A standard 500 response poll is $1000, to include recording the questions, buying the phone sample, performing the poll and producing basic analysis with cross-tabs. The entire poll should be four minutes or less, as the hang-up rate increases beyond that point. Further questions may not reflect the population accurately. Questions should be brief and to the point for best results. Long, wordy questions tend to increase the hang-up rate. Short questions keep the poll interactive and maximize response rate.

IVR Polls uses RBS (Registration Based Sampling) rather than RDD (Random Digit Dial). While there are arguments both for and against each method, each can produce accurate results. Voter history can be evaluated using RBS, producing a more accurate 'likely voter' model. On the downside, voters who don't list a phone number on their registration may not be included in the sample. RDD can generate these unlisted numbers, but it can also produce numbers for precincts outside the district if the polling area is less than statewide.

Beyond the quality and accuracy of sample types, RBS has the added benefit of demographic data, allowing a greater portion of the four minute poll to be devoted to the voter's opinion rather than the voter's description. IVR Polls strives to offer their clients a high yield of information at an affordable cost, and RBS achieves this goal better than RDD. Custom polls using RDD can be requested, but fees will be higher.

IVR Polls purchases their phone sample from a third party. If campaigns or elected officials have access to a current vote file, the sample can be drawn from the client's vote file, reducing charges from $1000 to $850. IVR Polls must be given access to the entire vote file to insure that the sample is representative of the population.

Volume discounts are available. Prepaid blocks of six different polls will discounted $100 per poll. Prepaid blocks of twelve different polls will be discounted $200 per poll. Tracking polls that do not change month to month will be discounted an additional $50.

Polls with a larger 1000 response sample size are priced at a 50% premium. This is recommended if your poll forks based on the answer to an early question. For example, you would ask one series of questions if the voter indicated that they voted Republican and a different series if they indicated they voted Democratic. The larger sample size would yield a reasonable margin of error for both series.