Skip to content
Idea Atlas
Sign in

DocPress

As described:StopNarrowed:Continue

No as a generic PDF or document API. Go as a German e-invoice API for developers.

e-invoicing · ~$75 a month per customer · last reviewed (next review by 29 Dec 2026) · 69 cited sources

Save
Share

← All research

5. Pre-Mortem: It's September 2028 and DocPress Is Dead

Three most likely causes of death, ranked by probability. They compound: #1 usually triggers #2.

Cause #1: The buyer solved it without you (probability ~45% of failure cases)

What happened. The Q4 2026 rush was real, but it went elsewhere:

  • Developers at German SaaS companies found Mustangproject and horstoeko/zugferd, spent a weekend on them, and shipped.
  • Small businesses moved their invoicing into DATEV, lexoffice or sevdesk.
  • Stripe-billed companies clicked "Install" on Billit, which was free until 2027.

DocPress launched in December with a good API and 6 customers. The validator page got traffic, and people used it once and left. By mid-2027, "we'll just use the library" was the answer on every call.

Early warning signs

  • Day 21: fewer than 2 of 10 design-partner calls end with payment. The usual objection is "we'll use Mustang".
  • Validator users convert to signups at below 2%.
  • Most signups are from companies under the €800k threshold, which have until 2028 and aren't in a hurry.
  • The paying customers are agencies doing one project each, not companies with steady monthly volume.

Fixable? Yes, if caught by day 90. Move to the buyer who can't use a library alone:

  • vertical software vendors shipping e-invoices to all their users;
  • platforms with complex invoices (credit notes, multiple entities, many languages);
  • anyone who also needs inbound parsing.

Sell certainty and saved maintenance ("the rules change; we track them"), not file generation.

Cause #2: Commoditisation and price collapse (probability ~35%)

What happened.

  • Apify actors at $10 per 1,000, AI-built clones (as happened in VAT validation in 2025–26) and a ZUGFeRD feature added to APITemplate.io or CraftMyPDF pushed the market price toward zero.
  • DocPress cut Starter to €9. ARPA fell to ~€25, and churn rose to the ~6% typical of that price band.
  • With $600 new MRR a month and 6% churn, the ceiling became $10k. The business plateaued at $4k, the same shape as ScreenshotOne's curve at a lower level.

Early warning signs

  • Blended ARPA falls below €45 by month 9.
  • Monthly revenue churn is above 4% after month 6.
  • More than 30% of new customers pick the cheapest plan and stay there.
  • A known PDF API (APITemplate.io, CraftMyPDF, DocRaptor, PDFMonkey) announces e-invoice output.

Fixable? Partly.

  • Hold the price, and add value competitors don't copy quickly: inbound parsing, readable errors, rule-version tracking, Stripe connector, German support, batch at month-end.
  • Sell to the €79–199 segments (platforms, agencies, vertical software).
  • Don't fix it by cutting the price. That's the move that caps you.

Cause #3: A compliance or trust event, or the rules moved (probability ~20%)

What happened. One of these:

  • (a) A validation rule update (XRechnung versions change regularly) wasn't applied in time. A customer's large client rejected 400 invoices at month-end, and the customer posted about it in a German developer community.
  • (b) Germany softened or delayed the 2027 duty, or tolerated plain PDFs longer, and the urgency disappeared for a year.
  • (c) Peppol became the expected channel in Germany faster than planned, and file-only APIs looked incomplete.

Early warning signs

  • No automated check that watches the KoSIT and XRechnung release pages.
  • Any document returned "valid" that a receiver rejected.
  • Support tickets saying "is this still required?" after political news on the mandate.
  • Receivers asking your customers for a Peppol ID.

Fixable? Mostly.

  • (a) is fully fixable: automated rule tracking, a release test suite of 100+ reference invoices, and a changelog with email alerts.
  • (b) is a timing risk. The ViDA 2030 cross-border rule and France (2026/2027) keep demand alive, so keep costs low enough to wait.
  • (c) is handled with a partner access point integration (e-invoice.be or Recommand), not by becoming one.

Honourable mentions (less likely, still real)

  • Founder-market distance. If nobody on the team reads German, docs, content and support sound foreign to a trust-sensitive buyer. That's why it's a condition.
  • Month-end outage. An API that fails on the 31st loses the customer on the 1st. Status page, queues and retries from day one.
  • Stripe ships native e-invoicing. Stripe currently relies on partners. If it builds the output itself, the Stripe-connector segment shrinks. Inbound parsing and non-Stripe stacks stay yours.
  • Liability claim. A customer blames DocPress for a tax problem. Handle it with clear terms ("validated against rule version X; tax content is your responsibility"), professional indemnity insurance, and never calling it tax advice.