What AI-generated answers look like in an open-end
Six patterns you can automate — and two you should never use.
PandaPanel
Owned panel, 47 direct suppliers, vetted experts. Every source traceable.
SourcesNEW Semantic QC — 41% more bad open-ends stopped
Launch in an hour.
Data you can defend.
Audience sourcing, fielding, cleaning and analysis in one workspace. Brief today, analyse tomorrow — and never pay for a complete you cannot use.
4.9 / 5
Used daily by 320+ research teams
Used daily by 320+ research teams
Urban commute study · Wave 3
Owned panel + 3 direct suppliers
COMPLETED
1,284/1,600
QUALITY
96.4
CPI
$2.90
Tier-1 cities, 18–3492%
Tier-2 · 60+ min commute78%
EV owners (supplemental)64%
37 completes blocked
Flagged before they bill
6.4h median field time
From launch, not kickoff
Trusted for everyday decisions at
0M+
Completes delivered
Trailing 12 months
0
Markets covered
Sourced locally
0+
Targeting attributes
Dated and refreshable
0h
Median field time
Consumer studies
Built in
Four parts, one project.
Finding the audience, waiting on quotes, chasing fielding and cleaning the file used to be four vendors.
Quality control
Identity, behaviour and semantics checked on every complete.
How QC worksPandaScope
Feasibility, CPI, conversion and spend on one screen, live.
DashboardPandaStudio
18 question types, branching, quotas and 28 languages.
BuilderQuality control
Dirty data is rarely malice.
It is a missing gate.
Three gates on every complete. Fail any one and it stops there — at our cost, with the reason written into your report.
Identity
Cross-referenced identity, employment and device signals. Multi-account clusters are frozen.
Behaviour
Question timing and consistency modelling catches straightliners at submit.
Semantics
Open-ends reviewed for AI cadence, drift and zero information — with a readable reason.
QC engine · live feedTHR 0.82
R-48192 · open-end: "why did you switch"Pass0.94
R-48193 · text restates the stemReject0.31
R-48194 · 11 questions in 92 secondsReview0.58
R-48195 · device linked to 3 accountsReject0.12
R-48196 · matrix consistencyPass0.88
R-48197 · open-end semantically emptyReject0.27
This wave: 37 / 1,321 rejectedCost carried by us
Audiences
The people you need to ask —
we probably already know them.
People who can answer for themselves
Identity-verified individuals with a purchase history. Attributes carry a last-collected date, and stale profiles are down-weighted automatically.
Audience detailTargeting attributes
AgeGenderCity tierRegion / DMAHousehold incomeChildren by ageTenureCar ownershipBrand affinityPurchase frequency
Comparison
Panel is slow. Traffic is rough.
Not everyone should buy from us. So here is the comparison, laid out to argue with.
| Capability | PANDAOPINIONSelf-serve + QC | Traditional panel | Ad-network traffic |
|---|---|---|---|
| Live audience size and price | Yes | No | Partial |
| Sources compared side by side | Yes | No | No |
| Semantic review on open-ends | Yes | Add-on | No |
| Cross-project de-duplication | Yes | Manual | No |
| Every rejection documented | Yes | On request | No |
| Pay as you go, no minimum | Yes | No | Yes |
| Median start, consumer sample | 3 hours | 2–5 days | 1 day |
| Export and API delivery | Yes | Ticket | Limited |
What teams say
Nobody was talked into it.
They just never went back.
A tracker wave used to sit in a queue for four days. Now we lock the questionnaire in the morning and read results the next day.
Elena Marsh
Head of Consumer Insight · Northwind
It tells you whether the data is usable before you have to ask. Quality score, rejection reasons, supplier performance — one view.
Linda Zhou
Research Director · Cloudline
B2B decision-makers are the hardest audience there is. Their verification was stricter than our own CRM.
Marcus Reid
Research Manager · Harbor Bank
Start your first project
Brief it today. Work with usable data tomorrow.
$500 in sample credit on your first order. No minimums, no contracts, no platform seat fees. If you would rather have someone run it with you, one of our researchers will field the whole study.
No credit card 14-day refund window Export to SPSS / Excel / CSV
FAQ
A few things worth
confirming first.
Not covered here? Email support@pandaopinion.com — you will hear back within two hours during business hours.
Talk to usMinimum spend, or platform fees?
Neither. You pay only for valid completes at the CPI quoted at launch. Your first order includes $500 in sample credit.
Can I keep my current questionnaire platform?
Yes. Paste your survey link and add four callback parameters. Qualtrics, Decipher, Voxco, SurveyMonkey, Typeform and custom builds all work.
What counts as a valid complete?
One that passes all three quality layers and satisfies your routing. Rejects are not invoiced, and every reason is itemised in your report.
My audience is very small. Can you still find them?
Low-incidence groups go through supplemental recruitment, usually in 5–12 business days. You get a feasibility range up front, not a promise.