBook, Yumpu, Issuu, Publuu) just make a PDF flip on screen — pinch-and-zoom, desktop-first. Even their own marketing admits static PDFs "don't adapt well to different screens." None let you build from scratch.
The few modern ones are enterprise (Pugpig, Nxtbook — built for Condé Nast, Hearst) or hide their pricing (Joomag). Not for independent publishers.
ZinUp's edge, in plain terms: a real mobile magazine (not a flipbook), the ability to build from scratch with AI, plus podcasts, translation, ask-the-archive, and an installable app — at a transparent price, with no developer.

Honesty guard for copy: don't claim rivals have "no AI" (some have AI helpers), and don't claim we're first/only on shoppable cards (a cheaper rival, Flipsnack, has them). Lead instead on the combination and on mobile-native + AI-from-scratch, where we are genuinely ahead for our audience.

5. Audience & tone

Who: independent and small-to-mid-size magazine publishers — print, digital-only, and hybrid. Not giant enterprise media groups, not consumers.
Tone: capable, modern, friendly, "you don't need a developer." Confident but not hypey. Show, don't oversell.

6. Pricing positioning

Transparent, self-serve, roughly £15–90/mo band. Cheaper and clearer than the capable enterprise tools; more capable than the cheap flipbook tools.
Current model: £89 base per title + add-ons, multi-title discount, archive setup (first issues free). The site should make pricing visible — a deliberate contrast to competitors who hide it behind "book a demo."

7. How to tell it on the website (structure guidance)

Hero: lead with the two/three paths — Convert your PDF · Create from scratch · Weave them together. Not PDF-only.
A section per mode (Convert / Create / Weave) with the matching journey.
A "modern vs old flipbook" comparison — our strongest story; show the card reader vs a pinch-zoom flipbook.
Feature groups as in §3: Convert · Create · Read & Engage · Grow & Monetise · Manage.
Visible pricing page.
Honesty: tag anything 🗓️ as "coming soon," keep claims defensible (§4 guard).
Carry the brand mission line (§1) somewhere prominent.

Questions or anything unclear → ask the ZinUp CEO agent. Canonical strategy lives in the Oracle: company/strategy.md (Convert · Create · Weave), company/vision.md, company/products/publisher/create-mode-roadmap.md.

Strategy — Zinup AI

Synthesised 2026-05-08 from Jason's 2026-05-05 audio feedback (round 1) and 2026-05-08 round-2 commercial pivot (see company/meeting-notes/). Companion to the live pitch in company/p3-pitch/App.tsx.

What this company is

Zinup AI is an AI-native publishing platform for indie, niche, B2B and small/independent magazine publishers — across the full print-to-digital spectrum. It is not only a PDF-conversion tool. It is a single product surface that meets a publisher wherever they are on the journey from print to digital, and lets them be print, digital, or both — in one place.

The product model: Convert · Create · Weave (canonical — owner direction, Jason 2026-06-08)

Convert — ingest legacy print archives and current-issue PDFs and turn them into a modern publisher property (the original wedge): mobile-native card reader, AI reflow, archive Q&A, audio (TTS + the publisher's own podcast), SEO with AI meta tags + auto Google Ads creative, OCR for non-digital archives, AI translation.
Create — author a digital magazine entirely from scratch, with no PDF at all. A magazine-native authoring studio (NOT a general-purpose website CMS) that produces the same modern reader, with AI assistance. Serves digital-only / born-digital publishers who never had — or no longer want — a print workflow.
Weave — combine the two in one property: keep a print/PDF edition and complement it with digital-only articles, drop born-digital content into a print archive, or transition off PDF entirely (switch the PDF off once content has moved across).

Why this is unique. The market splits into PDF-to-flipbook converters (no authoring) and a handful of from-scratch tools (no print heritage, mostly enterprise/quote-only) — evidence in research/competitor-analysis-digital-publishing-2026-06.md. No competitor in our segment does all three, and none weaves print and digital in a single publisher property. That hybrid — one publisher can be print, digital, or both — is ZinUp's distinctive model and the centre of the launch story. The whole stack stays AI-native and mobile-first: a swipeable card reader (not a flipbook), AI-assisted authoring, archive Q&A, podcasts, translation, SEO, installable PWA.

The brand "Zinup AI" sits with ZinUp. Anvilda is the build/ops partner — we operate the product, write the code, run the platform, support the customers. Anvilda is paid via revenue-share on sales we generate (no equity, no fees, no fixed retainer). This is the only co-* company in the Anvilda portfolio where the customer-facing brand and IP do not sit with Anvilda.

Who the customer is

The ICP is sharp:

Indie / niche B2B magazine publishers and small / independent consumer publishers with archives as small as two issues.
Digital-only / born-digital publishers (no PDF at all) who want a modern mobile magazine via Create mode — and hybrid publishers who run a print/PDF edition but want digital-only content alongside it.
Publishers who are too small to be on Pugpig's radar but big enough to want a real digital product.
The B2B / corporate / private-publication universe is a large slice of TAM that the consumer-magazine-only framing under-counts. ABM / SIIA / SIPA aggregate ~7,000 B2B+specialist titles before you count corporate and private publications globally.
Channel reach: a ~500-publisher channel pipeline. Trade events: PPA, FIPP, CRMA (City and Regional Magazine Association — citymag.org), InPublishing.

Out: large consumer publishers (already on Pugpig / native apps); pure newsstand aggregators (Magzter, Zinio).

What we sell

A platform subscription, per-title, with optional add-ons — the same subscription covers all three modes (Convert, Create, Weave); a title can be PDF-sourced, authored from scratch, or a blend, with no change to how it's billed. Each distinct title = its own platform instance. Publishers will not mix distinct titles on a single platform (advertising and subscriber database are title-specific) — exception: highly-similar titles (e.g. two fishing magazines) can share.

Pricing model — v4, APPROVED 2026-06-03 (see company/products/publisher/pricing.md):

Monthly per title: £89 base + £5 per 100 published pages/year, hard-capped at £189. Formula: min(189, 89 + 5 * round(issuesPerYear * avgPages / 100)). Typical title (~monthly, 100pp): ~£149/mo. Nothing over £200.
Multi-title: −15% on the base for titles 2+ (base becomes £75.65); volume component still applies per title.
Archive setup (one-time): free for first 12 issues, then £5/issue (×1.5 if scanned/OCR). Formula: max(0, archiveIssues - 12) * 5.00 * (scanned ? 1.5 : 1).
Add-on: training-data licensing pipeline — separate revenue line, rev-share with publisher.

The AI-native moves

Lifted directly from the live pitch (round 2). No incumbent in the competitive set is AI-native — that is the wedge. The nine moves:

PWA-first reader. Dominant pattern is native iOS/Android apps. PWA wins specifically in the indie/niche pocket — instant install, no app-store revenue share, frictionless trial. (FT and Forbes have shipped PWAs at scale; the mid-market default is still native, which is fine — that's not our segment.)
Bring PDFs to life with creative judgement. Movement, animations, swap a static print ad for a video ad on the digital edition — done with editorial creative judgement, not a mechanical text-strip.
Article extraction → vertical-mobile reflow. Today this is brutally manual. We automate it.
Contextualised semantic Q&A across the full archive. A few services do PDF-upload Q&A; nobody ships this for entire magazine archives in an easy-to-use reader interface.
Rights tracking, revenue sharing, archive monetisation. Critical to niche/smaller publishers specifically.
Podcast-style audio per issue — and integration of the publisher's existing podcast. Many publishers already have podcasts. The platform should let them plug those in so readers switch between read and listen — not just generate TTS from scratch.
SEO + Google indexability with AI meta tags + auto Google Ads creative + Analytics. Per-article AI-generated meta tags; Performance Max creative auto-generated from article content; native Google Analytics. Native apps get zero organic search traffic by default (App Indexing deprecated) — PWA + AI meta tags is the SEO play.
OCR for old archives without extractable text. Cloud-grade OCR (Google Document AI / AWS Textract — ~98% accuracy on archival scans) makes 20–30-year-old back-issues searchable end-to-end.
AI translation across HTML. DeepL HTML-preserving translation — no incumbent in our competitive set has shipped this.

Plus retrospective publisher tools so publishers can fix imperfect AI extractions and replace print-only assets (static ad → video ad) post-ingestion.

The integration philosophy: strict modern self-serve

Zinup integrates only with modern self-serve subscription / membership platforms — Shopify, WooCommerce, Chargebee, Memberful, MemberPress, Recurly, Seal Subscriptions. We do not integrate with legacy bureaux (Abacus, CDS Global, Myriad). This was an explicit round-2 reversal: legacy-bureau integrations were a time and revenue sink without strategic upside in our ICP. Headline: "Strict modern self-serve. No legacy bureaux."

The partnership shape

Brand and IP: ZinUp.
Customer contracts and domain (zinup.ai): ZinUp.
Build, ops, platform code, AI runtimes, customer support: Anvilda.
Revenue-share: Anvilda gets a percentage of sales Anvilda generates (definition of "Anvilda-generated sale" is one of the open commercial parameters). ZinUp's existing channel pipeline is ZinUp's; outbound we drive is shared.
No equity exchange. No retainer. No fixed fees.
An Anvilda ↔ ZinUp services agreement will be drafted at first-customer-close — deliberately deferred until there is real revenue to allocate.

Onboarding promise

8–10 hours per new publisher, AI-driven, no human bureau. Large-archive ingestion (500–600 editions) takes longer behind the scenes — naming-convention normalisation, HTML enrichment, TTS setup, OCR for non-digital scans — but framed as value (we do the heavy lift), not cost.

Sales motion

AI-native personalised demos built from the prospect publisher's own archive (Jason: "brilliant"). Highest-leverage move in the funnel.
Channel: the ~500-publisher channel pipeline.
Events: PPA, FIPP, CRMA, regional magazine bodies.
Outbound: AI-generated proof-of-product against named publisher archives.
Growth tier paid-acquisition: Google AdWords Performance Max + LinkedIn — shipped as part of the platform's growth automation, not just our own marketing.

Stage gates

M0 — first runner talks to Jason. Scaffold complete; runner up; CEO agent introduces itself; first work item agreed.
M1 — first paying publisher closed. First Anvilda-generated sale. Triggers the services-agreement draft.
M2 — repeatable funnel. Demos productionised; channel pipeline activated; pricing validated.
M3 — multi-title customer. Scale-tier proven against a real publisher running multiple distinct titles.

Out of scope — explicitly

Any direct-to-consumer reader play. Zinup is publisher-side.
Native iOS/Android apps as the primary reader (PWA first; native shell only when a customer needs it).
Legacy-bureau integrations (Abacus, CDS Global, Myriad).
A separate consumer-facing brand entity (ZinUp is the home of the brand).

Create Mode — Prioritised Build Roadmap

Owner-requested (Jason, 2026-06-08). Sequenced to get ZinUp's digital-first authoring demonstrably ahead of competitors before launch. Each stage carries the competitor reasoning from research/competitor-analysis-digital-publishing-2026-06.md (ticket #530). "Why" = what it buys us vs the field.

Guiding logic

Lead with what no affordable rival has: mobile-native card reading, AI from-scratch authoring, podcasts, pre-translated editions, archive Q&A, installable app. Several of these already exist — surface them, don't rebuild them.
Close the one table-stakes gap: reader subscriptions/paywall.
Reach parity where a cheap rival beats us: shoppable product cards (Flipsnack).
Don't overclaim: match shoppable cards, never market "world-first" there.

Already done

Tier-1 Create mode — author articles (headline/standfirst/byline/body), add images, set a cover, draft→publish. Built, verified end-to-end, on preview (design-preview), pending owner sign-off. (#506)
Differentiators already live (no build, just surface): card feed + Discover, RSS podcasts, pre-translated editions + language switcher, archive Q&A, read-aloud TTS, automatic SEO + analytics, per-publication installable PWA, custom domains.

The roadmap

Stage A — "Make it a real magazine" (authoring richness) · my lane (Studio)

Rich content blocks in the editor: pull-quotes, stat/number callouts, galleries, video & audio embeds, button/CTA, section dividers — plus advert cards so publishers can run display ads (the reader already renders ads).

Why: Joomag is the only from-scratch rival and it's enterprise/quote-only; the flipbook tools can't author at all. Rich, mobile-native blocks make our from-scratch output feel premium and beat both. Size: medium · no dependency.

Stage B — "AI-native authoring" (our single biggest leapfrog) · needs AI key

AI assist inside Create mode: draft/rewrite/expand article text, suggest headlines, generate a starter-issue outline, auto-write image alt-text + SEO.

Why: AI-native build-from-scratch is the one differentiator no affordable competitor has — this is the headline of the whole product. Size: medium · Dependency: ANTHROPIC_API_KEY (owner-action #396 — getting this key unlocks the marquee feature).

Stage C — "Get paid" (credibility + parity) · runner-buildable

Reader subscriptions / paywall — let publishers charge readers (#50, Stripe).
Shoppable / product cards — match Flipsnack; novel on a card feed.

Why: paywall is the only table-stakes feature we lack — needed to be taken seriously. Shoppable cards are the one spot a cheap rival (Flipsnack, ~$52/mo) beats us — close both. Size: paywall large (ticketed), shoppable medium · Dependency: Stripe keys (test keys in vault; live at launch).

Stage D — "PDF + digital together" (the transition audience) · my lane

Insert digital-only articles into an existing converted issue; "switch the PDF off" toggle for a digital-only edition of a converted title.

Why: serves publishers who keep a PDF but want digital alongside — or who want to migrate off PDF entirely. A journey the flipbook tools don't support. Size: medium · builds on Tier-1.

Stage E — Polish & onboarding speed

Schedule / publish-later; starter templates; "keep the channel alive between issues" filler post.

Why: scheduling is near-table-stakes; templates speed first-issue onboarding. Size: small–medium.

Recommended order & rationale

A (real-magazine blocks) — fastest path to a Create demo that looks ahead.
B (AI assist) — the headline leapfrog; start the moment the AI key lands.
C (paywall + shoppable) — credibility + parity; paywall can run on a runner in parallel with A/B.
D (PDF-alongside) — completes the three audiences.
E (polish) — last.

Dependencies to unblock now

AI key (#396): gates Stage B (our biggest differentiator). Owner-action.
Stripe keys: test keys stored; live keys at launch. Gates Stage C paywall.

Ownership (no-overlap)

CEO/my lane: Studio authoring UX — Stages A, D, parts of E (Create-mode editor).
Runners: Stage C paywall (#50, billing engine), and any backend/engine work — routed via the queue with full pipeline metadata, runners-first.

Competitor Analysis — Digital Magazine / Self-Publishing Platforms (2026-06)

Owner-requested (Jason, 2026-06-08). Two fact-checked deep-research passes (adversarial verification, primary sources preferred). Ticket #530. Pricing is 2026-current — re-verify exact tiers before publishing any public comparison.

Bottom line

The market is dominated by flipbook-era technology — PDF→page-turn converters that are desktop-first and don't reflow for phones. No competitor in the independent/SMB segment combines true mobile-native reading + AI-native from-scratch authoring + ZinUp's add-on stack (podcasts, translation, archive Q&A, installable app). That slot is empty. ZinUp can own "modern, mobile-first, transparently priced."

One honest exception: ZinUp is NOT unique on shoppable/product cards — Flipsnack offers them at ~$52/mo. Do not claim first/only there.

The three camps

Flipbook/PDF-conversion (the volume market) — FlippingBook, Yumpu, Issuu, Publuu, FlipHTML5, Paperturn, YUDU, 3D Issue (Flipbooks). PDF in → page-flip out. No from-scratch authoring; no true mobile reflow. Even their own marketing admits static PDFs "don't adapt well to different screens" (Publuu KB) and reflow "quickly shows its limitations… a lot of pinching and zooming" (FlippingBook).
Modern/mobile-native but ENTERPRISE — Pugpig (native apps; Condé Nast, Hearst, The Independent), Nxtbook PageRaft (responsive — but a separate product from their flipbook), Twixl (native apps from InDesign), MagLoft (native apps), 3D Issue Flow/Experios (single-column reflow from a PDF), Foleon (responsive content experiences, B2B). Capable, but quote-only/expensive and aimed at media groups — out of our segment.
From-scratch authoring — Joomag (Crater/Creative Studio editor + Gen-AI; the closest analog to Create mode — but pricing is quote-only/sales-led), Madmagz (template-based, SMB-priced but feature-thin, reader is a flipbook), Marq/Readymag (design/templating & web-design tools, not magazine platforms).

ZinUp differentiator scorecard (at the SMB price point)

Differentiator	Anyone else at SMB price?	ZinUp can claim
Swipeable card/story feed reading (not page-flip)	No — none of the field	✅ first/only (SMB)
AI-native from-scratch authoring (no PDF)	No — Joomag has it but enterprise/quote-only; Madmagz from-scratch but not AI	✅ first/only (SMB)
Podcasts via publisher RSS into the reader	No	✅ first/only
Pre-translated parallel editions + language switcher	No (rivals at most auto-translate one edition)	✅ first/only
Archive Q&A ("ask the catalogue")	No (medium confidence)	✅ soft-claim
Per-publication installable PWA + custom domain	Largely no (rivals build native apps at enterprise cost)	✅ strong
Shoppable / product cards	YES — Flipsnack (~$52/mo Professional)	❌ do NOT claim first/only

Pricing landscape

Entry, cheap-and-basic (~$7–30/mo): Publuu $7, Yumpu adFREE €14.95, FlippingBook Lite $20, Madmagz €9.99–49.99. Flipbook tools, thin features.
Mid/pro (~$50–280/mo): Flipsnack Professional ~$52 (shoppable), FlippingBook Optimal/Advanced $93–176, Issuu ~$23–270, Yumpu PROKiosk €277, 3D Issue Flipbooks $100/$225/$799.
Enterprise / quote-only (opaque): Joomag, Pugpig, YUDU, Twixl, MagLoft, 3D Issue Experios, Foleon. (PageSuite excluded — enterprise, out of segment.)

Recommendation: transparent self-serve pricing in the entry-to-mid band (~$15–90/mo). This undercuts the opaque enterprise tools (you must "book a demo" to even learn YUDU/Joomag's price) while out-featuring the cheap flipbook floor. ZinUp's current model (£89 base + add-on upcharges) sits right in this sweet spot and is defensible because we out-feature the cheap tier and are transparent unlike the capable tier.

Roadmap implications

Table-stakes (must-have to be credible): PDF import/convert; basic analytics + CSV export; SEO meta; custom domain; video/image/link interactivity; paywall/subscription access control; an installable/app output. (ZinUp has all but paywall/subscription — that's the gap to close for credibility; tracked with Stripe billing #50.)

Differentiators (where we win — lead the marketing with these): mobile-native card feed + Discover; AI-native Create mode; RSS podcasts; pre-translated editions; archive Q&A; read-aloud TTS; per-publication PWA. Shoppable cards = parity, not lead (match Flipsnack), but novel combined with the card feed.

Honesty / caveats (do not overclaim)

Do not say rivals have "no AI" — several have AI design/translation/conversion helpers. Our edge is the specific combination, especially AI from-scratch authoring + card-feed reading.
Do not claim first/only on shoppable cards (Flipsnack has them at SMB price).
Archive-Q&A uniqueness is medium confidence (not exhaustively probed on every vendor) — phrase as "we haven't found another that does this," not "the only one in the world."
Unresolved (low priority): NXT Publisher/4YouGo disambiguation vs Nxtbook PageRaft; Paperturn specifics; Foleon SMB-vs-enterprise placement. None likely to change the verdict, but flagged for honesty.

Primary sources (selection)

FlippingBook (order-online, digital-publishing-features), Publuu (KB, prices), Yumpu, Joomag (crater-editor, pricing, genai), Nxtbook (digital-platforms, responsive, PageRaft reflow guide), Pugpig (home, Condé Nast case study), Twixl (platform, pricing), 3D Issue (experios, experios-pricing, Flow KB), Madmagz (pricing, FAQ), MagLoft (mobile-app, universal-app), YUDU (online-magazine-software), Flipsnack (product-tag feature + help + shoppable-magazine blog), Marq, Readymag, Foleon.

Product Brief — Zinup AI Publisher Platform (MVP)

Status: first GM draft — 2026-05-12. Jason signs once co-zinup exists (post-#1100).

What we're building

An AI-native publisher platform for indie, niche, B2B, and small independent magazine publishers, spanning the full print-to-digital spectrum (Convert · Create · Weave — see company/strategy.md). One product surface that lets a publisher:

Convert legacy archives + current-issue PDFs into a modern publisher property (PWA reader, AI mobile reflow, archive Q&A, audio, SEO with AI meta + auto Google Ads, OCR, AI translation);
Create a digital magazine from scratch with no PDF (magazine-native AI authoring → the same reader);
Weave the two — complement a PDF with digital-only articles, or transition off PDF entirely.

The promise: From your PDFs — or from scratch — to a live mobile magazine in hours. No legacy bureaux. From £89/mo.

MVP feature list (M1 — first paying publisher)

Must have

PDF ingestion pipeline — drag-and-drop or SFTP upload of issues; automated extraction → article reflow.
PWA reader — installable, offline-capable, per-title subdomain or vanity domain.
Article extraction → vertical-mobile reflow — automated, with publisher-side edit override.
Contextualised semantic Q&A across the archive — built on pgvector; conversational UI in the reader.
Audio per issue — TTS-generated, plus a slot for the publisher's existing podcast feed.
SEO + AI meta tags — per-article AI-generated meta, native GA4 wiring.
OCR for archives without extractable text (~98% accuracy on archival scans).
AI translation across HTML (DeepL HTML-preserving).
Modern self-serve subscription integrations — Shopify, WooCommerce, Chargebee, Memberful, MemberPress, Recurly, Seal Subscriptions. No legacy bureaux.
Stripe billing for the publisher — one platform per title.
Multi-tenant RLS — publishers cannot see each other's data; readers see only their own title.
AI-native personalised demo generator — for sales (upload 3–6 PDFs → 6-hour demo URL).

Nice to have (M1 stretch)

Google PMax auto-creative (Growth-tier feature). Generates per-article PMax creative; publisher activates with their own Google Ads account.
LinkedIn ads integration (Growth tier).
Retrospective publisher tools — fix imperfect AI extractions; replace print-only assets (static ad → video ad) post-ingestion.
Rights tracking — per-article rights ledger; revenue-share accounting per rights-holder.

Explicit non-goals for M1

No native iOS/Android app. PWA only.
No legacy-bureau integrations. Hard line.
No newsstand-aggregator features. We are not Magzter; we are not Zinio. We build platforms; we don't aggregate.
No reader account system — readers authenticate via the publisher's existing membership platform.
No publisher-facing AI training pipeline. Training-data licensing is a separate revenue add-on, not an M1 feature.

Pricing model — v4 (APPROVED 2026-06-03)

Authoritative detail in company/products/publisher/pricing.md.

Monthly per title: £89 base + £5 per 100 published pages/year, capped at £189. Most publishers pay ~£149/mo. Nothing over £200.
Multi-title: −15% on base for titles 2+; volume component applies per title.
Archive setup (one-time): free for first 12 issues, then £5/issue (×1.5 ≈ £7.50 if scanned/OCR).
Add-on — training-data licensing rev-share.

Segments (M1 targeting)

Indie / niche B2B magazine publishers — small, ABM/SIIA/SIPA-shaped publishers with 2–500 issues in archive.
Small / independent consumer publishers — regional + special-interest titles.
Out of scope M1: large consumer publishers (already on Pugpig), pure newsstand aggregators, vanity self-publishers.

Customer-journey acceptance scenarios

I'm an editor at a 12-year-old B2B magazine with 144 back-issues as PDFs. → upload via demo generator → 6-hour demo against my own archive → demo email arrives → book a call → ~£149/mo (monthly 100pp title) + £410 archive setup (82 billable issues × £5).
I'm a small independent consumer publisher with 8 issues per year going forward and no archive. → ~£99/mo (quarterly-ish volume) → upload current issue → live PWA within 24 hours.
I'm a multi-title publisher running two fishing magazines. → both titles on the platform → can share the same advertising database (titles are "highly similar"), per Jason's exception rule. Second title at −15% base.

Go/no-go gates

M1 unlock (first paying publisher)

Brand assets (wordmark) finalised at minimum-viable level (Jason approval).
zinup.ai live with hero + pricing + demo CTA.
Demo generator producing working demos within 6 hours, end-to-end.
At least 2 ingestion integrations live (Shopify + Memberful at minimum).
OCR pipeline producing ≥95% accuracy on test archives.
Stripe billing live.
Acceptance agent's Playwright suite green on production.

M2 unlock (repeatable funnel)

≥5 paying publishers across Starter + Growth.
AI-native demo conversion rate measured (target: ≥20% demo→close).
The 500-publisher channel pipeline activated.

M3 unlock (Scale tier validated)

≥1 paying Scale-tier publisher running multiple distinct titles.
Training-data licensing add-on negotiated with at least one publisher.

Onboarding promise

8–10 hours per new publisher, AI-driven, no human bureau. Large-archive ingestion (500–600 editions) takes longer behind the scenes; framed as value (we do the heavy lift), not cost.

Open decisions for Jason

Demo generator quota — how many demos/mo before we cap? Quality-vs-cost tradeoff.
Free pilot trials — needed if AI-native demos already let prospects see the product against their own archive?
Multi-title pricing edge cases — two near-identical fishing magazines share one platform per Jason's exception rule; what's the formal boundary?
Reader-side authentication fallback — what if a publisher uses a legacy membership system not on our integration list? Walk away (current draft) or build a JWT-import bridge?
Per-article rights ledger granularity — at M1, do we model rights at the issue level or article level? Article = better; takes longer.

Vision — Zinup AI

Draft authored by Gerben on behalf of Jason, 2026-05-17. Subject to Jason sign-off via owner-action ticket; Anvilda agents may treat as authoritative pending that sign-off.

Vision statement

In a world where every magazine's archive is searchable, listenable, and rendered as a modern reading experience — and where the publisher, not an aggregator, owns the audience — independent publishers stop being a long tail and become the centre of trade-specific media.

Mission statement

Zinup is a perfect re-rendering of editorial intent of a printed magazine — with a mobile visual rhythm.

Coined by Jason on 2026-05-25, immediately after rejecting the placeholder hero "Magazines, lifted." with the verdict "doesn't mean anything to anyone." The mission is brand-binding: every customer-facing surface must carry it (verbatim where space allows, or with the editorial intent / mobile rhythm concept landing if the literal sentence is reshaped for headline cadence). It supersedes the May-20 "positioning principle" as the prime brand line; the principle remains as the long-form argument behind the mission. See .claude/agent-memory/co-zinup-ceo/mission_statement.md for the operating rules and rendering rules per surface.

Why this matters

Tens of thousands of indie, niche, B2B, and corporate publishers sit on years (sometimes decades) of high-quality content trapped inside PDFs. Pugpig and similar incumbents won't serve them — too small. Issuu/MagLoft/3D Issue ship flipbooks wrapped in mobile shells. Magzter sells consumer subs but captures the SEO traffic for itself. The result: real publications with real audiences but no modern digital surface. Reader expectations have moved on; the platform tools haven't. Zinup AI closes that gap — and not only for publishers with a PDF: a growing band of born-digital publishers want a real mobile magazine with no print step at all, and many print publishers want to weave digital-only content alongside their PDF. ZinUp serves all of them in one property (Convert · Create · Weave — see strategy.md).

3-year picture (end of 2028)

Dimension	Target
Revenue (platform subscriptions)	£3M ARR
Paying publishers	~300
Geography	UK lead, US #2, AU/NZ/IE/CA established, Spanish-language pilot live
Product scope	Full reader + ingestion + archive Q&A + audio + SEO + AI-creative pipeline + DeepL translation + training-data licensing add-on
Team	Zero human employees on the Anvilda side. AI agents handle 100% of build/ops. ZinUp closes top of funnel.
Channel	A 500-publisher channel pipeline activated; trade-event presence at FIPP, PPA, SIPA, CRMA

What success looks like

A publisher with a 25-year archive of PDFs can self-onboard end-to-end in a working day, with no human from Anvilda in the loop.
"Indie publisher" stops meaning "second-class digital experience" — Zinup AI publishers ship a reader, archive Q&A, and audio that match or beat the FT's product on any device.
The existing channel doesn't have to sell Zinup AI — they sell what the publisher gets, because the demo (built from the publisher's own archive in <6 hours) does the convincing on its own.
A second revenue line (training-data licensing) earns publishers from archives they never thought were monetisable — turning the long tail of content into the second balance sheet for the next decade of small-publisher economics.

What we are NOT building

A consumer newsstand. We are not Magzter, Zinio, Apple News+, or Readly. The publisher's audience is the publisher's audience — we never aggregate it, never disintermediate it, never sell to consumers directly.
A native iOS/Android app. PWA only. The economics of the long tail don't support per-publisher native development; PWA + a single platform stack is the only model that scales.
A bureau service. No production-staff overhead. Every operation is AI-driven self-serve. If a publisher needs a human bureau, they're outside our ICP.
A flipbook tool. "PDF reformat" means article extraction + creative reflow — not a digital page-turner.
A general-purpose website CMS. We do offer from-scratch authoring (Create mode — see the Convert · Create · Weave model in strategy.md), but it is magazine-native: it produces our mobile card reader, not a blog or marketing site. Publishers who want a general website builder belong on WordPress, Ghost, or similar. (This revises the earlier "no from-scratch CMS" non-goal — from-scratch magazine authoring is now in scope; a general web CMS is not. Owner direction, Jason 2026-06-08.)

What this changes about how we work

Every decision is gated on "does this serve the long-tail publisher buyer?" — if it serves only large publishers or only consumers, it's not us.
No equity partnership with publishers. Platform subscription + (optional) training-data revenue-share. Clean, scalable, no captable mess.
Brand sits with ZinUp, not Anvilda. This is the only co-* company where customer-facing brand and IP do not sit inside Anvilda; Anvilda is the build/ops partner on rev-share-on-sales. Future strategic decisions must respect that.

How this maps to `goals.md`

Gate 1 — Conversion quality proves the wedge claim (AI reformat beats TouchTree's PDF render). ✅ PASS 2026-06-04.
Gate 2 — Onboarding velocity — self-serve platform + the three commercial features (podcasts/translation/Q&A). ✅ PASS 2026-06-05.
Gates restructured 2026-06-05 (owner). Gates 3 & 4 are the pre-launch build gates; the funnel/scale gates renumbered to 5 & 6.
- Gate 3 — Launch readiness: core 🔓 ACTIVE — Archive upload (bulk back-catalogue import), Issue Review HTML editor, Settings·Billing (Stripe + add-on upcharge).
- Gate 4 — Launch readiness: content → GO LIVE — Audio read-aloud (TTS, for the ~95% of publishers with no podcast) + SEO, then public launch.
- Gate 5 — Repeatable funnel (post-launch) — proves the channel + paid-acquisition scales beyond the warm channel pipeline.
- Gate 6 — Scale validation (post-launch).
Post-launch is where the £3M ARR vision lives. Don't build for it before then.

ZinUp — Website Feature & Story Brief

1. The one-line story

2. The strategy: Convert · Create · Weave

3. The full feature set (group these on the site)

A. Convert — from PDF to mobile

B. Create — build from scratch (no PDF)

C. Weave — print + digital together

D. Read & engage (the reader experience)

E. Grow & monetise

F. Manage (the publisher Studio)

4. Why we win (the competitive angle)

5. Audience & tone

6. Pricing positioning

7. How to tell it on the website (structure guidance)

Strategy — Zinup AI

What this company is

The product model: Convert · Create · Weave (canonical — owner direction, Jason 2026-06-08)

Who the customer is

What we sell

The AI-native moves

The integration philosophy: strict modern self-serve

The partnership shape

Onboarding promise

Sales motion

Stage gates

Out of scope — explicitly

Create Mode — Prioritised Build Roadmap

Guiding logic

Already done

The roadmap

Stage A — "Make it a real magazine" (authoring richness) · my lane (Studio)

Stage B — "AI-native authoring" (our single biggest leapfrog) · needs AI key

Stage C — "Get paid" (credibility + parity) · runner-buildable

Stage D — "PDF + digital together" (the transition audience) · my lane

Stage E — Polish & onboarding speed

Recommended order & rationale

Dependencies to unblock now

Ownership (no-overlap)

Competitor Analysis — Digital Magazine / Self-Publishing Platforms (2026-06)

Bottom line

The three camps

ZinUp differentiator scorecard (at the SMB price point)

Pricing landscape

Roadmap implications

Honesty / caveats (do not overclaim)

Primary sources (selection)

Product Brief — Zinup AI Publisher Platform (MVP)

What we're building

MVP feature list (M1 — first paying publisher)

Must have

Nice to have (M1 stretch)

Explicit non-goals for M1

Pricing model — v4 (APPROVED 2026-06-03)

Segments (M1 targeting)

Customer-journey acceptance scenarios

Go/no-go gates

M1 unlock (first paying publisher)

M2 unlock (repeatable funnel)

M3 unlock (Scale tier validated)

Onboarding promise

Open decisions for Jason

Vision — Zinup AI

Vision statement

Mission statement

Why this matters

3-year picture (end of 2028)

What success looks like

What we are NOT building

What this changes about how we work

How this maps to goals.md

How this maps to `goals.md`