Pollster Directory
UK Polling Companies: Full Directory & Methodology Guide
The UK polling industry is dominated by around ten active firms publishing regular voting intention data. Each uses different methodologies, sample sources, and weighting approaches — which is why their figures often diverge. This directory covers every major UK pollster, their methodology, typical sample sizes, publication frequency, and performance at the 2024 General Election.
Active UK pollsters at a glance
| Pollster | Method | Typical Sample | Frequency | 2024 GE MAE | Profile |
|---|---|---|---|---|---|
| YouGov | Online panel | 1,600–2,000 | Weekly+ | 2.5 pts | View → |
| Ipsos | Online (KnowledgePanel) | 1,000–1,500 | Monthly | 3.0 pts | View → |
| Survation | Online + phone | 1,000–2,000 | Fortnightly | 3.8 pts | View → |
| Kantar | Phone + online | 1,500–2,000 | Monthly | 3.2 pts | View → |
| Opinium | Online panel | 2,000 | Weekly | 2.8 pts | View → |
| Redfield & Wilton | Online panel | 1,500 | 2–3× weekly | 2.7 pts | View → |
| Techne UK | Online panel | 1,664 | Weekly | 3.1 pts | View → |
| Deltapoll | Online panel | 1,500–2,000 | Irregular | 3.5 pts | View → |
| More in Common | Online panel | 2,000+ | Monthly | 2.9 pts | View → |
| Savanta | Online panel | 2,000 | Fortnightly | 3.3 pts | View → |
Methodology comparison: online vs. phone vs. hybrid
The single most important methodological divide in UK polling is between online panel polls and telephone (CATI) polls. Different recruitment pathways produce samples with subtly different political profiles, creating systematic inter-firm differences even when all firms apply best-practice methodology.
Online panel polling
Who uses it: YouGov, Redfield & Wilton, Techne, Opinium, Savanta, More in Common
Strengths: Fast turnaround, low cost, high volume, large sub-group samples.
Limitations: Risk of panel conditioning, internet-only population, younger skew before weighting.
Telephone (CATI) polling
Who uses it: Kantar (hybrid), Ipsos (KnowledgePanel)
Strengths: Historically gold-standard coverage, captures non-internet users.
Limitations: Expensive, slow, response rates under 5%, hard to reach younger voters by phone.
Hybrid methods
Who uses it: Survation (online + CATI), Kantar (dual-frame)
Strengths: Better demographic coverage, reduces mode-specific bias.
Limitations: More complex weighting, higher cost, potential for mode effects.
2024 General Election accuracy
The actual result: Labour 33.7%, Conservatives 23.7%, Reform UK 14.3%, Liberal Democrats 12.2%. Most pollsters overstated Labour and understated Conservative support. YouGov performed best on mean absolute error. See the GE2024 accuracy league table for full per-party analysis.
| Pollster | Lab | Con | Reform | LD | MAE |
|---|---|---|---|---|---|
| Actual result | 33.7% | 23.7% | 14.3% | 12.2% | — |
| YouGov | 37% | 21% | 13% | 13% | 2.5 |
| Redfield & Wilton | 36% | 22% | 15% | 12% | 2.7 |
| Opinium | 36% | 22% | 13% | 13% | 2.8 |
| Ipsos | 36% | 20% | 16% | 13% | 3.0 |
| Kantar | 37% | 20% | 15% | 13% | 3.2 |
| Savanta | 37% | 20% | 15% | 13% | 3.3 |
| Techne | 36% | 21% | 16% | 12% | 3.1 |
| Survation | 37% | 18% | 17% | 13% | 3.8 |
MAE = mean absolute error across Lab, Con, Reform UK, LD. Lower is better. Full 2024 accuracy analysis →
House effects
Persistent differences between pollsters — called house effects — arise from panel composition, weighting choices, question wording, and undecided treatment. In 2026, Techne shows Reform UK approximately 2–3 points higher than YouGov on average. More in Common consistently shows the Conservatives 2–3 points above the cross-firm average. The house effects explainer covers the causes and adjustment methods.
BPC membership and transparency
The British Polling Council sets transparency standards. All major UK pollsters are BPC members, required to publish full data tables, methodology notes, and fieldwork dates within two working days. Data is deposited at the UK Data Archive at the University of Essex.