>
According to the study conducted by the Pew Internet & American Life Institute, "robo-calls" are the second most popular method of contacting the voter, trailing only direct mail as a key tool of political communication.

Political Marketing International, Inc. has been doing automated political calls for more than a decade and has represented more than 750 political candidates for polling and over 2500 candidates utilizing various other applications of automated calls. PMI works directly with consultants at all levels of the political spectrum including candidates for the White House to "down ballot" candidates, school committee and county races, in addition to statewide ballot initiatives.

Automated polling is inexpensive, extensive and, above all, accurate. This polling style reaches and reflects the voice of the voter in a manner that person-to-person or volunteer calls cannot match. One of the benefits to automated polling is that we call every household on the voters list.

The majority of our automated polling questions range from 30-90 seconds in length, although we have conducted some polls lasting 4-5 minutes. Responses usually vary from 10-40% depending on the issue(s) and/or candidate(s) and demographics involved.

Please keep in mind that it has been our experience that in most cases the person polling with the highest number seems to attract 50% of the undecided vote when we give undecided as an option when polling. Only after consulting with our client do we decide on including undecided as an option or just listing the candidates in a given race. We also suggest that candidates be listed in ballot order.

PMI, Inc., gives our clients a 4% + or - (MOE) margin of error with at least 5,000 households polled.

As you can see the response rate will vary. This is due in part to many factors including the accuracy of data, day and time of day when we call, the length of the survey as well as how many live people we actually connect with when we conduct a poll. Out of 5,000 records called we may only connect with 1,500 live participants and 200 actually participate in the survey. You must take that into consideration when calculating your response rate, i.e., 200 participants divided by 1,500 live connects gives you over a 13% response rate.

Automated polls can be just as accurate as a live poll and much more cost effective for a candidate running on a limited budget. Several case studies follow in which PMI was involved that show the precision of our automated polling.



State of Florida, House District 5, General Election
Date of poll: October 27, 2008.
Candidates: Brad Drake and Johnny "Mac" MacDaniel.
*Polling Cost: $1750.00
The total number of households called: This was a district-wide survey of 28,367 likely voters. 3,748 voters participated in the survey which is a 9.8% representative opinion of District 5 voters.

The percentages of our polling and election-day vote were:
  • Drake: 59.5% - election-day vote: 59.9%
  • MacDaniel: 40.5% - election-day vote: 40.1%
*This case study did not involve either candidate as a client and was a case study only.



Jackson County Commission, District 5, Republican Primary
Date of poll: August 7-8, 2008.
Candidates: Ben Odom and Jerry Neel.
*Polling Cost: $500.00
The total number of households called: approximately 1,000 by party affiliation. Over 60 voters participated in the survey. In this race we offered the option of Undecided.

The percentages of our polling and election-day vote were:
  • Neel: 22% - election-day vote: 45.45%
  • Odom: 30% - election-day vote: 54.65%
  • Undecided: 48%
By including an undecided in this poll and based on past polling experience, we projected that the top vote getter would attract at least 50% of the undecided vote. If you take our figure of 30% and give Mr. Odom 50% of the undecided vote (24%) that would give him a projected number of 54%.
*This case study did not involve either candidate as a client and was a case study only.

As a follow up to the above mentioned race we conducted a poll among the two Democratic candidates in the same commission district.



Jackson County Commission, District 5, Democratic Primary
Date of poll: August 18-19, 2008.
*Polling Cost: $500.00
Candidates: J. Milton Pittman and Kenneth Stephens.
The total number of households called: approximately 2,500 by party affiliation. 250 voters participated in the survey. We did not offer an Undecided option in this race.
*This case study did not involve either candidate as a client and was a case study only.
The percentages of our polling and election-day vote were:
  • Pittman: 47.2% - election-day vote: 47.65%
  • Stephens: 52.8% - election-day vote: 52.35%


Jackson County Sheriff, Democratic Primary
Date of poll: May-August, 2008.
*Polling Cost: $750.00
Candidates: A crowded field of candidates including a retired Chief of Police, a Major in the Sheriff's department, a State Attorney's office employee as well as those active in reserve in law enforcement.
Total number of households called: 10,000. Over 2,250 voters participated in the survey. In this race we offered the option of Undecided.
The percentages of our polling and election-day vote were:
  • Charles (Chucky) Anderson: 8% - Actual vote: 8.47%
  • John Dennis: 7% - election-day vote: 11.84%
  • Lou Roberts: 31% - election-day vote: 45.96
  • Robbie Western: 9% - election-day vote: 15.52%
  • Darryl Williams: 12% - election-day vote: 16.15%
  • Undecided: 33%
If you take our figure of 31% and give Mr. Roberts 50% of the undecided vote (16.5%) that would give him a projected number of 47.5% which is well within our 4% + - MOE.

Due to the unusually high undecided factor and the threat of rain on election-day, PMI suggested that Lou Roberts send out an automated GOTV message "reminding voters that tomorrow is election-day and that he would appreciate their vote."

We also suggested that Mr. Roberts record his message that we delivered to answering machines only, i.e., "hi this is Lou Roberts and I am sorry that I missed you today" and a separate message was recorded from a 12 year old child saying "hi, my name is Brandon and I'm not old enough to vote but if I could vote I would vote for Mr. Lou Roberts in tomorrow's election".

*This case study did not involve any candidates as a client and was a case study only.



Jackson County Sheriff, General Election
Date of poll: October 20, 2008.
*Polling Cost: $1250.00
Candidates: Jim Peacock, Lou Roberts, William Nelson and Zannie Williams. Total number of households called: 13,000. Over 850 participated in the survey which is a representative opinion of 7% of registered voters.
The percentages of our polling and election-day vote were:
  • Peacock: 33% - election-day vote: 36.37%
  • Roberts: 57% - election-day vote: 57.86%
  • Nelson: 5% - election-day vote: 2.16%
  • Williams: 5% - election-day vote: 3.61%
*This case study did not involve any candidates as a client and was a case study only.



Mojave County, Arizona, Superior Court Judge Division 5
Date of poll: August 18, 2008
Candidates: John Taylor, Dana Hlavac, Rick Williams.
*Polling cost: $500.00
The total number of households called: 2,500. 250 voters participated in the survey.
The percentages of our polling and election-day vote were:
  • Taylor: 41% - election-day vote: 35.30%
  • Hlavac: 25% - election-day vote: 28.10%
  • Williams: 34% - election-day vote: 36.34%
*This case study did not involve either candidate as a client and was a case study only.



Mojave County, Arizona, County Assessor
Date of poll: August 18, 2008
Candidates: Ron Nicholson and David Smallwood.
*Polling cost: $500.00
The total number of households called: 2,500. 200 voters participated in the survey.
The percentages of our polling and election-day vote were:
  • Nicholson: 70% - election-day vote: 65%
  • Smallwood: 30% - election-day vote: 35%
*This case study did not involve either candidate as a client and was a case study only.



Mojave County, Arizona, County Recorder
Date of poll: August 18, 2008
Candidates: Kim Stewart and Carol Meier.
*Polling cost: $500.00
The total number of households called: 2,500. 277 voters participated in the survey.
The percentages of our polling and election-day vote were:
  • Stewart: 49.5% - election-day vote: 48%
  • Meier: 50.5% - election-day vote: 52%
*This case study did not involve either candidate as a client and was a case study only.



Mojave County, Arizona, County Treasurer
Date of poll: August 18, 2008
Candidates: Lee Fabrizio and Melissa Havatone.
*Polling cost: $500.00
The total number of households called: 2,500. 360 voters participated in the survey.
The percentages of our polling and election-day vote were:
  • Fabrizio: 50.5% - election-day vote: 46%
  • Havatone: 49.5% - election-day vote: 54%
*This case study did not involve either candidate as a client and was a case study only.



Jackson County Superintendent of Schools, General Election
Date of poll: October 20, 2008
Candidates: Steve Benton and Lou Miller.
*Polling cost: $1250.00
The total number of households called: 10,000. Over 700 participated in the survey which is a 10% representative opinion of registered voters.
The percentages of our polling and election-day vote were:
  • Benton: 48% - election-day vote: 47.16%
  • Miller: 52% - election-day vote: 52.84%
*This case study did not involve either candidate as a client and was a case study only.



Leon County, Florida, Circuit Judge Group 3
Date of poll: August 17-18, 2008.
*Polling Cost: $2750.00
Candidates: David Frank and Kevin Davey.
Total number of households called: 50,000 Republicans, Democrats and Independents. Over 2,800 voters participated in the survey. In this race we did not offer an Undecided option.

The percentages of our polling and election-day vote were:
  • Frank: 47% - election-day vote: 49.2%
  • Davey: 52% - election-day vote: 50.8%
*This case study did not involve either candidate as a client and was a case study only.



Congressional District 1, Democratic Primary, State of Maine
Date of Poll: May 5-6, 2008
Polling Cost: $2500.00
Total number of households called: 30,000 Democrats. Over 1,800 voters participated in the survey. In this race we offered the option of Undecided.
A survey was conducted during a two day period of May 5 and May 6, asking voters how they intended to vote on June 10, 2008 in the Democratic primary for congress.

The percentages of our polling and election-day vote were:
  • Brennan-11.66% - election-day vote: 10.83%
  • Cote-6.42% - election-day vote: 28.63%
  • Lawrence-5.2% - election-day vote: 4.55%
  • Meister-.0125% - election-day vote: 1.17%
  • Pingree-32.38% - election-day vote: 44.19%
  • Strimling-8.32% - election-day vote: 10.63%
  • Undecided-34.73%
If you gave Pingree 50% of the undecided vote (17.36% and added that to his 32.38% polling figure) that would put him at 49.74% or well within our 4% + - MOE.



City of Tulsa, Oklahoma
Date of Poll: March 5-6, 2008
Candidate: Eric Gomez and Maria Barnes
Polling cost: $750.00
The total number of households called: 8,000. 700 voters participated in the survey.

This survey is unique in the sense that we not only polled what candidate voters supported but what issues were most important to them. The survey results revealed that out of 700 voters surveyed, over 27% supported the creation of conservation districts, compared with around 13% who opposed it and an overwhelming 60% having no opinion at all. The incumbent spearheaded the conservation district concept and we speculated that the incumbent might do something at the end of the campaign that we would not have time to answer. The survey also conveyed over 61% of the voters wanted the city to devote all available resources to repair and rebuild the streets. Surprisingly over 26% said no and 13% had no opinion.

This poll also gave helpful information about how Mr. Gomez was doing. The incumbent was at 20.8%, Mr. Gomez was at 18.9% and the undecided were at 60.3%.

With this information, PMI began cross referencing which voter supported Mr. Gomez, who also supported or opposed the conservation districts, as well as the repair and rebuilding of the streets and put a strategy in place for GOTV efforts. There is no wisdom in calling a person asking them to vote for you if you already know they are supporting your opponent!

Mr. Gomez won by 152 votes, (53% to 47%).



Mayor of Shawnee, Oklahoma, General Election
Date of poll: October 14, 2008.
Polling Cost: $1250.00
Candidates: Linda Peterson and Linda Agee.
Both candidates shared the cost of the survey. The race for Mayor was the first question and then we asked voters who they preferred for Commissioner.
Total number of households called: 10,309, city wide survey. 1400 participated in the survey which is a 14% representative opinion of registered voters.

The percentages of our polling and election-day vote were:
In the race for Mayor…
  • Linda Peterson: 50% - election-day vote: 54%
  • Linda Agee: 50% - election-day vote: 46%
In the race for Ward 2 Commissioner…
  • Frank Sims: 54% - election-day vote: 54.6%
  • Jerry Chastain: 46% - election-day vote: 45.6%


PMI recommends at least 5,000 records be called in a small race and at least 10,000 records in a larger race, however we conduct polls with as few as 2,500 records. The more records called, the more responses obtained.

Data is vital in our polling. We advise clients to provide the most current up to date data available and then we recommend having that data tele-matched. PMI's experience indicates that we get anywhere from 12-28% tele-matched with new numbers. If a client is not going to provide data, we can purchase data for an additional cost to the campaign.

In a close race the data you use can determine the outcome. Surprisingly, some candidates use older data in order to save a few dollars.

If we could tell you the outcome of your election before election-day would that information be useful to you? You could plan and execute last minute mail and crucial GOTV efforts more effectively.

Automated polling works. It's not only quick but its proficient and cost-effective. Please contact us for more information.WhereIvote

Jerry Dorchuck, 215-370-5509 cell - Office: 850-482-5079
© 2008 PMI all rights reserved
Website design and hosting by Web Showcase