How to Hire a Public Opinion Pollster
If all opinion pollsters were all the same, they would produce the same outcome every time. But, hiring an opinion pollster is like hiring a surgeon: there are great surgeons, average surgeons,
and butchers. This document will help you understand the differences between the best and the worst polling firms. Who you hire to conduct your opinion polling and market research makes
a difference in what results you get back. There are many excellent opinion research firms; use this checklist to make sure you are about to hire one.
At the highest level   Survey
The best Opinion Research Firms
1 Actually Conduct the Research        
    Sad to say, but some marketing companies masquerading as pollsters will take your money and not conduct any research at all. There are a number of prominent cases in recent years, and several additional cases where the research company owners thought they were in fact conducting research, but where the interviewers got tired of asking questions and just made up the answers. When you start with a reputable research firm, everything follows from there.  
2 Do Not Make Up the Data        
    Opinion pollsters can "cook the books," "put their thumb on the scale," and in general, manipulate data to say whatever they want. The best research companies won't do this, ever. The best research companies will show you unweighted data, if asked, and will show you individual respondent data, timestamped, to show the disposition of each "case" in the sample. There should be a clear, replicatable path from each interview to published results. Ask to see unweighted data.  
3 Are 100% Transparent        
    Honorable research companies have integrity, are 100% above board, 100% transparent in all of their dealings. At the highest level, you can find out who owns the research company, and whether the research company has a hidden agenda. On a project-by-project basis, you should insist that every aspect of the project be disclosed, so that once the results are known, other scientists can re-create the project and replicate the results.  
4 Respect Respondents        
    The easiest mistake for a research company to make is to forget that research respondents are human beings, with busy lives, who are not sitting around waiting to be interrupted by a 20-minute to 30-minute "barge-in" unscheduled phone call. SurveyUSA works hard to keep respondents tied up just a few minutes, out of respect for a respondent's busy life. The best respondents are "normal" respondents, not respondents with nothing else to do but talk to a pollster.  
5 Are Run by Scientists, Who Love Science        
    All science requires the research scientist to tell you exactly what he/she has done, so that if another scientist wants to re-create the project, he/she can. SurveyUSA discloses each step of its research process, so that others may examine our work to make sure it is free of bias, full of fairness. Others tell you next to nothing. SurveyUSA goes out of its way to make sure that all sides of a particular issue are able to answer the questions we write.  
6 Are Not on One Side of Any Particular Issue        
    SurveyUSA is fiercely independent. We are not associated with any particular cause or ideology, and this frees us to write fair and unbiased questions. "Garbage-in garbage-out" applies directly to how questions in a research questionnaire are drafted. SurveyUSA editors, writers and pollsters have combined 60 years of experience writing fair, unbiased, neutral, uncomplicated questions.  
7 Are Beholden to None        
    SurveyUSA is not just an independent research company, it is an apolitical research company -- beholden to no cause or ideology, able to work for conservative and progressive causes with equal ease. Many competitors are aligned with one political party or the other, or are in business to promote themselves. These "agenda" pollsters see the world one way. If you want actually to learn something, hire us. If you want to reinforce preconceived notions, hire them.  
8 Have an Outstanding Track Record, Year-After-Year        
    Any pollster can get an occasional election correct. But no pollster over the past 20 years has worked on more separate election contests than has SurveyUSA, and few pollsters can match SurveyUSA's year-after-year track record for excellence. In presidential contests, contests for US Senate, Governor elections, county elections and municipal elections, SurveyUSA consistently distinguishes itself.  
9 Include Hard-to-Reach Respondents        
    1/3 of American Homes, and more than half of young people, do not have a home telephone. Pollsters must make a special effort to reach these homes. If the pollster does not make a special effort, the survey will be undercounting the young. Ask your pollster point-blank if he/she includes hard-to-reach cell-phone respondents, and if so, how many cell respondents. Ask your pollster to contrast home phone and cell phone respondents.  
10 Have Signed the Codes Governing Professional Conduct        
    Professional research associations require those who join to sign codes of professional conduct. SurveyUSA has signed the code of professional conduct of the American Association for Public Opinion Research, the National Council on Public Polls, the Council of American Survey Research Organizations, and the Market Research Assocation. These codes hold the research firm to the highest standards. If your pollster has not signed these codes of conduct, why?  
11 Are Surgeons, Not Surgical-Supply Warehouses        
    There is a difference between hiring a surgeon to perform an operation and going to a surgical supply store and buying a scalpel. Many SurveyUSA competitors want you to operate on yourself. SurveyUSA is a full-service research firm that works from the first point of contact to ensure that you do not accidentally cut off the wrong limb. A full-service research firm is involved in every aspect of the project's conceptualization and implementation.  
12 Publish Research Results at Scientific Conferences        
    SurveyUSA's scientific process is documented in presentations made at conferences of research professionals, where SurveyUSA discloses the results of its internal testing, the results of its side-by-side testing, and provides a window into how it weights its data. SurveyUSA is an authority on measuring the accuracy of election polls and has published one of only eight existing methods for measuring the accuracy of election polls.  
13 Understand 'Total Survey Error'        
    The key to producing the most accurate research results is to avoid obsessing over any one part of the research process at the expense of others, and concentrate instead on eliminating possible error from each and every part of the research process with equal zeal, resulting in a lower "Total Survey Error." Lower error is better, zero error is perfect. Ask your pollster to talk you through the different ways he/she works to reduce Total Survey Error.  
14 Are Innovators, But Not Risk Takers        
    There is a fine line between research innovation and indefensible risk taking. SurveyUSA tries always to strike the right balance between having excellent data hygiene while not being needlessly wedded to the past. At the same time, we observe in some of our competitors risk taking and sloppy data hygiene that compromises the entire methodology and calls into question the results.  
And at a lower level   Survey
The best Opinion Research Firms
15 Are not charging you for their failure to modernize        
    Many SurveyUSA competitors use elevator operators to run their elevators. They are using 1950s technology in 2012. You pay a fortune for their failure to modernize. SurveyUSA automation allows respondents to push their own elevator buttons. This makes our research more affordable. We are proud of our innovation. We are embarrassed by our competitors who defend their elevator operators as "best practice." Best practice it may have been 40 years ago, not today.  
16 Have written their own software        
    SurveyUSA's proprietary SurveyQSA software was written by SurveyUSA employees for exclusive use by SurveyUSA. No research company has a more powerful platform on which to conduct research. Quality is assured from the first to the last step in the process. SurveyUSA maintains total in-house control over your project. Our end-to-end process is owned by SurveyUSA, not licensed, not leased, not rented. Ask your pollster if the same is true.  
17 Don't Outsource Your Phone Calls to a Call Center        
    SurveyUSA phone calls are made on proprietary software written by SurveyUSA programmers, resident on SurveyUSA hardware, located in SurveyUSA call centers. SurveyUSA maintains quality control from one end of the process to the other. Ask your pollster where his/her phone calls originate. Ask what software program they run on? Ask who wrote the software?  
18 Have Breadth and Depth        
    Some pollsters work in just part of the country, and some pollsters work on only one type of project. Few can match SurveyUSA's ability to work in all 50 states, on issues of national, regional, and local significance. SurveyUSA conducts public opinion polls that you may see the results of on TV or in the newspaper. SurveyUSA conducts private market research that you do not see the results of.  
19 Can Handle Complicated Skip Patterns        
    Many SurveyUSA competitors are limited: within a questionnaire, they can't screen, they can't branch, they can only allow every respondent to hear the same question. SurveyUSA's proprietary software was custom engineered to allow the most demaning research requirements. There is no challenge that SurveyUSA's software can't rise to meet. Infinite flexibility means you don't have to compromise your research design.  
20 Can Split Sample When Necessary        
    Sometimes subtle differences in question wording make all the difference in survey results. To study the effect that a particular wording may have on respondents, SurveyUSA routinely "splits" its sample into 2 or more random groups, and then randomly assigns one question wording or another to a particular group. This allows us to better understand exactly what respondents are reacting to. Ask your pollster if they will split the sample for you.  
21 Can Rotate Questions        
    Certain questions early in a questionnaire can influence the answers to questions asked later in a questionnaire. The best researchers rotate the order in which respondents hear questions, to eliminate any chance of question-order bias. SurveyUSA can rotate your questions to ensure that bias is not baked-in.  
22 Can Rotate Answer Choices        
    If you always read a list of names in the same order, you may end up with order bias. The first name you read and the last name you read will be over-counted. To fix this, good researchers rotate the names that appear in a list. Many SurveyUSA competitors do not, because they are using cheap, immitation software that does not permit question rotation or answer-choice rotation. SurveyUSA's proprietary software pays dividends each time you hire us to conduct research for you.  
23 Can Re-try Respondents Who Do Not Answer the First Time        
    Some SurveyUSA competitors only conduct research on a single night. They do not have software capable of carrying forward onto subsequent nights phone numbers of those respondents who are not home on Day #1. SurveyUSA proprietary software allows SurveyUSA to program, on a job-specific basis, the exact re-try parameters for busy signals, no answers, answering machines, and phone-network problems. This helps ensure your research is more accurate.  
24 Can Select a Respondent From Within the Household        
    SurveyUSA has experimented with every possible method of selecting a respondent from within the household. While sometimes this makes no difference to a research outcome, on telephone projects, SurveyUSA now consistently requests to speak with the youngest male adult who is home at the time of our first attempt. This helps ensure that our unweighted data more closely resembles our weighted data.  
25 Can be Breathtakingly Nimble When Required        
    Some SurveyUSA research projects are completed overnight, when conditions demand it. If you are photographing sports, you use a fast shutter speed to freeze the action; a slow shutter speed produces blur. Some SurveyUSA competitors cannot go fast; they have too much practice going slow. You cannot retrofit a cumbersome research company into a nimble one. But when you need responsiveness, SurveyUSA stacks itself up against anyone in the business.  
26 Can Be Methodical When Required        
    Sometimes, SurveyUSA takes weeks to gather data, when systematic thoroughness is more important than speed. Routinely, when SurveyUSA conducts research for state governments, and for academic institutions, SurveyUSA is "in the field" for weeks at a time. This allows us to make sure we have contacted both the easy-to-reach respondents, who are always at home, and the hard-to-reach respondents, who are seldom at home.  
27 Regularly Conduct Research for Top Universities        
    SurveyUSA is proud that a number of leading universities turn to SurveyUSA for data collection. SurveyUSA has worked for more than a dozen major universities, including some of the nation's most prestigious, on complicated research projects in English and in other languages.  
28 Are Published in Scientific Journals        
    SurveyUSA is proud of the work it conducts for universities and leading academics across the country. Our research appears in peer-reviewed scientific journals, and is used to vet the research conducted by other research companies. Ask your pollster to list the authors and list the scientific journals that have cited their work.  
29 Don't Use Typewriters to Show You Your Results        
    Look at the data released by firms that compete with SurveyUSA and you'll think it's 1957: black typewriter text on white paper. These firms cannot do better, because they have not written their own report-generation software. They are licensing an off-the-shelf solution. By contrast, SurveyUSA has written from scratch the software it uses to display your results; for 20 years, our results have been easier to digest; they still are.  
30 Display Time-Series Data Using Interactive Tracking Graphs        
    SurveyUSA is the only opinion pollster that routinely releases interactive tracking graphs whenever the same question has been asked on different occasions. Interactive tracking graphs, a SurveyUSA exclusive, allow you to study how each separate demographic group is changing over time. SurveyUSA has $1 million invested in this software.  
31 Let You Compare Home and Cell Answers in Real-Time        
    At this moment in the evolution of opinion research, it's not just enough to make a special effort to reach out and interview those unreachable on a home telephone, but it's equally important to contrast the results of the home phone and cell phone respondents. SurveyUSA is one of 2 firms who do this. Ask your pollster why he/she does not.  
32 Can Speak to Respondents in Any Language        
    When appropriate, SurveyUSA can ask questions in Spanish, French, or any other language your respondents may speak. A majority of our projects are, of course, conducted in English. But there are parts of the United States where SurveyUSA will decline to poll unless the client is willing to pay for bilingual research. Bilingual research ensures that everyone is represented fairly in a survey, not just those in the majority.  
33 Can Let You Listen to Respondents Speak, in Their Own Voices        
    Can your pollster allow you to listen to the actual voice of your respondents? SurveyUSA can. Our real-time focus group capability allows us to gather qualitative feedback instead of, or in addition to, quantitative feedback, on any study. SurveyUSA returns to you .wav files that are indexed to a particular respondent. You will know that the gender, age and other characteristics of the person offering you such high praise, or such damning criticism.  

SurveyUSA, 1-800-786-8000,