oll.am · oll.in Positioning · research-grounded (3 streams · verified pricing) · 2026-07-08

oll.in — the no-brainer sell

v1 is built. Before it deploys, this is the answer to one question: why does a stranger pay CHF 9 on day one instead of pasting their CV into free ChatGPT? Three research streams — competitors + positioning, conversion UI/UX, and the user journey — converge on a single, uncomfortable, useful conclusion: the wedge is not "we bundle everything." Someone in Zürich already bundles. The wedge is a finished, trustworthy, genuinely-Swiss dossier you can send — grounded in your real CV, cited, and honest about what it doesn't know. This doc turns that into landing copy, a conversion journey, a pricing-unit decision, and the five calls that make it a no-brainer.
Reframe: the bundle is taken — grounding + Swiss-correctness is the moat Lead with: the felt outcome · prove with: a citation + an honest gap on screen Fix: the pricing unit, not the price

What's inside

  1. The finding that changes the plan
  2. The field — 13 competitors, verified prices
  3. What we actually own (ranked)
  4. Why buy over free ChatGPT
  5. The no-brainer journey — 6 moments
  6. The conversion moves
  7. The pricing-unit decision
  8. The honest risk & the demo that beats it
  9. Five calls for Sam

The finding that changes the plan the bundle is not greenfield

The plan-of-record assumed the wedge was "all-in-one Swiss Bewerbungsdossier" — headshot + CV + cover letter in one flow. Research kills that as the lead: Profile Bakery (Zürich, Mamendo GmbH) already bundles headshot + CV template + cover letter + LinkedIn analysis + ATS checker + a job-search Kanban into €24.90 / €29.90 / €49.90 one-time packs. A Swiss company, at our price point, occupying "we bundle."

So "we bundle" is table stakes, not a wedge. But read what Profile Bakery — and everyone else — doesn't do:

What nobody in the field does
  • Grounds output in your real CV with line-by-line citations.
  • Abstains honestly — flags a claim it can't support instead of inventing one.
  • Gets Swiss conventions right — tabellarischer Lebenslauf, photo, nationality/permit (B/C/G/L), Arbeitszeugnisse, Motivationsschreiben, DE/FR/IT.
Why that's exactly oll.in
  • The oll-memory RAG spine already retrieves + cites + abstains — it's built and proven.
  • The dossier composes from your corpus, not a blank prompt — grounding is native, not bolted on.
  • The moat is the one thing a chatbot and a template bundle both structurally cannot reproduce.
The pivot in one line: stop selling "everything in one place" (taken) and start selling "a finished Swiss dossier built from your real CV — cited, never invented." Same product, sharper knife.

The field verified July 2026 · pay-once vs subscription

Thirteen competitors, real prices. Two facts jump out: subscription dominates (a ~CHF 24–30/mo anchor), and among resume tools only Rezi and CoverDoc are genuinely pay-once. Nobody grounds, nobody abstains, nobody does Swiss format.

CompetitorWhat it isVerified priceThe gap oll.in exploits
ChatGPT / Claude (DIY)paste CV+ad → draftFree · $8 · $20/mofabricates experience · no Swiss PDF · no headshot · needs prompt-craft
Profile Bakery (ZH)headshot + bundled CV/letter/ATS€24.90 / €29.90 / €49.90 oncespeaks EUR + "EU infra"; CV/letter read as generic template bolt-ons, not grounded or Swiss-tailored
LinkedIn PremiumAI profile/letter + insights$39.99 · €29.98/moEnglish-first · assistive side-layer · no headshot · recurring
Swiss Bewerbungsservicehuman-written dossierCHF 120–250 once2–3 day turnaround — untenable for 20 applications
Swiss career coachinghuman coach / CV reviewCHF 170–180/hr20–50× pricier · must book · free RAV/BIZ path exists
ReziATS-first resume builder$29/mo · $149 lifetimeUS-ATS monoculture · irrelevant to Swiss photo/permit/Zeugnis norms
Kickresume / Enhancv / Zetytemplate + AI writer~$13–26/mo (dark-pattern renewals)no pay-once · ungrounded bullets · no Swiss format · fake-free trials
DACH Abofallen (cvmaker.de)cheap teaser CV builders€1.99 trial → ~€15/moconsumer-watchdog warnings — "pay once, no trap" is a direct answer
Headshot pure-plays (HeadshotPro/Aragon)AI headshots only$29–79 onceno job tooling — don't fight them on standalone photo price
Read: our headshot alone loses on price/volume to the pure-plays — so it only ever sells inside the dossier flow. And "pay once" is real leverage against 7 of 9 resume tools + the distrusted Abofallen, but Rezi/CoverDoc/Profile Bakery partly match it — so it's a reinforcer, not the headline.

What we actually own ranked by no-brainer pull

Four candidate wedges, ordered by how fast a cold buyer feels the value — which is not the same as how defensible they are. The trick is to lead with the felt one and prove with the deep one.

1 · Swiss-correct finished dossier, done-for-you the lead
A sendable A4 tabellarischer Lebenslauf + Motivationsschreiben in DE/FR, with photo, permit line, Zeugnis handling — no prompt-craft. ChatGPT structurally can't emit it; the DACH builders are trap-subs; the human services take days. This is what earns the CHF 9. It's the value a buyer feels in ten seconds.

2 · Grounded + cited from YOUR CV, honest abstention the trust layer
The deepest, most defensible moat — nobody has it, and it rides the oll-memory RAG stack we already built. But it's the hardest to sell cold: a buyer doesn't feel hallucination risk until they've been burned. So make it the visible proof under the headline, not the headline itself — show a citation resolving and one honest gap on screen.

3 · Pay-once, no subscription, no auto-renew reinforcer
Strong against the subscriptions + the distrusted Abofallen — but partly matched by Rezi/CoverDoc/Profile Bakery. Keep it loud on the page; don't make it the whole pitch.

4 · All-in-one bundle demote
Profile Bakery already occupies it. Don't claim "we bundle" — win within the bundle on Swiss-correctness + grounding.

Lead with the outcome a buyer feels in ten seconds. Prove it with the moat nobody else has. Reinforce with pay-once. Never lead with "we bundle."

Why buy over free ChatGPT the honest answer

This is the question every skeptic asks, so answer it head-on. The honest verdict: oll.in is a real product — only if it's pitched as "a finished, trustworthy, Swiss-correct dossier you can send," never as "better cover-letter text." We will not beat GPT-5 on prose, and we shouldn't try. We win on three things a chatbot structurally cannot do:

1 · Emit a formatted Swiss PDF dossier
  • A4, tabellarisch, photo, permit line — a sendable artifact, not a chat transcript to reformat.
2 · Make the headshot
  • A real Bewerbungsfoto in the same flow — ChatGPT can't.
3 · Ground & abstain vs a source CV
  • Cites your real experience; refuses to invent. The chatbot cheerfully fabricates.
CHF 9 is impulse-priced against the pain of one important application. The pitch is never "cheaper AI writing" — it's "the friction and the fabrication risk, removed, for the one job you actually care about."

The no-brainer journey 6 moments · where the franc enters

Conversion research + the user-journey study agree on the shape: value before login, paywall at peak wow, price visible from screen one. Each moment below carries a named lever from the research.

① Paste the job — price already visible
Home is one field: paste a posting. The CHF 9 pay-once price sits on screen one (price-anchor-from-the-start beats a surprise at checkout). No signup, no wall.

② Upload your CV once — parse, don't interrogate
One upload, parsed into the corpus. Never a 12-field form. The buyer cares about the outcome; every field we ask for is friction we chose to add.

③ Watch it compose — the labor illusion + the IKEA effect
The dossier forms live, token-by-token; citations resolve as you read; the headshot develops in the corner. Visible work builds trust (the labor illusion); watching their CV become the dossier builds ownership (the IKEA effect) — both before a single franc.

④ See one honest gap — the trust proof
Somewhere on screen, a claim the engine couldn't ground is flagged, not faked. This is the moment that separates oll.in from every ungrounded competitor — and the single most persuasive thing a skeptic can see.

⑤ Take it — paywall at peak wow, guest checkout
The download/email button is the paywall, placed at the peak of the wow (finished dossier in hand). Guest checkout — no account required — with TWINT + card. Sign-in is optional, only to save.

⑥ Come back — the RAV loop
Swiss job-seekers on RAV owe 8–12 applications/month. After the first dossier, offer to save the profile so the next job is one paste away. This is the returning-user engine, not a one-shot sale.

Value → trust → ownership → pay — in that order, all before an account. The franc enters at "take it," at the exact peak of felt value.

The conversion moves what to change in v1 + the landing

Concrete, research-backed levers. Most are copy + placement, not new engineering — which is why they're cheap and belong in the first deploy.

MoveWhy (research)Where
Price visible on screen onesurprise pricing kills trust; anchoring early convertslanding hero + compose view
Show a citation resolving + one honest gapthe moat only converts if it's felt, not claimedthe live compose (⓸ above)
Guest checkout (no account)account walls are the top drop-off; pay first, save laterCore guest-checkout endpoints (PR exists)
TWINT + cardTWINT is the default Swiss impulse-payment railcheckout
"Pay once · no subscription · no auto-renew," louddirect answer to the distrusted Abofallen + subshero + checkout
Headshot as +CHF 20 add-on, never standalonepure-plays win on standalone price/volume; inside the flow it's pure margincompose + checkout
Money-back guaranteede-risks the first franc from a stranger who can't touch it yetcheckout microcopy
Live "dossiers built" countersocial proof; labor illusion at the brand levellanding (only once real)
Honesty guardrails hold: the counter goes live only when the number is real. No "Swiss-hosted"/data-residency claim — "Swiss format" (the Bewerbung layout) is the true claim. No auto-apply, no scraping.

The pricing-unit decision fix the unit, not the number

CHF 9 is the right zone — it undercuts the CHF 120–250 human services ~15×, dodges the $24–30/mo subs, and is impulse-priced against free ChatGPT's friction. The critical fix is the unit. If CHF 9 buys unlimited dossiers, a 20-application job-seeker mentally defaults to free ChatGPT (marginal cost ~0). If CHF 9 buys one finished target application, the pitch becomes "the price of getting this important one right" — and the RAV loop brings them back per job.

Recommended
  • CHF 9 = one finished target application (one job → one tailored, grounded, Swiss dossier).
  • + CHF 20 to add the headshot to that dossier.
  • Optional small pack — 3 for CHF 19 — for the RAV-quota applier.
  • A CHF 29 decoy tier (dossier + headshot + a small extra) makes CHF 9 feel obvious.
Make it unmistakable
  • State the unit in plain words: "first application CHF 9." CHF 9 is the trial rung, not the everyday unit.
  • Cap it against the Pass, visibly: "the Pass pays for itself by application #11." A RAV seeker owes 10–12/mo, so per-app pricing silently out-costs a CHF 99 Pass — the page must do that arithmetic for the buyer or the ladder confuses the everyday professional it's built for.
  • Headshot is never a standalone SKU — only an add-on (the pure-plays beat us standalone).
  • Pay-once framing stays loud; no recurring anything on the CHF 9 product.
Why the cap matters: at CHF 9 × 11 applications = CHF 99, per-application pricing makes the Active Search Pass look like break-even and Radar (CHF 19/mo) look strangely cheap next to per-dossier pricing. Fix the unit, not the number: CHF 9 wins the trial, the Pass/Radar wins the volume applier. Full ladder in the v2 money model.

The honest risk & the demo that beats it where the thesis is thin

The uncomfortable truth: the strongest differentiator (grounding / honest abstention) is the hardest to make a cold buyer feel — hallucination pain is invisible until after you've been burned. And the easiest-felt differentiator (Swiss PDF + headshot in one flow) is partly occupied by Profile Bakery at the same price. So the skeptic bounces to: "I'll paste my CV into free ChatGPT and drop it into a jobs.ch Word template." For a format-literate applicant, that's genuinely good enough — oll.in wins the buyer who doesn't know Swiss conventions or won't run the prompt/format loop. That's narrower than "everyone job-hunting."

The mitigation is a demo, not an argument. Lead with the finished-dossier outcome + headshot in one flow (the felt value). Make grounding the visible trust proof — show the inline citations and an actual abstention on screen. That specific moment — "it refused to invent, and cited the rest" — is the thing a chatbot and a template bundle both can't show, and it's the demo that converts the skeptic. Then price the unit so a volume-applier can't trivially default to free.
Don't argue that grounding matters. Show one honest gap on screen and let the buyer feel the difference.

Five calls for Sam batched · recommendations included

These convert the research into the build. Each has a recommendation; say yes/no/adjust and the landing + v1 copy follow.

  1. Pricing unit — capped trial rung"first application CHF 9," visibly capped against the CHF 99 Active Search Pass ("pays for itself by #11"); + CHF 20 headshot add-on. recommend yes Un-capped per-app pricing silently out-costs the Pass for a RAV seeker (10–12 apps/mo) and confuses the everyday professional — CHF 9 wins the trial, the Pass/Radar wins volume.
  2. Headline positioning — adopt: "Your real CV → a ready-to-send Swiss application. Grounded, cited, never invented. Pay once, no subscription." recommend yes Leads with the felt outcome, proves with the moat, reinforces with pay-once.
  3. Headshot is add-on only — never a standalone SKU; sells at +CHF 20 inside the dossier flow. recommend yes Standalone loses to the pure-plays on price/volume.
  4. TWINT + guest checkout + money-back — the Swiss impulse-payment + no-account + de-risk trio for the first stranger franc. recommend yes Guest checkout needs the Core guest-endpoint merge (PR exists).
  5. The converting demo — make "a citation resolving + one honest abstention on screen" the hero moment of the live compose. recommend yes It's the one thing neither ChatGPT nor Profile Bakery can show.
What changes in the build (all cheap, mostly copy): hero + compose show price and the pay-once line · the compose view surfaces one real citation + one honest gap prominently · checkout is guest + TWINT + money-back with the unit stated in words · headshot demoted to an in-flow add-on. None of this is new engineering on the spines — v1 already does the hard part. Companion docs: feature scope · v2 roadmap · architecture · the live demo.