Tell the truth before chasing growth
Quote the smaller inventory, the longer timeline, and say no when it is a no. The most expensive thing in this business is a client making a decision on bad data.
About us
We just wanted "ask the right people"
to stop being the hard part.
PandaOpinion Limited is a global survey sample provider and market research solutions partner, committed to delivering high-quality, verified survey sample resources and reliable supply services to market research firms, survey platforms, and business partners worldwide. Through our global sample network and professional project operations, we support online research projects across multiple countries and regions, helping our partners improve sample sourcing efficiency, data quality, and overall project performance.
2019Founded
86Full-time team
320+Client organisations
92Markets covered
What we think
The problem was never
"not enough" sample.
Sample prices halved over the past decade, and confidence in the conclusions did not rise with it. Cheap sample turned "completes delivered" into the KPI, and nobody owned whether the data could support a decision.
So we sell three things together: real people, reviewable quality control, and a conclusion we will put our name on. Price is negotiable. Those three are not.
Something said internally a lot: if we would not make this decision on this file ourselves, it should not reach a client.
What we optimise for, in orderInternal agreement
1Fixed
Data credibility
Rejects not billed, sources traceable, results reproducible
2Fixed
Respondent experience
Capped invitations, on-time rewards, protected privacy
3Improving
Speed to data
Standard consumer studies start inside 24 hours
4Negotiable
Price
Pay as you go; we do not compete on being cheapest
How we got here
Seven years, mostly
pushed by projects.
Every milestone below maps to one client's problem. Nobody wrote a product roadmap; we filled the hole in front of us and moved on.
1,860Attributes maintained
47Direct suppliers
2019
It started as an internal tool
The team was running consumer research and was tired of waiting a week for sample, then cleaning the file themselves. So they wrote a fielding tool for their own projects first.
2021
Quality control v1
Timing models and consistency checks became automated rejection, with the rejected completes no longer billed to clients. The first month ran at an 11% rejection rate; within six months it was under 4%.
2023
Cross-border and local sourcing
Local panels in Southeast Asia and Europe, so sample never passed through a reseller. Multi-language fielding on a single logic tree launched the same year.
2024
White label and API
Research agencies started delivering our engine under their own brand. That year, 40% of new revenue came from partners.
2026
Semantic QC
Semantic review joined the pipeline, lifting invalid open-end detection by another 41%. The same year, the client base passed 320 organisations.
What we believe
Three rules that are not
always comfortable
If you cannot recalculate it, it is not transparent
Price broken down by source, a reason for every rejection, and a report stating where the sample came from. Any conclusion should survive a client doing the maths again.
Respondents are not consumables
Panel experience determines data quality. That is why we cap invitation frequency, pay rewards on time, and never sell member data to a third party.
The team
Research, engineering and success
sit in the same room.
There is no wall between pre-sales and delivery. The person who scopes your study is the person who runs it.
Careers
If "close enough"
grates on you too.
We are not chasing a category. The work is unglamorous: getting people to tell the truth, then making the file survive a difficult question. It rewards people who care about the detail.
- Remote-friendly, with research and engineering able to work hybrid.
- No performative overtime; teams set their own schedules.
- $2,000 a year learning budget, spent on whatever you choose.
Senior front-end engineer (data viz)
Own the PandaScope charts and dashboard experience. Canvas or WebGL at volume expected.
Singapore / remote · Full-time
Research manager (CPG)
Run trackers and concept tests end to end. Three years plus in quantitative research, comfortable in front of clients.
London / remote · Full-time
QC algorithms engineer
Behavioural and semantic rejection models. Needs production experience in NLP and anomaly detection.
Singapore / remote · Full-time
Customer success (cross-border)
Support European and North American clients in overlapping hours.
Singapore · Full-time
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