Pilot

One study, measured two ways.

Prove the data-quality lift on your own panel before you subscribe. We run a continuous-human check alongside your normal intake on a single study, and score it against the duplicate and fraud rate you already track.

How the pilot runs

  1. Scope one study. You pick a study you are about to field. Nothing about the study changes; we sit at intake.
  2. Two arms, same respondents. Every prospective participant is scored by your existing intake and by the Pupul continuous-human check. We never see your survey content.
  3. Measure against your own signals. We compare the two arms on the data-quality measures you already use: duplicate rate, failed attention, straight-lining, impossible metadata, and any fraud flags.
  4. You get the readout. A short report with the sample size, the two-arm comparison, and the null rate, plus the raw pass/flag list so you can audit it. Honest nulls: no record returns null, never false.

What you keep

The lift

Whatever cleaner data the check produced on that study is yours, whether or not you continue.

The method

A repeatable intake step you can keep running, on the same four claims, over your normal tooling.

The proof

Every continuous-human claim is checkable against a public transparency log, so your quality story is auditable, not asserted.

Your participants' trust

The record is theirs, opened by consent, revocable in two clicks. You verify without collecting anything they wrote.

What it costs

$999 for the pilot study

Runs on the Department plan for one study. If you continue, it credits toward your first paid month. Larger panels are quoted; smaller studies can start on the Solo or Lab plan instead.

Book a 30-minute pilot call

Bring one study you are about to field. We will scope the two-arm setup, confirm the measures, and get you a start date. No slides.

Pupul is a proof-of-human signal, not a consumer report, and the record is owned by the person it describes. The pilot measures your own data-quality signals; we do not fabricate or promise a specific number before it is measured. See the open methodology.