NSON® Opinion Strategy has conducted thousands of political polls of every type since 1992. We have completed over 1,000,000 individual voter interviews in that time making us one of the most experienced polling firms in the United States. NSON has helped campaigns at the local, statewide, Congressional, Senatorial, Gubernatorial and Presidential level.  NSON conducts all telephone surveys in our own 150-station phone center. Having our own call center allows us to directly manage our projects and keep our prices competitive.

Our services include questionnaire design, sampling strategy, scheduling advice and issue development. We can create and manage every kind of political poll including Benchmark Surveys, Mini-Benchmarks, Brushfire Polls, Rolling Tracking Polls and Short Tracking Polls.

Polling Analysis

Polls can be very complicated.  They must be designed correctly in order to work properly. The purpose of a political poll is to gather information from a small group of voters and use that information to make predictions about how a larger group of voters will act in the future.

The accuracy of every political poll depends upon the wording and length of the questionnaire (the survey) and the diligence and precision with which the polling data is analyzed. NSON analysts have been conducting political research since we started NSON in 1992. When we make predictions about the outcomes of political campaigns, we are usually correct.

Voter ID, GOTV with Live Interviewers

We have many years of practical experience handling live voter contact projects of every size including huge, multi-million call GOTV campaigns. We can provide “live” interviewers with neutral accents to make GOTV and Voter ID calls from the NSON call center in Salt Lake City.

A Brief Review and Glossary of Polling, Voter Contact Services and Political Terms

NSON is a nationally recognized leader at voter contact services including political polling, Voter ID, GOTV, and pre-recorded message delivery. We have extensive practical experience conducting political surveys at all campaign levels including National, Congressional, Gubernatorial, statewide and local elections.

Political polling is a fundamental tool of modern electoral campaigns. It is a complicated and multi-faceted business. The following information is meant to explain the basic fundamentals of polling and other voter contact services and to define the normal terminology pollsters use. We hope that you find this information helpful

Fundamentals of Polling

Political Polling is a process where a defined group of potential voters are asked questions about an up-coming election. The answers given are then used to make predictions about how a larger group of voters will act in the future. The process has real value because the voters in the small group represent a cross-section of the voters in the larger group. The accuracy of the poll is influenced by the design and wording of the survey questionnaire, the accuracy of the data collection method used, and the number and random distribution of the voters polled.

Polling Methods

There are many ways to conduct political polls: In person, by mail, on the Internet or by telephone. Of these, telephone interviewing is the preferred method of contacting a sufficient number of qualified voters in a short period of time to produce data that is both statistically accurate and affordable.

Closed Questions

Pollsters learn what voters think by asking them either closed or open-ended questions. Closed questions give voters a specific list of choices from which to select an answer. A simple closed question might ask, for example, if a voter prefers candidate A or candidate B. Such a question, to a pollster, has five possible answers: A, B, undecided, neither or refused.

Open-Ended Questions

Open-ended questions ask voters to give an opinion about a specific topic. For example, an open-ended question might ask, “why do you prefer candidate A?” Open-ended questions have no predetermined answers. They take longer to complete over the telephone, they cost more, and the responses are more difficult to tabulate once the survey is finished. Pollsters classify open-ended responses into groups using a technique called “coding”.

Glossary of Standard Political Terms

Advocacy or Persuasion Calls

Persuasion calls deliver positive messages to voters. The message is often pre-recorded by the candidate or a famous person who is supporting the candidate. Often, the messages are delivered to answering machines. These calls are often directed to undecided or swing voters late in a campaign.

Automated Telephone Calls

Modern political campaigns rely heavily on the use of automated telephone calls to reach thousands or even millions of voters during a campaign. These calls are sometimes known as IVB (Interactive Voice Broadcast) or IVR (Interactive Voice Response).

Ballot Test

Ballot tests are polls that compare different candidates running for a given office and are often called “head to head” tests.


A banner is a complicated set of cross tabulations used to help pollsters and campaign consultants understand how voters responded to particular survey.

Benchmark Survey

Benchmark surveys typically take place early in a campaign. These surveys can have twenty-five to fifty or more questions and they often include several open-ended questions. The survey involves such things as the importance of specific issues, hypothetical ballot tests, demographic matters, name recognition, an opponent’s weaknesses or advertising recall. The data that is collected is used to develop strategies as the campaign progresses.

Brushfire Survey

A brushfire survey is similar to a benchmark only it is a smaller study. These surveys can be conducted any time during a campaign, but frequently take place after a benchmark has been completed. They usually have some open-ended questions.

Call Center

Most political polling firms house all of their telephones and computers in a central location known as a call center. The NSØN® call center has 150 CATI stations.


This is an acronym that means “Computer Aided Telephone Interviewing.” Most polling companies rely on computers to facilitate their political research.


A completed survey occurs when an interviewer reaches a voter on the telephone and the voter cooperates by answering all of the questions in the survey. The total cost of a survey is partially calculated by multiplying the number of completes required by the amount charged per completed interview. One problem that is common to all telephone surveys is known as a “mid-termination.” This occurs when a respondent hangs up after finishing a portion of the survey. This invalidates the entire survey.

Cooperation Rate

Over the years, market research firms have determined how likely people who reside in specified geographical areas are to cooperate with telephone surveyors. The rate is designated as a percentage. The cooperation rate is one of the factors used in pricing polls.

Cross Tabulation

Cross tabulation means the frequencies (or answers) to one question are compared to the frequencies of another question. In a general political survey, the demographic questions (gender, age, party, geography) are crossed against the other questions in the survey.


This is a term to describe the answers to a question. For example, if 100 people are asked if they will vote for A or B or C, 25 might answer A, 40 answer B, 20 answer C, 10 are undecided and 5 refuse to answer. Those are the frequencies of that particular question.


GOTV stands for Get Out The Vote. GOTV calls are generally twenty to thirty seconds long. They are not polls since no questions are asked. Positive messages are delivered by a live caller or by a recording of the candidate’s voice. GOTV calls are often made just before Election Day or before another important event in the campaign.

Incidence Rate

The ratio between the number of people contacted and the number who actually qualify to take a survey is called the incidence rate. This rate is one of the main factors used in calculating how much a poll will cost.


Interviewers are the people who do the calling, ask questions, leave messages and collect data. All NSØN® interviewers receive continual training and daily monitoring to improve and refine their telephone skills.


Quality control is an essential part of political polling. Companies gauge the quality, accuracy and reliability of their interviewers by monitoring them during actual calls. We monitor every interviewer daily and we make it possible for our clients to monitor our staff when they believe it is necessary.

Predictive Dialer

Modern political polling requires sophisticated equipment in order to complete large scale calling projects. Many GOTV and advocacy calling projects require a predictive dialer. These ultra-efficient machines dial numbers and direct calls to available interviewers. This equipment makes it possible to leave recorded messages in the candidate’s own voice and to record the answers voters make on open-ended questions.


Determining the cost of a polling project involves analyzing many factors including the number of completes required, the computer programming necessary, the length of the script, the incidence rate, the cooperation rate of the voters being called, the amount of cross tabulation involved and the number of open-ended questions in the poll.


Political polling revolves around computers. Every poll has to be carefully programmed into our computers so that our interviewers have a precise script to follow and a reliable method of tabulating answers to questions.


Polling companies utilize scripts or questionnaires for their interviewers to follow during a survey call. Most political clients provide us with their own polling scripts, but we offer questionnaire development services to those who require it.

Random Digit Dialing

Many political polls are conducted using telephone numbers that are selected at random by the predictive dialer. RDD samples reach a very broad spectrum of the populace.


Sample is a term used to designate the group of people who are going to be called to respond to a poll. Clients often provide their own sample of registered voters obtained from state party lists. Lists of names can also by purchased from brokers. The sample can by targeted to reach the desired cross section of the population at large.


This word means “data analysis.” When a survey is conducted, all the data that is collected needs to be organized (or tabulated) into a standardized format that makes the information easy to understand and interpret.

Tracking Polls

Tracking polls are short surveys that are repeated using the same questions (but calling different voters) to gain insight into how particular issues or trends are moving within a campaign. Tracking polls are often sold as packages of four or more surveys that are repeated over the course of several days or weeks.

Voter ID

Voter ID means voter “identification.” Voter ID calls normally have from one to eight questions. They often do not have open-ended questions. The calls are made to a listed sample and sometimes ask for a specific person within a household or maybe for any registered voter. Normally no messages are left on voter ID calls. The information gathered from these calls is used to generate follow-up direct mail ads or GOTV calls.