How it works
The internet is loud. Comments are chaos, polls are gamed, and real signal gets buried. NoiseGate is a structured filtration system — a place to ask real questions and get crowd judgment you can trust.
Question types
Present 2–6 options and let respondents choose one. Good for logos, names, concepts, or any A/B decision. Results show vote counts, percentages, and signal clarity.
Ask respondents to rate something on a 1–5 scale. Results show mean, median, distribution, and signal clarity based on how much respondents agreed.
Collect short written impressions (140 characters). Good for qualitative texture alongside structured questions. No signal clarity score — words resist reduction.
A binary choice — the clearest possible question. Results show the split as a visual bar. Signal clarity reflects how decisive the crowd was.
Ask respondents to rank up to 6 options from best to worst. Results show average position per option, aggregated across all respondents.
Signal clarity
Signal clarity tells you what the crowd said — whether responses clustered around one answer or scattered across many. It appears on results pages once enough responses have been collected.
Confidence tiers
Clarity tells you what the crowd said. Confidence tells you how much to trust it. Every results page shows a confidence tier based on how many responses have been collected.
Statistically robust. Cite this with confidence.
Signal is reliable for most decisions. Safe to act on.
A pattern is visible but more responses are needed.
Thresholds vary by question type. Ratings and rank order require fewer responses than polls. See exact thresholds →
Filtration capacity
Plants filter. Wetland grasses pull toxins from water. Forests clean the air just by existing. On NoiseGate, your participation builds filtration capacity — the more you contribute thoughtfully, the more your responses carry weight in filtered results. The gate gets better as the garden grows.