Survey research is the use of structured questionnaires, polling, and statistical analysis to produce defensible numbers about what people think, do, and want. Communications teams use it for three purposes: shaping public opinion, earning media coverage, and understanding customers. The Edelman Trust Barometer, the Gallup Poll, and the J.D. Power studies show how a recurring survey becomes a permanent press asset.
What are the three purposes of survey research?
Survey research serves three purposes: it shapes public opinion, earns media coverage, and tells a company what its customers and employees want. Most major franchises serve more than one, and the strongest programs serve all three.
The first purpose is shaping public opinion. The Gallup Poll and the Pew Research Center studies produce the numbers on candidates, policies, and social attitudes that journalists, regulators, and executives treat as the baseline for a debate.
The second is earning media coverage. The Edelman Trust Barometer, the J.D. Power Vehicle Dependability Study, and the Salesforce State of Marketing give reporters proprietary data on a recurring schedule, which produces coverage that no advertising budget buys.
The third is understanding customers and employees. Net Promoter Score programs, segmentation studies, and engagement surveys such as the Gallup Q12 inform product, marketing, sales, and HR decisions. This work is the most valuable internally and the least visible outside the company.
Edelman shows how the purposes stack. The Trust Barometer produces public-opinion data on trust, gets covered every year, and gives Edelman's own clients a customer-insight asset.
How has survey research changed across five eras?
Survey research has moved through five eras since 1935, and each era cut the cost or time of collecting data while testing the discipline's standards.
The Gallup era (1935 to 1980) began when George Gallup founded the American Institute of Public Opinion. His 1936 call of the Roosevelt and Landon election, against the Literary Digest's wrong prediction of a Landon win, made polling credible. Telephone and in-person interviewing made it slow and expensive.
The direct mail era (1980 to 2000) extended surveys into commercial market research through consumer mail panels, focus groups, and B2B methods, and it built the proprietary-data businesses of Nielsen, Forrester, Gartner, and J.D. Power.
The online era (2000 to 2015) arrived with SurveyMonkey, Qualtrics, and panel providers such as Toluna, Cint, and Lucid. Turnaround fell from weeks to days and brands could field their own research. Panel quality varied widely.
The mobile and programmatic era (2015 to 2022) added mobile delivery, ad-platform recruitment, and links to first-party CRM data. Speed and scale grew while sample quality and response bias became harder to control.
The AI era (2022 to the present) brought AI-assisted analysis, fraud screening, and synthetic respondents. The section on AI below covers what is settled and what is still contested.
Which survey franchises earn lasting press coverage?
The franchises that earn lasting coverage are recurring surveys with a branded methodology, run for years by an organization that benefits from the data. The table groups the most cited ones by what they measure.
| Franchise | Operator | What it measures |
|---|
| Edelman Trust Barometer | Edelman (launched 2001) | Trust in government, business, media, and NGOs across about 28 countries |
| Axios Harris Poll 100 | The Harris Poll and Axios | U.S. consumer views of the 100 most visible companies |
| State of the Global Workplace | Gallup | Employee engagement worldwide; the 2024 edition reported 23% of employees engaged |
| Employee Benefit Trends Study | MetLife | Benefits preferences, mental health, and financial wellness among employees and employers |
| WorkForces Report | Aflac | Employee benefit preferences and the gap between employer and employee views |
| Vehicle Dependability Study | J.D. Power | Long-term vehicle quality, one of many J.D. Power category studies |
| National Retail Security Survey | National Retail Federation | Retail shrink, organized retail crime, and loss prevention |
| American Customer Satisfaction Index | University of Michigan (founded 1994) | Customer satisfaction across more than 40 industries |
| Cost of a Data Breach Report | IBM with Ponemon Institute | Average breach cost; the 2024 edition put the global average at $4.88 million |
| State of Marketing, Sales, and Service | Salesforce | Technology adoption, AI use, and priorities among practitioners |
| Tech Trends | Deloitte | Annual forecast of enterprise technology shifts |
| Global CEO Survey | PwC (annual since 1998) | CEO views on strategy, economic outlook, and risk |
| CIO Survey | Gartner | Technology priorities and IT spending among enterprise CIOs |
| Gallup Poll | Gallup (since 1935) | Presidential approval and public opinion on policy issues |
| American Trends Panel | Pew Research Center | Social, political, and technology attitudes among U.S. adults |
| Health Tracking Poll | KFF | Monthly U.S. attitudes on health policy and insurance coverage |
What makes a research franchise compound?
A research franchise compounds when five conditions hold at once. Missing any one usually produces a single press cycle and nothing after it.
- The survey runs every year for many years, so each edition adds a dated, citable record. The Edelman Trust Barometer and the PwC Global CEO Survey each have decades of editions.
- The methodology carries a name, such as the Trust Barometer or the Q12, so journalists and AI engines cite it by that name.
- The sample is large enough to support sub-group claims, such as country-by-country comparisons.
- The operator builds distribution around the release: conference timing, press briefings, and syndication.
- The research feeds the operator's business. Edelman's data supports its client work, and J.D. Power's studies support its licensing business.
For a brand deciding whether to launch its own research, the practical test is multi-year value. A survey that earns one news cycle is a campaign. A survey that earns citations for ten years is an asset.
How do employee surveys feed corporate reputation?
Employee surveys feed corporate reputation through three channels: published research findings, public review sites, and annual employer rankings. Between 2015 and 2026 they moved from an HR diagnostic into a corporate-affairs asset.
The foundation is the Gallup Q12, a 12-question engagement instrument Gallup developed in the 1990s. Gallup's studies link engagement scores to outcomes such as retention, productivity, and customer loyalty, which is why the instrument anchors most corporate engagement programs.
Platforms built on continuous listening, meaning short pulse surveys in place of one annual survey, now run most enterprise programs. The main platforms are Microsoft Viva Glint (acquired by LinkedIn in 2018 and later folded into Viva), Culture Amp (founded 2009), Lattice (founded 2015), Qualtrics EmployeeXM, and Peakon (acquired by Workday in 2021).
Public review sites add an uncontrolled layer. Glassdoor and Indeed publish unstructured employee reviews that candidates, reporters, and AI engines read next to company-issued material. Companies that respond to reviews with specifics shape what those pages say. Companies that stay silent leave the reputation to reviewers.
Annual rankings turn survey data into press. The Fortune 100 Best Companies to Work For, produced with Great Place to Work since 1998, draws on an employee survey and a review of company programs. The Glassdoor Employees' Choice Awards use Glassdoor ratings. LinkedIn Top Companies, Built In Best Places to Work, and the Comparably awards each run their own annual cycle.
The communications takeaway is coordination. The HR team that runs the survey, the corporate-affairs team that publishes findings, and the team that answers reviews should work from the same data. Internal results that contradict published external research, such as the Gallup or MetLife findings, create a communications risk.
How is AI changing survey research?
AI now touches four layers of survey research: panel recruitment and fraud screening, questionnaire design, data collection, and analysis. It has produced clear gains in analysis and fraud detection and an open dispute over synthetic respondents.
Panel operators use AI to flag bot respondents and survey farms, which raises respondent quality. Researchers use language models to draft questionnaires and spot biased wording. Those drafts still need an experienced human to catch subtle bias, so the working practice is human-led design with AI refinement.
Analysis is the clearest win. Coding thousands of open-ended answers used to take analysts weeks, and language models now produce first-pass codes and themes in hours. The safeguard is human review. Models can misread sarcasm, cultural references, and context, so a person should check the codes before anyone publishes them.
Synthetic respondents are the contested layer. These are AI-generated respondents that simulate survey answers from patterns in training data. They can help with early concept testing and with pre-testing a questionnaire. They cannot measure attitudes that emerged after the model's training cutoff, and most established research operators have not adopted them as a primary measurement method.
Communications teams that publish AI-assisted research should disclose where AI was used, which steps humans checked, and how the sample was built. Ask any research vendor for the same disclosure before commissioning work.
Why are polls wrong sometimes?
Polls are wrong when sampling bias, non-response bias, mode effects, or weighting errors push results away from the true population value. The 2016 and 2020 U.S. presidential cycles showed systematic underestimates of support for Donald Trump in key states, and 2024 showed a partial recovery.
Election misses also damage brand research. Companies that use the same online-panel methods inherit part of the credibility hit when election polling fails. Polling Errors That Change Headlines covers the three cycles and the lessons for brand surveys.
Where do the other survey guides fit?
This page is the hub. The guides below cover the specific questions teams face when they plan, commission, or interpret a survey.
- Survey Methodology covers sampling, weighting, margin of error, mode effects, question wording, and how consumer and B2B surveys differ in sample source, size, and cost.
- How Large Should a Survey Be? explains what 400, 1,000, and 30,000 respondents can and cannot support.
- Types of Research Design compares experimental, survey, case study, and longitudinal designs.
- Polling Errors That Change Headlines analyzes the 2016, 2020, and 2024 polling cycles.
- PR Surveys Done Well shows what separates a survey that travels in the press from one that sits unused.
- Omnibus Surveys explains how agencies share survey costs to produce media-ready data.
- Six Months Into the 2026 Edelman Trust Barometer reviews what the year's flagship trust survey implies for communications strategy.
Four sector guides map the research franchises in specific industries: technology surveys, hotel and travel surveys, food and beverage surveys, and cannabis research and industry data.
