Fact-checked by Grok 2 weeks ago

YouGov

YouGov is a British-headquartered international and group founded in 2000 by and . The company specializes in providing , polling, and consumer insights through a proprietary global panel of over seven million respondents, pioneering the use of methodologies to deliver real-time on political, social, and commercial topics. YouGov has gained prominence for its political polling, frequently demonstrating high accuracy in election forecasting, including being rated the most precise pollster for the 2025 Australian federal election and the 2025 German federal election. As a publicly traded entity on the London Stock Exchange's market, it has expanded operations across , , the , and , emphasizing data-driven over traditional survey techniques. Despite its empirical successes, YouGov has faced allegations of methodological tweaks, such as last-minute adjustments in polls, and scrutiny over potential Conservative bias linked to its founders' affiliations with the UK , though independent assessments describe its output as minimally biased with strong factual reliability.

History

Founding and early development (2000–2010)

YouGov was founded in 2000 by and , who had previously met while working in British politics, as YouGov Dot Com Limited, establishing it as a pioneer in internet-based with both serving as joint chief executive officers. The company focused on leveraging online panels for rapid polling and , differentiating itself from traditional or in-person survey methods prevalent at the time. In 2001, YouGov accurately forecasted the outcome of the General Election and the leadership contest, building early credibility, while , a former political analyst, joined as chairman to guide its political polling efforts. By 2002, the firm re-registered as YouGov Limited, relocated its headquarters to Farringdon in , correctly predicted the winner of the Pop Idol competition, and formed a partnership with for regular polling. In 2003, it launched the YouGov Omnibus service for efficient multi-client surveys and became a founding member of the British Polling Council, committing to transparency in methodology disclosure. The mid-2000s marked significant growth, with the panel expanding to 100,000 members by 2005, enabling YouGov to accurately predict the 2005 results; that year, it also listed on the (AIM) as YouGov plc and introduced BrandIndex, a real-time brand tracking tool. expansion began in 2006 with the acquisition of Siraj and opening of a office, followed in 2007 by the purchase of Polimetrix, which facilitated entry into the and markets and grew the global panel to 1.5 million respondents, alongside Roger Parry assuming the role of non-executive chairman. Further developments included establishing a office in 2008 and accurately forecasting the London Mayoral election, with the acquisition of Clear Horizons in 2009 positioning YouGov as the most frequently cited polling source in media as its panel exceeded 2 million members. By 2010, retired from his CEO role, leaving as sole CEO, while the company acquired Harrison Group to bolster consumer insights capabilities.

Expansion and technological advancements (2011–2020)

During the 2011–2020 period, YouGov pursued aggressive international expansion through organic office openings and strategic acquisitions. In 2011, the company established its first organic office in , , and acquired U.S.-based Definitive Insights, gaining a West Coast presence with an office in . Further U.S. growth included the acquisition of Harrison Group in . By 2014, YouGov entered the market via the acquisition of Decision Fuel, while 2017 saw the purchase of Sydney-based Galaxy Research and 2018 brought offices in and , alongside acquisitions of SMG Insight (rebranded YouGov Sport for media research), Portent.io (rebranded YouGov Signal for audience insights), and Inconversation Media (rebranded Inconvo for content strategy). These moves diversified YouGov's offerings into specialized sectors like and while extending its footprint across , , and . Technological infrastructure advancements supported this growth, including the 2012 establishment of a Technology and Hub in , , and data processing centers in Bucharest, Romania (2015), and Mumbai, India (2017). YouGov's proprietary panel expanded significantly, reaching over 3 million members across 30 countries by 2012, 5 million in 38 countries by 2016, 8 million worldwide by 2019, and 11 million by 2020. In 2013, the acquisition of Opigram introduced a for proactive panel opinion sharing, enhancing real-time data collection capabilities. Key innovations focused on and polling methodologies. YouGov launched Profiles in 2014, an audience planning and segmentation tool, and began developing Crunch (a cloud-based platform) and YouGov Cube (a profile ). The company applied multilevel regression and post-stratification (MRP) modeling to accurately predict the General Election outcome, demonstrating advancements in probabilistic polling techniques. By 2019, YouGov Direct introduced technology for transparent advertising verification. In 2020, amid the , launches included Recommend+ for and brand tracking, DestinationIndex for daily travel tracking, and a Behavior Tracker in partnership with , leveraging the expanded panel for rapid, large-scale insights. These developments shifted YouGov toward integrated platforms, emphasizing scalable, technology-driven research over traditional survey methods.

Global scaling and recent innovations (2021–present)

In February 2021, YouGov launched online panels in 15 new markets across , , the , marking its largest simultaneous geographic expansion to date. This initiative built on its proprietary panel model, enabling rapid access to consumer insights in emerging regions. Concurrently, the company pursued strategic acquisitions to bolster international capabilities, including Charlton Insights in for sports research expertise in March 2021, LINK Marketing Services AG in , and Wizsight in . These moves extended YouGov's footprint into specialized sectors and underrepresented markets, with its global panel surpassing 24 million members by 2023. The scale accelerated in with the €315 million acquisition of GfK's Consumer Services (), integrating behavioral tracking data from established European panels and adding over 3,000 employees across multiple countries. This deal enhanced YouGov's hybrid data offerings, combining attitudinal surveys with purchase behavior analytics, and propelled its presence to over 60 countries. membership further grew to 27 million by late and exceeded 29 million in 2025, supporting to £388.9 million for the fiscal year ending July 31, 2025—a 16% increase from prior levels. On the innovation front, YouGov integrated capabilities through the 2024 acquisition of Yabble, a , to develop AI-enabled products that accelerate analysis while preserving human-validated . This followed the purchase of KnowledgeHound, enhancing automated data querying and visualization tools. New offerings included Recommend+, a recommendation service leveraging for brand strategy, and expanded trackers for real-time trends in destinations and consumer behaviors. These advancements emphasized scalable, distortion-free that augment proprietary datasets with computational efficiency, positioning YouGov to address evolving demands in global .

Corporate Structure and Governance

Leadership and key figures

Stephan Shakespeare co-founded YouGov in May 2000 alongside Nadhim Zahawi, pioneering online polling methods in the UK. As the company's long-serving leader, Shakespeare resumed the role of Chief Executive Officer on February 4, 2025, following the mutual agreement for Steve Hatch to step down from the position after serving since 2021. Under Shakespeare's direction, YouGov expanded from a UK-focused pollster to a global data analytics firm, emphasizing proprietary panels and advanced modeling techniques. Deborah Davis serves as Non-Executive Chair of the Board, providing oversight on governance and strategy since her appointment. Alex McIntosh, appointed in , manages financial operations and reporting for the publicly listed company. Key technical figures include Douglas Rivers, Chief Scientist, who oversees methodological innovations such as multilevel regression and post-stratification (MRP) modeling, and Jonathan Van Parys, Chief Platform Officer, responsible for and . Recent board enhancements include the September 2025 appointments of Belinda Richards, a senior partner at Brunswick Group with expertise in communications and public affairs, and Ian Griffiths, former CFO at Kantar, to strengthen audit, risk, and remuneration functions. These additions followed the March 2025 departure of Nick Prettejohn from the board. Other non-executive directors, such as Ashley Martin, contribute to committees focused on audit and risk management.

Ownership, financial performance, and regulatory compliance

YouGov plc is publicly listed on the London Stock Exchange under the ticker symbol YOU. The company exhibits high institutional , with institutions holding approximately 78% of shares as of July 2025, positioning them to benefit significantly from any appreciation in stock value. Major institutional shareholders include Global Investments with 4.85% (5,718,713 shares), Brown Capital Management LLC with 4.879% (5,699,652 shares), and Liontrust Investment Partners LLP with 4.696% (5,485,676 shares). Insiders control about 9.71% of shares, while the remainder is distributed among companies, private entities, and employee schemes. No single entity holds a controlling , reflecting a dispersed structure typical of mid-cap firms in the sector. Financially, YouGov reported of £388.9 million for the ended July 31, 2025, marking a 16% increase from £335.3 million in 2024, driven by organic growth and acquisitions. Adjusted operating profit rose to £60.7 million, with adjusted at 31.7 pence, indicating operational stability amid demand. The prior year (2024) saw a 30% uplift to £335.3 million, underscoring consistent , though the company noted a one-off loss impacting trailing metrics into early 2025. YouGov maintains a focus on momentum, with full-year 2025 results slightly exceeding expectations despite broader market pressures on share price. On regulatory compliance, YouGov adheres to global data privacy standards, integrating principles from the EU's (GDPR) across its operations despite operating internationally. The company's Global Code of Conduct and Ethics mandates processing of in line with applicable laws, including GDPR equivalents, with no reported major violations or fines as of 2025. Proactive measures, such as integration for consent management, have been implemented to meet GDPR requirements for granular data handling, reflecting an emphasis on responsible data practices without evident enforcement actions. YouGov's compliance framework aligns with and EU listing rules as a FTSE 250 constituent, supported by regular disclosures under the Quoted Companies Alliance Code.

Methodology and Technical Approach

Online panel recruitment and active sampling

YouGov recruits members for its online panels through a combination of open invitations and targeted outreach. Potential participants can join voluntarily via the company's website or , where they provide demographic details during registration. Additional recruitment occurs via web-based campaigns, including ads that direct users to sign-up pages, as well as advertisements in print such as newspapers and magazines. The process is opt-in and non-probabilistic, allowing broad while emphasizing self-selection among users. To enhance , YouGov partners with third-party websites and networks that provide access to underrepresented groups, such as younger individuals and ethnic minorities, who may be harder to reach through standard channels. These strategic alliances enable proactive efforts beyond passive sign-ups, aiming to build panels that reflect varied population segments despite the inherent biases of opt-in methods, such as overrepresentation of more digitally engaged demographics. As of recent reports, YouGov's global panels exceed 29 million members, maintained through ongoing to replace inactive or churned participants. For conducting surveys, YouGov employs active sampling from its , a technique that involves selectively inviting pre-qualified members to participate based on predefined demographic targets. Invitations are distributed via or app notifications to specific subsets of the panel, with controls ensuring that only invited individuals can respond and preventing duplicate or unauthorized completions. This approach functions as a form of quota fulfillment, where the proportion of invites aligns with benchmarks (e.g., , , ) to approximate representativeness prior to post-survey adjustments. Unlike open or passive surveys where any panelist might self-select, active sampling restricts access to maintain sample integrity and efficiency, particularly for time-sensitive research. Active sampling supports rapid data collection from large, pre-profiled pools, enabling surveys of 1,000 to 2,000 respondents within hours or days, but it relies on the 's recruitment quality to mitigate selection biases inherent in non-probability online methods. YouGov has applied this for the majority of its commercial and political polling since the early , combining it with proprietary matching algorithms to optimize respondent selection against census-like targets. Empirical evaluations, such as those in collaborations, note that while active sampling improves demographic balance over unrestricted opt-in approaches, its effectiveness depends on panel freshness and invitation algorithms, with potential undercoverage of offline populations.

Multilevel regression and post-stratification (MRP) modeling

Multilevel and post-stratification (MRP) is a statistical technique that integrates multilevel modeling with demographic post-stratification to generate population-level estimates from survey samples, particularly useful for non-probability samples like online panels. In MRP, individual survey responses are modeled using a hierarchical framework that accounts for variation across subgroups and geographic units, allowing predictions for undersampled areas by borrowing information from similar groups. Post-stratification then adjusts these predictions to align with known population distributions from census data, producing estimates at fine-grained levels such as electoral districts or states without requiring direct sampling from each. YouGov employs MRP to enhance the accuracy of its polling by modeling vote intentions as a function of covariates including , , level, social grade, , past , and EU referendum vote (in contexts), with random effects incorporated for constituencies or regions to capture local variations. The process begins with large-scale surveys, often exceeding 100,000 respondents, where logistic or multinomial regression predicts probabilities for each outcome within demographic-geographic cells. This multilevel structure enables the model to estimate effects that vary smoothly across similar areas, addressing the limitations of quota-sampled data that may not perfectly mirror probability samples. Following , YouGov applies post-stratification by aggregating cell-level predictions weighted by proportions of the in each geographic unit, such as parliamentary constituencies or states, to yield vote shares and seat projections. For instance, in the 2019 general election model, this method produced constituency-level forecasts validated against 2017 results, achieving outcomes within 2 percentage points in many seats. YouGov has extended MRP to incorporate turnout modeling and multi-party dynamics, using it routinely since 2017 for elections in the , , , and , with sample sizes like 45,000 for the 2024 model to support robust uncertainty estimates via simulations exploring thousands of scenarios. While MRP improves granularity over aggregated polling averages, its reliability depends on the quality of covariates, model specifications, and the assumption that relationships observed in the sample hold population-wide; YouGov mitigates this through validation against historical data, such as correctly predicting 93% of constituency winners in both 2017 and 2019 elections. Limitations include potential biases from unmodeled variables or shifts in voter behavior not captured by demographics, though the technique's hierarchical borrowing reduces variance in sparse cells compared to direct .

Data processing and weighting techniques

YouGov applies weighting to survey data as a post-collection adjustment to align the sample with known distributions, ensuring representativeness despite the use of non-probability online panels. This process involves calculating individual weight values for respondents based on demographic and other benchmarks derived from official sources such as censuses or estimates. The primary technique employed is rim weighting, also known as raking or , which iteratively adjusts weights to match the marginal proportions of the sample to target population totals across multiple variables simultaneously. This method minimizes distortions by sequentially scaling weights for each variable until convergence is achieved, typically within predefined tolerances for demographic alignment. For instance, in surveys, raking targets include , , education level, social grade (a classification based on ), and ; in the United States, it incorporates , , , education, and . Additional variables, such as ethnicity, newspaper readership, political attention, or turnout propensity (e.g., self-reported likelihood to vote and past ), may be included depending on the survey's focus, particularly for political opinion polls. Data processing precedes and entails automated cleaning to remove invalid responses, such as speeders or inconsistent answers, followed by verification against quality controls. Weights are then applied in to produce estimates, with effective sample sizes adjusted to account for weighting inflation, though YouGov maintains that their large sizes (over 1 million members in ) support robust adjustments without excessive variance. This approach contrasts with pure by allowing flexibility in initial recruitment while relying on statistical post-hoc correction, which YouGov claims enhances efficiency and accuracy for online .

Services and Business Model

Political and public opinion polling

YouGov delivers political polling services including intention tracking, assessments on issues, evaluations, and electoral to clients such as governments, political organizations, entities, and . These offerings utilize surveys distributed via the company's panels to capture insights into voter behaviors and societal attitudes, often through formats for rapid deployment or custom serviced research for in-depth analysis. In the United States, YouGov partners with to produce ongoing polls on national political topics, including presidential approval ratings and congressional races, while collaborating with on the SLU/YouGov Poll to gauge Missouri-specific voter sentiments on issues like and elections in a framework. Internationally, the firm conducts similar polling across , , and other regions, supporting media and political campaigns with data on local elections and public referendums. YouGov also maintains YouGov Blue, a dedicated unit offering bespoke polling and analytics for progressive and Democratic-leaning clients, staffed by specialists in political strategy, survey design, and academic research to inform campaign tactics and messaging. This segment integrates with broader public opinion services, enabling clients to access weighted datasets from panels exceeding 29 million registered respondents for enhanced representativeness in polling outcomes.

Commercial market research and brand tracking

YouGov provides services encompassing consumer surveys, audience profiling, and analytics to inform strategies on product , , and entry. These offerings leverage the company's panels to deliver targeted insights, often integrating qualitative and quantitative for clients across sectors such as , , and . For instance, YouGov conducts -specific surveys tracking consumer trends in devices and software adoption to aid strategic decision-making. In 2024, the company's overall revenue reached £335.3 million, with growth attributed in part to expanded research capabilities, though specific segmentation for commercial activities is not publicly detailed beyond broader services. A core component of YouGov's tracking is BrandIndex, a syndicated platform offering daily monitoring of health through metrics including awareness, impressions, quality perceptions, value, satisfaction, and recommendation likelihood. It aggregates thousands of interviews daily from representative online samples, enabling against competitors and market in over 30 countries. Clients utilize BrandIndex for campaign evaluation and crisis response; for example, employed it to assess shifts among post-campaigns, revealing improvements in indicators. Complementary tools include CategoryView for industry-specific holistic views combining , , and data, and trackers tailored to client needs for continuous, monitoring. YouGov enhances tracking through ongoing methodological refinements, such as improved response validation to mitigate biases in online self-selection, ensuring metrics reflect genuine sentiment. These services support by linking brand scores to sales outcomes, aiding ROI optimization, though empirical validation of causal impacts varies by client implementation and external market factors. The platform's always-on nature allows for rapid detection of perception shifts, as evidenced by its sensitivity to media events influencing responses.

Data analytics and custom solutions

YouGov offers custom research services that deliver bespoke market intelligence, enabling clients to obtain tailored data optimized for specific business challenges through targeted surveys and analysis drawn from its proprietary online panels. These projects encompass full-service execution, including sample design, questionnaire development, , advanced , and interpretive reporting to inform strategic decisions. As of 2023, such custom solutions have been utilized by brands for audience profiling, competitive , and growth strategy formulation, with YouGov emphasizing integration of for actionable insights. In data , YouGov provides platforms like Crunch, a specialized tool for processing large datasets from surveys, enabling cross-tabulation, segmentation, and to uncover patterns in consumer behavior and preferences. This capability supports custom applications such as performance tracking, where clients commission ongoing monitors with bespoke metrics to assess , equity, and perceptual shifts over time. Analysts and data scientists collaborate on these initiatives, applying techniques like and predictive segmentation to derive causal inferences from . YouGov's custom solutions extend to integrated partnerships, such as with Analytic Partners since April 2023, embedding its consumer data into commercial frameworks for optimized and ROI . These offerings prioritize through proprietary weighting and validation methods, ensuring representativeness in outputs for sectors including , , and . Clients benefit from scalable solutions that combine syndicated data assets with ad-hoc custom modeling, facilitating rapid iteration and evidence-based adjustments to business models.

Performance and Empirical Accuracy

Historical election forecasting outcomes

YouGov has demonstrated varying degrees of accuracy in forecasting UK general election outcomes, particularly through its multilevel and post-stratification (MRP) models introduced in later cycles, which have often outperformed traditional uniform swing projections used by competitors. In the 2010 election, YouGov's final prediction underestimated Conservative support at 35% versus the actual 37%, while overestimating Liberal Democrats at 28% against 24%; its average absolute error across major parties was 2.3 points, comparable to 's 1.8 but better than ICM's 3.3. The 2015 election saw YouGov predict Conservatives at 34% (actual 38%) and at 34% (actual 31%), with an average error of 1.4 points, matching and surpassing Survation's 2.0.
ElectionPartyYouGov Prediction (%)Actual Vote (%)Absolute Error
2010Conservative35372
28302
Liberal Democrats28244
2015Conservative34384
34313
Liberal Democrats1082
UKIP12131
2017Conservative42442
35416
Liberal Democrats1082
The 2017 election highlighted YouGov's strengths, as its MRP model correctly anticipated a with Conservatives at 42% (actual 44%) and Labour surging to 35% (actual 41%), avoiding the widespread error of other pollsters who projected a Conservative ; its average error of 2.1 points exceeded Survation's 0.9 but was below BMG's 2.3. In 2019, YouGov's final MRP forecast Conservatives at 43% and at 34%, aligning closely with actual results of 43.6% and 32.1%, respectively, contributing to accurate seat projections under first-past-the-post dynamics. For the 2024 UK election, YouGov's MRP overestimated 's seat haul at 431 (actual 412) and underestimated Conservatives at 102 (actual 121), though Liberal Democrat projections matched exactly at 72 seats; vote share errors were modest but systematic, with overpredicted relative to the final margin. In US presidential elections, YouGov's performance has been solid but subject to systemic polling challenges, such as nonresponse bias favoring Democrats. Its 2016 MRP model projected leading by margins akin to Barack Obama's 2012 edge, underestimating Trump's Electoral College win amid broader polling misses. For 2020, YouGov polls showed ahead nationally by around 8-10 points in late surveys, closer to the actual 4.5-point popular vote margin than many aggregates but still overestimating in key states like those contributing to the "blue shift" errors. YouGov's MRP approach has earned high marks in post-election analyses for transparency and methodological rigor, though like peers, it has not fully mitigated turnout underestimation among low-propensity voters.

Comparative reliability against traditional pollsters

A 2016 evaluation of U.S. survey methodologies found that YouGov outperformed nine other opt-in online panels and even probability-based samples in matching demographic benchmarks from sources like the and , achieving higher accuracy in variables such as age, race, and education through weighting on non-demographic factors like party identification and ideology. This methodological edge contributes to YouGov's competitive reliability relative to traditional pollsters like Gallup or , which rely more on telephone or mixed-mode probability sampling but often struggle with similar weighting challenges. In general elections from 2001 to 2019, YouGov's final polls exhibited average errors in the Conservative-Labour lead of 1.1 percentage points (), outperforming the industry average of 1.8 points across all pollsters.
ElectionYouGov Error (Con-Lab Lead)Industry Average Error
2001-0.7-4.8
2005-2.0-2.5
2010-0.21.0
2015-6.5-6.4
20174.64.2
2019-1.8-2.1
Positive errors favor Conservatives; YouGov ranked first in 2005 with a 1.0-point vote share error but mid-tier in others like (2.1 points vs. Survation's 0.9). YouGov's multilevel and post-stratification (MRP) models have further distinguished its performance in seat-level forecasts; in the 2019 election, it predicted a Conservative (actual: 80-seat ) while traditional uniform-swing aggregations of polls anticipated a . U.S. election outcomes show similar patterns of industry-wide errors affecting both online and traditional pollsters, with systematic Democratic overestimation in national polls (93% overstated Biden's margin). YouGov's polls contributed to these aggregates without unique outliers, maintaining parity with firms like . In the 2024 UK election, YouGov's MRP forecast underestimated Labour's vote share (37% vs. actual 33.7%) and overestimated Reform UK's (16% vs. 14.3%), mirroring errors across pollsters including traditional ones like , which also overpredicted Labour's margin by boundaries of . YouGov paused intention releases post-election for methodological review, highlighting ongoing challenges in online opt-in sampling such as bogus respondents and subgroup distortions (e.g., for young or adults). Despite these, its empirical track record positions it as reliably equivalent to traditional methods, with MRP providing causal advantages in heterogeneous electorates over simplistic averaging.

Factors influencing predictive success and failures

YouGov's predictive successes have often stemmed from its multilevel regression and post-stratification (MRP) methodology, which leverages large online to generate constituency-level estimates by modeling voter behavior across demographic, geographic, and historical voting covariates before post-stratifying to benchmarks. In the 2017 general election, this approach accurately forecasted outcomes in 93% of seats and correctly identified the national vote distribution, attributing reliability to the model's ability to borrow strength from national samples for local predictions and incorporate like past turnout patterns. Similarly, in the 2019 election, YouGov's MRP projected Conservatives at 43% nationally (actual 43.6%) and a , with errors limited primarily to overestimating seats due to minor tactical voting deviations rather than core model flaws. The firm's ownership of a panel enables rapid, high-volume —often exceeding 10,000 respondents per MRP —reducing sampling variance and allowing frequent updates to capture trends, a factor credited for superior accuracy in stable electoral environments like the 2024 UK and 2025 Australian federal elections where YouGov outperformed peers in vote share projections. protocols, including adjustments for education, political interest, and recalled past vote, further mitigate panel biases toward engaged or urban respondents, enhancing representativeness when benchmarked against census data. Failures, conversely, have arisen from unmodeled shifts in differentials and late-campaign volatility. In the 2015 election, YouGov underestimated Conservative support (projected 32% vs. actual 36.9%) and overestimated (34% vs. 30.4%), primarily due to inadequate capture of higher Conservative turnout among older and rural voters, which standard weighting failed to fully correct despite demographic adjustments. MRP's dependence on assumed stable correlations between covariates (e.g., age-education-vote linkages) can amplify errors if underlying behaviors change, as seen in sparse subgroups where priors dominate, leading to over-smoothing or divergence from actuals during issue-driven realignments like Brexit-era shifts. Online panel recruitment introduces non-response biases among low-engagement groups, such as less digitally literate or apathetic voters, which MRP post-stratification partially offsets but cannot eliminate if panel recruitment skews toward frequent responders; this contributed to inconsistencies in YouGov's polling, where methods have historically underperformed live-phone surveys in volatile races due to higher reluctance among certain partisans. Methodological tweaks, such as ad-hoc adjustments for perceived past errors (e.g., 2017 upward revision of Conservative estimates), risk introducing house effects if driven by internal incentives over data, though YouGov maintains these stem from refined turnout modeling rather than . Overall, predictive variance correlates with election timing—closer fieldwork yields better results—and external validations highlight that MRP excels in multiparty systems with strong demographic predictors but falters amid rapid turnout surges or unobservable enthusiasm gaps.

Controversies and Critiques

Allegations of methodological adjustments and bias

In June 2022, former YouGov political research manager Chris Curtis alleged that during the 2017 UK general election campaign, the company suppressed an internal poll showing strong performance for Labour leader Jeremy Corbyn, including a post-debate survey where Corbyn outperformed Theresa May among some Conservative voters, due to an unrepresentative sample favoring Labour. Curtis further claimed that senior executives pressured staff to implement minor last-minute methodological adjustments to the final voting intention poll published for The Times on June 7, 2017, which increased the projected Conservative lead from 3 points to 7 points by reallocating approximately 2 percentage points from Labour to the Conservatives, despite internal objections. YouGov rejected these claims, attributing the non-publication of the debate poll to its methodological flaws from sample imbalance and noting that their separate MRP model accurately forecasted a hung parliament, unlike the adjusted national poll which overestimated the Tory margin (actual result: 2.5-point Conservative lead). Curtis retracted his allegations the following day, June 9, 2022. Peter Kellner, YouGov's former president, acknowledged in 2022 that the firm had indeed made such methodological tweaks to the 2017 final poll under external pressure to avoid repeating errors from the 2015 election (underestimating Conservatives) and 2016 referendum (underestimating Leave), framing it as a defensive response to rather than deliberate partisan manipulation, though he conceded the changes reduced accuracy by turning a potentially strong prediction into a mediocre one. MP cited Curtis's initial claims to accuse YouGov of pro-Conservative bias, alleging intimidation by senior Conservatives over poll releases unfavorable to Theresa May's campaign. Similar methodological critiques arose in YouGov's 2016 EU referendum polling, where the firm admitted post-vote that its approach overrepresented Leave supporters, contributing to an inaccurate forecast of a Remain victory. Allegations of inherent political bias in YouGov often reference its co-founders—, a Conservative peer, and , a former Conservative cabinet minister—as evidence of structural Conservative favoritism influencing operations. However, empirical analyses of YouGov's historical polling errors, such as comparisons of Conservative-Labour leads across UK elections, find no systematic deviation favoring Conservatives beyond industry-wide variances attributable to sampling and turnout modeling challenges, with YouGov's aggregate performance aligning closely with peers. Independent evaluators, including , rate YouGov's output as minimally biased based on predictive track records, though critics argue online panel recruitment and weighting adjustments (e.g., interlocking recalled vote with demographics) can introduce subtle non-response biases favoring engaged respondents.

Challenges with online sampling representativeness

Online sampling in YouGov's relies on opt-in panels where participants voluntarily join and are recruited via online advertisements or partnerships, leading to inherent self-selection bias as not all segments have equal probability of . This non-probability approach contrasts with traditional random-digit-dial or address-based sampling, which theoretically ensures every individual has a known chance of selection, though both face declining response rates. Critics argue that opt-in panels disproportionately attract internet-savvy, politically engaged, or incentivized respondents, skewing toward and urban demographics unless aggressively weighted, but weighting cannot fully correct for unmeasured variables like motivation to participate. A prominent challenge manifests in demographic subgroups, particularly young adults and ethnic minorities, where opt-in samples exhibit higher rates of implausible or inconsistent responses indicative of "bogus respondents" or survey trolls seeking incentives or amusement. For instance, a December 2023 /YouGov poll reported that 20% of U.S. adults under 30 believed was a , a figure over six times higher than the 3% found in Pew Research Center's January 2024 probability-based survey of the same group. Similarly, opt-in polls have shown elevated claims of rare qualifications, such as 12% of young adults reporting nuclear submarine operation licenses—impossible in reality—compared to near-zero in rigorous panels, with Hispanics displaying even higher inconsistency rates (24% vs. 2% for non-Hispanics). These artifacts undermine representativeness, as fraud or inattentive responses amplify errors in low-propensity groups less inclined to engage online. Historical examples illustrate sampling volatility in YouGov's online exit polls with small subgroups. During the , an initial YouGov survey of just 14 teenagers estimated 71% support for "yes," but a subsequent larger YouGov poll of 16- to 24-year-olds revised this to 51% "no," aligning better with official turnout data and highlighting how limited sample sizes in opt-in recruitment exacerbate and non-coverage of offline or apathetic youth. While YouGov employs multi-level regression post-stratification (MRP) and matched sampling to adjust for known biases, empirical validations reveal persistent gaps against probability benchmarks, particularly in volatile electorates where unobservable traits like drive turnout discrepancies. Such methods improve aggregate accuracy but falter when panel recruitment fails to capture evolving , as seen in post-2016 critiques of online pollsters' underperformance relative to or traditional approaches.

Responses to external validations and peer reviews

In response to the 2016 report "Vendor Choice Matters," which evaluated online survey vendors' performance across benchmarks including question wording, question order effects, and demographic accuracy, YouGov highlighted its superior results, noting the study concluded the firm "consistently outperforms competitors on accuracy." The company incorporated this validation into its public FAQs and methodology defenses, attributing success to active sampling, large panel sizes exceeding 1.5 million members, and rigorous weighting to national demographics like , , , and region. Academic evaluations have further bolstered YouGov's position. A 2013 study analyzing YouGov's British samples against probability-based surveys found them "of high quality, similar to other nationally representative surveys, and representative of the target population," with minimal biases after adjustments; YouGov referenced such findings to counter critiques of opt-in panels' inherent non-representativeness. In addressing concerns over non-probability sampling, the firm maintains that empirical validation through and —updated against data—outweighs theoretical probability advantages, as evidenced by high recontact rates (up to 50% in panel follow-ups) indicating stable respondent behavior. YouGov has leveraged post-election analyses to affirm MRP (multilevel regression and post-stratification) efficacy amid peer scrutiny. Following the 2025 Australian federal election, independent assessments ranked YouGov as the most accurate pollster for vote shares and projections, prompting the company to publicize these outcomes as proof of methodological robustness against sampling doubts. Similarly, in contexts, YouGov cites accuracy via MRP—predicting a Conservative within 1%—to rebut claims of , arguing that transparent modeling and historical track records (e.g., outperforming aggregates in multiple cycles) validate adjustments over unweighted raw data. When confronted with negative validations, such as its 2016 EU poll's overrepresentation of Leave supporters due to unadjusted turnout models, YouGov acknowledged the error publicly, refined protocols for future surveys, and emphasized subsequent improvements like demographic to mitigate similar issues. Against broader critiques of panels' vulnerability to self-selection , YouGov responds by prioritizing observable performance metrics—such as alignment with actual results—over probabilistic purity, while conducting internal like detection and panel refreshment to sustain credibility. This approach aligns with industry guidelines from bodies like ESOMAR/WAPOR, where YouGov contributors advocate for transparency in non-probability methods.

References

  1. [1]
    Timeline - YouGov plc
    The company is co-founded by Stephan Shakespeare and Nadhim Zahawi, who first met while working in British politics. The official company registration name is ...
  2. [2]
    YouGov plc Corporate Homepage
    YouGov is an international online research data and analytics technology group. Our mission is to supply a continuous stream of accurate data and insight into ...About · YouGov stories · Strategy · Timeline
  3. [3]
    YouGov - Overview, News & Similar companies | ZoomInfo.com
    YouGov is an international research and data analytics group. Their core offering of opinion data is derived from their highly participative panel of over 7 ...
  4. [4]
    YouGov was the most accurate pollster of the 2025 Australian ...
    May 13, 2025 · An analysis of how YouGov's vote and seat projections fared in showing the reality of the 2025 Australian federal election.Missing: achievements | Show results with:achievements
  5. [5]
    YouGov was the most accurate pollster of the 2025 German federal ...
    Feb 24, 2025 · An early analysis of how YouGov's vote and seat projections fared in showing the reality of the 2025 German federal election.Missing: achievements | Show results with:achievements
  6. [6]
    YouGov PLC - Reuters
    YouGov plc is an international online research data and analytics technology group. The Company's principal activity is the provision of market research, data ...
  7. [7]
    YouGov - InfluenceWatch
    YouGov claims to be non-partisan. It is often accused of displaying bias toward the U.K. Conservative Party due to its founders' associations with the party ...
  8. [8]
    YouGov Polling US - Bias and Credibility - Media Bias/Fact Check
    Overall, we rate YouGov Polling US as Least Biased based on polling that minimally favors the left. We also rate them as Very High Factual pollsters based on a ...Missing: controversies | Show results with:controversies
  9. [9]
    Why do polling firms like YouGov tweak polls? Because they are ...
    Jun 9, 2022 · The pressure to avoid mistakes can force pollsters to second-guess themselves, just as YouGov did in 2017, says Peter Kellner, ...Missing: key achievements
  10. [10]
    [PDF] Admission to AIM - YouGov plc
    On 12 April 2005 the Company converted from a Limited Company to a Public Limited. Company. 2.2 Details of changes to and the Company's current share ...
  11. [11]
    Corporate News and Press Releases - YouGov plc
    Acquisition of Decision Fuel extends YouGov's presence to China and S.E. Asia. PDF - 161.3 KB. 04 Apr 2011. Acquisition of Definitive Insights. PDF - 21.8 KB.Missing: expansion | Show results with:expansion
  12. [12]
    YouGov completes the acquisition of GfK's CPS |
    Jan 8, 2024 · YouGov agreed to acquire CPS GfK in an all-cash transaction for a headline purchase price of €315 million.<|separator|>
  13. [13]
    YouGov Reports Stable FY25 Performance with Strategic Growth ...
    Oct 13, 2025 · YouGov plc reported stable financial performance for the fiscal year ending July 31, 2025, with a 16% increase in revenue to £388.9 million and ...
  14. [14]
    YouGov acquires Yabble
    Sep 26, 2024 · Initially, Yabble's revolutionary technology will be integrated with YouGov's core Data Products to accelerate the development of AI-enabled ...
  15. [15]
    Pioneering & Accurate Research Technology, Innovation & AI
    Discover how YouGov's pioneering technology accelerates research without compromising data integrity through advanced methodologies & human-validated ...
  16. [16]
    YouGov: Data Analytics & Market Research Services
    At the heart of YouGov is a global online community of responsive panel members, across all demographics, that share their opinions, behavior and reality.About · YouGov · YouGov Platform · YouGov Behavioral
  17. [17]
    Stephan Shakespeare | YouGov
    One of the pioneers of internet research, Stephan has been the driving force behind YouGov's innovation-led strategy. He is the CEO of YouGov.
  18. [18]
    [PDF] Appointment of CEO - YouGov
    Feb 4, 2025 · The Board and Steve Hatch have mutually agreed that he will step down from his position as Chief. Executive Officer (CEO) and Stephan ...
  19. [19]
    Stephan Shakespeare | YouGov
    One of the pioneers of internet research, Stephan has been the driving force behind YouGov's innovation-led strategy. He is the CEO of YouGov.
  20. [20]
    YouGov plc Board of Directors
    Board of Directors ; Deborah Davis. Non-Executive Chair ; Stephan Shakespeare. Chief Executive Officer ; Alex McIntosh. Chief Finance Officer ; Belinda Richards.
  21. [21]
    Our Team | YouGov
    Our Team ; Stephan Shakespeare. Founder ; Alex McIntosh. Chief Financial Officer ; Douglas Rivers. Chief Scientist ; Jonathan Van Parys. Chief Platform Officer.
  22. [22]
    [PDF] Directorate Change - YouGov plc
    Mar 31, 2025 · YouGov, the international research and data analytics group, announces that Nick Prettejohn has stepped down from the Board of Directors and as ...
  23. [23]
    Corporate Governance - YouGov plc
    Members: Ashley Martin (Chair, Non-Executive Director); Belinda Richards (Non-Executive Director); Ian Griffiths (Non-Executive Director). The Committee ...
  24. [24]
    YouGov plc (YOU.L) Stock Price, News, Quote & History
    Find the latest YouGov plc (YOU.L) stock quote, history, news and other vital information to help you with your stock trading and investing.
  25. [25]
    With 78% institutional ownership, YouGov plc (LON:YOU) is a ...
    Jul 1, 2025 · We can see that institutions own the lion's share in the company with 78% ownership. That is, the group stands to benefit the most if the stock rises.
  26. [26]
    Shareholder information - YouGov plc
    Major Shareholders ; Shareholder: T Rowe Price Global Investments ; Percentage: 4.85 ; No. Shares Held: 5,718,713.
  27. [27]
    YouGov plc (YOU.L) Stock Major Holders - Yahoo Finance
    Breakdown. 9.71%, % of Shares Held by All Insider. 65.03%, % of Shares Held by Institutions. 72.02%, % of Float Held by Institutions.
  28. [28]
    YouGov plc Insider Trading & Ownership Structure - Simply Wall St
    Ownership Breakdown ; Private Companies 0.428% 501,195 shares ; Employee Share Scheme 0.945% 1,108,012 shares ; Public Companies 1.23% 1,441,688 shares ; Individual ...
  29. [29]
    Results for the year ended 31 July 2025 - London Stock Exchange
    Oct 13, 2025 · "YouGov has delivered a stable performance ... The consolidated financial statements of YouGov plc are for the year ended 31 July 2025.
  30. [30]
    Financial reports - YouGov plc
    Summary of financial results for the year to 31 July 2025. Revenue. £388.9m. Adjusted operating profit. £60.7m. Adjusted earnings per share. 31.7p. All. All
  31. [31]
    Results for the year ended 31 July 2025 | Company Announcement
    Jul 31, 2025 · YouGov PLC announced its results for the year ended 31 July 2025, reporting revenue of £388.9m, a 16% increase from £335.3m in 2024, ...
  32. [32]
  33. [33]
    Our data commitment - YouGov
    We have therefore decided to incorporate the GDPR's principles as much as possible into our global operations, while complying with all other applicable privacy ...
  34. [34]
    [PDF] Global Code of Conduct & Ethics - YouGov plc
    Any information related to an identified or identifiable person must be collected and processed in compliance with applicable data privacy laws, such as the EU ...
  35. [35]
    Data Provider YouGov Pins Hopes On Blockchain To Ensure GDPR ...
    Apr 24, 2018 · Under GDPR, it's no longer sufficient for YouGov to gain blanket consent to sell information – over 5,000 data points about a panelist's likes, ...Missing: issues | Show results with:issues
  36. [36]
    Methodology | YouGov
    Panel members are recruited from a host of different sources, including via standard advertising, and strategic partnerships with a broad range of websites.How does YouGov conduct... · How does the YouGov Daily...
  37. [37]
    [PDF] ESOMAR 36: YouGov questions to help research buyers of online ...
    Our recruitment process is open to all, and members of the public can easily join our panel by signing up via our website or our app (both Android & iOS). We ...
  38. [38]
    YouGov Recruitment/Methodology - OMB 0920-1291
    Dec 1, 2020 · Recruiting methods include: • Web advertisement join YouGov campaigns - Web search based advertising in which a respondent is directed to a ...
  39. [39]
    [PDF] The YouGov Online Panel - Cloudfront.net
    YouGov actively recruits hard-to-reach respondents, such as younger people and those from ethnic minorities, via a network of partners with access to a wide ...
  40. [40]
    About our panel - YouGov
    Founded in the UK in 2000, YouGov is considered the pioneer of online market research. With operations in the UK, the Americas, Europe, the Middle East, ...
  41. [41]
    Methodology - YouGov
    Panel members are recruited from a variety of sources, including through standard advertising and strategic partnerships with a broad range of websites.
  42. [42]
    YouGov Research Methods
    Active Sampling ensures that the right people are invited in the right proportions. In combination with YouGov's statistical weighting, this ensures that the ...
  43. [43]
    YouGov Research Methods - Reuters Institute Digital News Report
    Jun 15, 2020 · When using Active Sampling, restrictions are put in place to ensure that only the people contacted are allowed to participate. This means that ...
  44. [44]
  45. [45]
    A national survey assessing public readiness for digital health ... - NIH
    Mar 16, 2021 · Survey development. An online survey was co-designed with qualitative experts from YouGov (YouGov PLC, London, UK), a market research company.
  46. [46]
    How YouGov's 2019 General Election model works
    Nov 27, 2019 · In recent UK and US elections, YouGov has been using a technique called Multilevel Regression and Post-stratification (or 'MRP' for short) ...
  47. [47]
    FAQs about YouGov's 2024 general election MRP model
    May 28, 2024 · It is a Multi-level Regression and Post-stratification (MRP) model that uses polling data. Polls aim to ask a representative sample of people a question and ...Missing: explanation | Show results with:explanation
  48. [48]
    How is the data weighted? - Help Center for YouGov Customers
    Oct 19, 2022 · We calculate weight values using rim weighting (raking), which ensures that the marginal proportions in the sample match those of the target ...
  49. [49]
    Government & Politics Surveys, Market Research & Brand Tracking
    Politics is powered by polling and intelligence. With wide-ranging access and insight on public opinion and behaviors, you'll expand your understanding today, ...
  50. [50]
    YouGov Surveys: Serviced | Omnibus Surveys, RealTime Results
    Work with a YouGov researcher with Omnibus surveys and international options. Get fast survey results from nationally representative or target audiences ...Children and Parents Omnibus · International Omnibus · US Omnibus Surveys<|separator|>
  51. [51]
    YouGov /Research Reality
    YouGov is a global public opinion and data company. Explore what the world thinks, discover our solutions, and join our community to share your opinion.United States · Surveys: Self-serve · Market research services and... · International
  52. [52]
    SLU/YouGov Poll : SLU - Saint Louis University
    The SLU/YouGov Poll delivers nonpartisan insights into Missouri voters' opinions to inform civic discourse and support policymaking that reflects citizens' ...
  53. [53]
    YouGov Blue | Democratic Political Research
    Custom research for progressive and Democratic clients with YouGov Blue, staffed with experts from politics, survey research, and academia.<|control11|><|separator|>
  54. [54]
    Technology Surveys, Market Research and Brand Tracking - YouGov
    Explore technology industry market research, latest trends surveys, consumer opinion and brand tracking services to improve strategic decision making.
  55. [55]
    [PDF] Annual Report & Accounts 2024 - YouGov plc
    Nov 5, 2024 · Acquisition risk register. Mergers and acquisitions inherently involve risks such as unforeseen liabilities, financial misrepresentations ...
  56. [56]
    YouGov BrandIndex | Daily Competitor & Brand Tracking
    Daily brand performance. Helping you analyze and benchmark your success against your goals, competitors, and entire market with daily brand tracking.
  57. [57]
    Methodology Overview - Help Center for YouGov Customers
    Oct 19, 2022 · YouGov Ratings measures the popularity and fame for anything and everything. It conducts thousands of interviews every year from a representative sample.
  58. [58]
    Lufthansa gained brand perception with YouGov BrandIndex
    The KPI-based brand tracking of YouGov BrandIndex provided Lufthansa with comprehensive, always up-to-date information on brand perception among existing and ...
  59. [59]
    YouGov CategoryView | Industry-specific brand tracking
    Our industry-specific brand tracking platform combines brand, audience and category-level data, giving you a holistic view of your brand and the wider market.
  60. [60]
    Bespoke Custom Brand Tracker - YouGov
    Custom brand research trackers offer dynamic and integrated performance monitoring across your brand, to get continuous, reliable insights with bespoke data ...
  61. [61]
    Improving response quality in brand tracking: YouGov white paper
    Nov 21, 2023 · ... YouGov has taken over the last year to improve data quality in BrandIndex, describing what they are and their impact on data in the tool.Missing: details | Show results with:details
  62. [62]
    Market Research & Brand Data for Marketing Mix Modeling - YouGov
    Discover YouGov's marketing mix modeling services for powerful MMM, accurate analytics, and custom solutions that drive smarter decisions and higher ROI.
  63. [63]
    [PDF] Introduction to YouGov's Key Products
    "We find Brand Index is a very sensitive and powerful tool for understanding consumer responses in a fast changing market place. If a story breaks in the media ...
  64. [64]
    Custom Research - Bespoke Market Intelligence - YouGov
    Get detailed market research projects tailored to your needs with YouGov's Custom Research. Access unique data optimized to solve your problems, ...Missing: analytics | Show results with:analytics
  65. [65]
    Six ways custom research helps grow your brand - YouGov
    May 17, 2023 · YouGov Custom Research offers a full end-to-end service from YouGov experts, including sample framing, questionnaire design, analysis reports, ...
  66. [66]
    YouGov Crunch: | Powerful Research Analytics Platform
    Explore our connected data solutions · Identify, size & profile your ideal customer · Monitor your brand across metrics that matter · Activate your audience.
  67. [67]
    Custom insights & strategy - YouGov
    Our expert analysts and data scientists collaborate with you to design a research roadmap tailored to your objectives. From comprehensive market reports to ...
  68. [68]
    Analytic Partners Integrates YouGov into its Commercial Analytics ...
    Apr 19, 2023 · We provide adaptive solutions for deeper business understanding and right-time planning and optimization – for marketing and beyond. We turn ...
  69. [69]
    Market research solutions and survey tool services - YouGov
    Data quality and accuracy means decisions you can trust. From self-serve and serviced surveys to custom research, you can rely on YouGov's unmatched, market- ...<|separator|>
  70. [70]
    YouGov Solutions: Market Research Services and Intelligence
    From a survey building platform to fast-turnaround researcher-supported omnibus, right up to bespoke, full-service custom projects, YouGov has market research ...
  71. [71]
    [PDF] YouGov's Record of Accuracy - Cloudfront.net
    YouGov's Record of Accuracy. Public Polling Results compared to other pollsters and actual outcomes. UK General Elections. UK General Election 2017. YouGov.
  72. [72]
    Final 2019 general election MRP model: small Conservative ...
    Dec 10, 2019 · Predicted vote shares in our final poll have the Conservatives on 43%, Labour on 34%, the Liberal Democrats on 12% and the Brexit Party on 3%.
  73. [73]
    How YouGov's seat and vote projections fared at the 2024 UK ...
    Jul 5, 2024 · YouGov's final MRP seat projection of the campaign was Labour 431, Conservative 102, Liberal Democrat 72, SNP 18, Reform 3 Plaid Cymru 3 and Greens 2.
  74. [74]
    The YouGov Model: The State of the 2016 Election
    Oct 4, 2016 · Clinton is currently ahead of Trump in the national popular vote by the same margin that Obama defeated Romney in 2012. Her electoral college ...Missing: 2020 | Show results with:2020
  75. [75]
    What are the best pollsters in America? - ABC News
    Jan 25, 2024 · Pollster A released three surveys of the presidential election in 2020. They showed now-President Joe Biden beating then-President Donald Trump ...
  76. [76]
    2024 Most Valuable Pollster (MVP) Rankings - ActiVote
    Mar 25, 2025 · Let's start with the final results: During the 2024 election season, thousands of polls were published by well over a hundred different ...
  77. [77]
    Pew Research: YouGov consistently outperforms competitors on ...
    May 13, 2016 · Most demographic variables come from either the Current Population Survey (CPS) or the American Community Survey (ACS)--very high quality ...
  78. [78]
  79. [79]
    YouGov: is it biased to the Conservatives? - Mark Pack
    Jan 6, 2022 · If there's an opinion poll published by YouGov with figures that do not look great for Labour or the left in general, it often triggers ...
  80. [80]
    Confronting 2016 and 2020 Polling Limitations - Pew Research Center
    Apr 8, 2021 · Looking at final estimates of the outcome of the 2020 U.S. presidential race, 93% of national polls overstated the Democratic candidate's ...
  81. [81]
    [PDF] Accuracy and Bias in the 2020 U.S. General Election Polls
    Feb 21, 2021 · The results show the pro- Democratic biases reflected in the final 2020 preelection polls were systematic. Keywords: 2020 election, elections, ...
  82. [82]
    A summary of YouGov's review into our 2024 general election ...
    Dec 18, 2024 · ... vote share – but there are ways we could have been more accurate still. Since the general election we've paused our voting intention figures ...<|separator|>
  83. [83]
    Election 2024 polls were wide of the mark on Labour's margin of ...
    Jul 12, 2024 · The pollsters over-predicted the Labour and arguably under-predicted the Tory vote, although in the latter case it was on the boundary of statistical ...
  84. [84]
    Online opt-in polls can produce misleading results about young ...
    Mar 5, 2024 · As polling takes center stage again this election year, understanding its strengths and weaknesses is critical. At Pew Research Center, ...
  85. [85]
    How YouGov's MRP model works for the 2024 U.S. presidential and ...
    Sep 24, 2024 · A guide to the methodology behind YouGov's use of multilevel regression with post-stratification (MRP) to project how Americans will vote in ...Missing: explanation | Show results with:explanation
  86. [86]
    YouGov called the election right. How? | by Max Nathan | Medium
    Jun 12, 2017 · YouGov's MRP (multilevel regression and post-stratification) method not only got the national result right, but correctly predicted results in 93% of seats.Missing: success | Show results with:success
  87. [87]
    Initial analysis of YouGov's election polling
    Dec 13, 2019 · Our seat projection was further from the result. Our MRP model told the correct story - we showed a Conservative majority, with the Labour vote ...
  88. [88]
    2019 UK General Election MRP Post-mortem - Benjamin Lauderdale
    Dec 17, 2019 · The headline Conservative seat prediction was too low (339 vs 365), but on many other metrics the model performed well, capturing many important ...Missing: accuracy | Show results with:accuracy<|separator|>
  89. [89]
    MRP Methodology - YouGov Australia
    Feb 13, 2025 · For the Australian 2025 federal election, each YouGov MRP model will explore over 25,000 different possibilities. The sample size involved - ...Missing: explanation | Show results with:explanation
  90. [90]
    We got it wrong. Why? - YouGov
    May 11, 2015 · In any event, whatever caused last week's errors was no respecter of polling methods. Online and telephone polls were equally wide of the ...Missing: success failures
  91. [91]
    MRP: what it is and will MRP polls get the next election right?
    Aug 20, 2023 · What is multi-level regression and post-stratification (MRP) polling, and how much trust should we put in the individual constituency ...
  92. [92]
    No One Picks Up the Phone, but Which Online Polls Are the Answer?
    Jul 2, 2019 · There is substantial evidence that it remains easier to construct a high-quality live-interview telephone poll than an online poll of equal quality.
  93. [93]
    [PDF] Failure and Success in Political Polling and Election Forecasting
    2020). 2. Survey Errors and Adjustments. Before discussing the successes and failures of political polls and forecasts, we briefly ...
  94. [94]
    YouGov sat on 2017 poll as it was too positive on Labour, claims ex-employee
    ### Summary of Content from https://www.theguardian.com/politics/2022/jun/08/yougov-sat-on-2017-poll-as-it-was-too-positive-on-labour-claims-ex-employee
  95. [95]
  96. [96]
    Ian Lavery on X: "These new allegations suggest that YouGov did ...
    Jun 8, 2022 · These new allegations suggest that YouGov did manipulate the release of polling data in 2017 to have a pro-Conservative bias and that senior ...Missing: methodological adjustments
  97. [97]
    YouGov's Polling Methods Were Flawed - Business Insider
    May 18, 2016 · Pollsters YouGov said in its latest EU referendum survey that the methodology it was using was flawed as it overrepresented supporters of ...Missing: controversy | Show results with:controversy<|control11|><|separator|>
  98. [98]
    The Problem With a Crowd of New Online Polls - The New York Times
    by 538's reckoning perhaps the best poll of 2012 — was arguably the single worst pollster in the 2016 election, and it ...
  99. [99]
  100. [100]
    Survey Trolls, Opt-In Polls, and the New Era of Survey Science
    Jun 26, 2024 · The Economist/YouGov poll was a so-called opt-in poll, in which pollsters often pay people they've recruited online to take surveys.Missing: achievements | Show results with:achievements
  101. [101]
    Sampling in epidemiological research: issues, hazards and pitfalls
    A later, more representative YouGov poll with a much larger sample reported that 51% of 16- to 24-year-olds had voted 'no'. In whatever way the first sample ...<|separator|>
  102. [102]
    YouGov Frequently Asked Questions - FAQs
    YouGov's most frequently asked questions - in-depth questions and answers on YouGov's panel, products & solutions and public data.Missing: involvement | Show results with:involvement
  103. [103]
  104. [104]
    Getting It Right: YouGov and Online Survey Research in Britain
    Aug 6, 2025 · Research shows that YouGov samples are of high quality, similar to other nationally representative surveys, and representative of the target ...Missing: peer | Show results with:peer
  105. [105]
    ESOMAR/WAPOR Guideline on Opinion Polls and Published Surveys
    former Director of Surveys, CBS News & Consultant, YouGov & member of ESOMAR Professional Standards Committee; Miroslawa Grabowska Professor, University of ...