A 5-Minute Test That Prints a Receipt of Your Calibration

By the BatchBrain founder — the competence check now ends in a downloadable receipt, not a badge. Here's what shipped and why it's built that way.

A certificate says you were in the room. A receipt says what you actually did, and what it was worth. We've spent months building BatchBrain around that distinction — and this week the thing we're most proud of finally shipped: the check now ends in a receipt. Not a badge, not a certificate. A receipt.

Here's what it looks like, why it's built that way, and where you can get your own in five minutes.

Certificates measure attendance. Receipts measure competence.

Every developer has a folder of certificates that prove one thing: at some point, you finished a course. They don't prove you can retrieve any of it three months later. They can't, because a certificate records completion, not ability.

A receipt is different. A receipt records a transaction — a specific thing that happened, with specific numbers on it. So the question that started BatchBrain was: what would a receipt for your knowledge look like?

The answer is a five-minute competence check. You answer questions from memory — no searching — and you rate how confident you are in each answer, from "guessing" to "absolutely certain." At the end you get two things:

• A competence level per skill (Beginner → Expert), derived from whether you were actually right.
• A calibration analysis — the gap between how confident you were and how correct you were. Do you overestimate yourself, underestimate yourself, or call it accurately?

That second one is the part we care about most. The most common failure mode in self-taught learning isn't not knowing. It's not knowing what you don't know. A confidence-rated check makes that blindness visible — in five minutes, in black and white.

The receipt

The results page now draws an actual receipt — a paper-white card with the vocabulary of a lab report, not a trophy case. It has:

• Your score in big type, with a color-coded level pill next to it (Expert, Proficient, Competent, Beginner).
• A competence meter — a gradient from Novice to Expert with a needle at exactly where you landed.
• The line that actually matters: SELF-AWARENESS — CONFIDENCE vs ACCURACY, with your average confidence on one side and your real score on the other.
• The verdict underneath it — "You rated your confidence 30 points higher than your actual performance" is a much more useful sentence than "Good job!"
• A reference code and a faux barcode, because a receipt should look like it can be audited.

It renders as a 2400×1260 PNG you can download or share directly — the image carries your real numbers, not a stock congratulation. The share line is the anti-brag: "I scored 70% (Proficient) on the BatchBrain 5-minute competence check. I rated my confidence 30 points higher than my actual performance. Take it free: batchbrain.com/check"

Why "receipt" instead of "share card"

We almost shipped this as a generic "share your score" card — a pretty frame around a number, the thing every quiz app does. Then we asked what would actually make someone post it, and the honest answer wasn't "look how smart I am." It's "here's what I actually know, including the gap."

A receipt is proof, and proof is shareable in a way a humblebrag isn't. If you're a bootcamp grad trying to stand out, "I can demonstrate exactly what I know and what I don't" is a stronger claim than "I have a certificate." If you're self-taught, the calibration number is the most honest signal about you that exists. The receipt is just that signal, rendered as an artifact you can put somewhere.

Where it stands

Honest numbers, because that's the whole point of a receipt: 23 active users, £0 revenue, one newsletter subscriber. The funnel is simple — content → the free check → the receipt → signup — and we're watching two numbers before anything else: check completion rate, and how often that receipt actually gets shared.

The check is free, needs no account, and takes about five minutes:

Take the check →

Download the receipt, and if it's worth a damn, post it somewhere you learn. Then tell us what breaks: would you share a result that showed your calibration gap, or is that too honest for public consumption?