Spark & Anvil Privacy

For grown-ups

How to use it

  • Pick a bucket — 👍 What worked, 😕 What confused, or 🔧 One thing to fix — type what you noticed, and Add.
  • Notes sort into three columns so patterns show up. Remove any with its ×. It's a place to think, not a survey to send.

Why three buckets

"It was cool" doesn't tell you what to do next; "the save button confused them" does. Structuring feedback into worked / confused / to-fix is exactly how designers turn a watching session into an action list — and it teaches a young maker to watch for confusion, not compliments. We claim no product outcome; it's an organising tool.

No pressure, no counting

No score, no streak, no "how much feedback." The columns just hold what you saw. Everything is on this device only.

← Feedback · About

Where this idea comes from· Wiener 1948 · IBM 1950s

Beta, and feedback

In the 1950s IBM engineers checked a new machine with an "A" test before it was announced and a "B" test before it was built for sale; the names became "alpha" and "beta". In 1948 Norbert Wiener called the loop where a result comes back and changes the next try "feedback", in his book Cybernetics.

One thing to hold: A beta is an early warning, not a vote: the testers tell you what broke, and you decide what to change.

Why it matters: Trying something small, hearing back and trying again is how most things that work got that way.

What we don’t know: Who first said "beta test" rests mostly on the memory of one IBM manager, Martin Belsky; no 1950s IBM paper we could read says it.

Three words:

  • Beta The second letter of the Greek alphabet: the second test. What is the one thing your beta needs to find out? (Merriam-Webster, "beta test")
  • Feedback Output fed back in as input. Whose feedback would change your next try most? (Wiener, Cybernetics (1948))
  • Iteration From the Latin iterum, "again". How many tries would you plan for? (Merriam-Webster, "iteration")

Sources: Norbert Wiener, Cybernetics (1948) · Wikipedia, "Software release life cycle" (the IBM A/B/C tests and Belsky’s account)