Building the next product during a 14-day verdict wait — productive impatience

14 days is a long time to do nothing. So I'm doing something productive that
won't contaminate the experiment: pre-building the next product.

If the verdict on June 5 comes back ≥ 1 sale, the $99 ISO 17025 Certificate
Generator launches June 12. Pre-built during the wait, that timeline becomes
7 days. Spec-locked, core-built, web-form-done. I'd otherwise need 14 days
post-verdict.

What I built in the last 6 days

Python CLI (gen_cert_from_form.py): takes a JSON form + country code
(KOLAS / A2LA / UKAS) → produces a filled .docx via python-docx → converts
to PDF via LibreOffice headless. Generates a 3-section certificate
(results table, Welch-Satterthwaite + Student-t uncertainty, conformity decision).
3-step web form (cert_form.html): vanilla HTML/CSS/JS — no framework,
no localStorage, no analytics tracking. ~5 min to fill. Submits to mailto
fallback so even if my webhook server is dead, the buyer's form data gets
to my inbox.
3 sample PDFs — one per country, all 11 sections filled with realistic
digital multimeter calibration data. These ship in the receipt of the $99
product, so the buyer's expectation is calibrated before they fill the form.

What I deliberately did NOT build

No auto-PDF webhook (manual workflow for first 10 buyers — by design)
No payment integration beyond Gumroad redirect (manual JSON-to-PDF for now)
No multi-currency for cert generator at launch (USD only)
No lab logo support (alpha v0.5 = uniform layout)
No bulk upload (single cert at a time)

These come in v0.6 → v1.0 after the first 10 sales prove there is a buyer.

Why I think this is productive, not contamination

The 14-day toolkit experiment is testing one question: does the niche
(ISO 17025 lab) buy a $19 Excel toolkit from a cold marketplace?

The certificate generator is testing a different question: do the same
buyers also pay $99 for the format layer that wraps the same math?

These are independent. Pre-building the cert generator doesn't change the
toolkit conversion rate. It only changes how fast I can ship product 2 IF
product 1 validates.

If product 1 fails on D+14 (Variant B: 0 sales), the cert generator pivots
into "alpha deliverable for consulting clients" — same files, different use case.

What this teaches me about waiting

The hard part of an experiment is not the work. It is the waiting. Most people
fill the waiting time with variable-changing (lowering price, adding features,
running promos). That contaminates the data.

Productive waiting means: do work that doesn't change the variables of the
experiment you're running, but does compound your option value when the
experiment resolves.

The cert generator alpha is exactly that. Whether D+14 is +1 sale or 0 sales,
the cert generator is shipped, ready, useful. No regret either direction.

What's coming next

D+8 (5/30): another dev.to article — "The split funnel"
D+9-13: Beehiiv Week 10 (D+7 mid-verdict) — will share metric snapshot
D+14 (6/5): verdict email + variant publish (3 pre-drafts written)
D+15+: depending on verdict, either cert generator launch path or direct outreach pivot