this story and using it as a personal example, the difference 2010 the obama and kerry campaign is there was a story on the obama campaign. there was a good reason that was given for why you would want to vote for obama and there was talk about these are the good guys. your candidate is the hero. you want them to be the good guy. they should be in office. and it is one of those things where the overall frame of the campaign is hope and change. four years earlier you had john kerry and i'm from massachusetts, love the man as a senator, forth a very successful campaign. where you can't answer the question what is the campaign standing for? where i could go around the room and get 15 different answers and there was even confusion sort of in the campaign of what does

this actually mean. so, it kind of came down there was not that clear a story line. the last thing i want to talk about is how to talk about policies an issues. i want to be clear on this. you shouldn't be polling to figure out what your policy s. you should know what your policy s. your candidate should know what issues they think are important. you should have a good handle on with the big issues are in your state and what is going to come up. right now education, economy, jobs are important things. fending on what state. there will be local issues. it is not rocket science to figure out what the important issues are. but what it really is is there are always six different ways to talk about an issue or a policy proposal and trying to figure out the best way to talk about it is ultimately going to be the difference between a successful or unsuccessful campaign.

there are two examples that i like to talk about. first is the, what started as the estate tax. it was the mid 1990's where the republicans came out with a bill it was a against the estate tax and it failed. you talk to people with to you care if they are taxed. then it changed to the inheritance tax. it did a little barrett. you heard people i may get into that. i don't want that tax. then they did more research and death tax. everyone is against. i will die some day and i sure as hell don't want my money being taxed. the policy didn't change. it was just how they talked about it, what they called it. that was the only thing that changed. that is what research did for them in that case is figuring

out the best way to talk in the way that would get the most people behind them. the other example, out of the political sphere but a great example is tivo. when it first came out they marketed it as on the premise of the technology that was behind tivo where all the ads was about how cool it was that they came one a way to record commercials and figure out how you -- the types of shows you want to record and figure out shows hau might also like and it failed. it was flat. no one cared. made no sense to anyone. they regrouped, did bunch of research and came out with a new advertising came that was based on this is how this is going to change your life. you never need to sit down at a set time. this box is going to record all the shows you want to watch.

you can watch tv on your time. and now i don't know how i lived without my tivo before. it was a very dark period in my life. but like nothing changed. the technology didn't change, the product didn't change. it was just how they talked about it. this is just something to remember. you don't want to be polling on what your policy is. you want to be polling on how to talk about your policy. that is the nutshell of what polling s. the other piece of research that us pollsteres do are focus groups. we basically go around the country, pull together 10, 20 people, sometimes less, sometimes more and talk about bakley what a the who of times what is the thing to put in the

poll. what focus groups do is in polls sort of tell you what focus groups are, how you figure out why. we know that you find out that 40% of people in your state support the healthcare package that will go through congress. that is interesting in and of itself but you don't have the why. what you can do with focus groups is figure out why. it usually lasts about two hours. with focus groups if you want to talk to voters who vote every third year and have bluies you can do that. you can get very specific with what you are doing in your groups. so, you tend to have tpwrups just maybe his educated women, or you have focus groups of

parents with kids in public schools, or voters over 65. you can do whatever you want in the focus groups. and you really sort of get the depth of the understanding of the issues. because at the end of the day a lot of in polling tends to be bina binary where if is a yes or no and very black and white picture of things. and most people's views on issues are not really black and white. they are very huge shades of gray and focus groups allow you to get in the gray and figure out why they do things they do and how you can best connect with those people. these are things that you do not

want to do without because this will give you a depth of understanding that you really need and want. this also is a great way to test media. there is no point in putting out ads that you have no idea if they resonate with performance. you would be amazed what a campaign with overlook in an ad. what somebody in a focus group will say you really can't say that in this state because -- i'm trying to think of a good political example. but the one that documents mind, this isn't political but the chevy nova. this going back a few years. they have the chevy nova. wasn't selling well in mexico so they put together a couple of focus groups that nova means don't go. not good for cars. we put up an ad that we thought was great and you just started hearing from people where they said either they picked out like that video wasn't shot in this district or that is actually not

our state in the background, that is somewhere else and things that people didn't pick up or it happens. it is good to have the average voter because they have a different eye than you will as a campaign and will save you a lot of heartache. and it is like in the corporate world no one is spending $10 million on an ad campaign without testing in thoroughly to make sure will are no mistakes that people just overlooked. the other big thing is candidate buy bios. i find in a waste of time to read a bio of somebody over the phone. it is very enlightening to see what it is like on the phone.

but if you read a one-page bio and ask them what they thought they possibly stopped listening and they will give you an answer in the middle unless they are very attentive. but that doesn't happen too often. so if you read somebody's b bio and ask if you have a positive or negative view, even if they say very positive you don't know what was in it that really resonated. but you read it to people and say read this and then you can talk about it and this is where you get into things that maybe the person had a military background, that is drawing people in because they had a military connection. or work they have done in the past, that is where you really learn about what is the best way to couch or present your candidate to the voters.

it is tough to pick that up in the poll but in the focus group that is where you can really do a great thing. >> do you always have time to do a focus group? >> typically you do the baseline and follow up with focus groups. there are times where you might do that differently. as a rule of thumb if you are dealing with two very well known candidates, then you could do focus groups first and really start getting a sense of what palmeiro think about the candidates. the other time i have done statewide ballot initiatives where you are stuck with a set paragraph of how this is described and how it will go on the ballot and again this is a tough thing to test on phone so focus groups may be more

enlightening because you have something concrete. but in general you will probably do the baseline and follow with focus groups. >> are focus groups more expensive? >> the one thing that you can't do with focus groups is can't learn something and apply it to the audience at large. you can't do a group where you sit down with 10 people and say ok, well, we talked to 10 women and all of the women think what these women do. bad idea. it can't be applied sort of to the bigger picture. the reason why they tend to be a little more expensive, it is depe depending on how you look at it. most of the expense is the facility that you are holding the groups% they need to make sure they get the right people and things like that. and so if you are on a lower

budget campaign, if you have to class between the two you are probably going to go with the poll than just doing focus groups. so if you had $15,000 to $20,000 to spend on research you should probably do a poll because you want numbers that you can look at the buyer universe. >> potential negatives are a good thing to test in particular groups because you can throw a lot of things out there on your own candidate, maybe on somebody you are running against. but it is a god controlled way to release information that you might not want to get out in the public sphere potentially. on the phone people don't enjoy hearing bad things. even if they are true a lot of times they don't believe you or you are just trying to get a

clear read on what this person was arrested for drunk driving six years ago, what dinkind of effect would that have on your support for the person. but a lot of times they can be better test on focus groups and it is tough to test an ad over the phone. maybe one day but we are not there right now. it tends to be done it focus groups the best. any questions? >> would you be able to give us an indication of what you would carjack or like a reasonable rate? how much do these things cost and what would be we responsible for? >> a good campaign at least 80% goes to polling research.

anything less than -- and i don't care what the other people say. they are full of it. no, i mean a good campaign you will probably spend about 10% of your budget on research. a great campaign, you may be up to 15% to 20%. that is a large campaign. more the presidential race. but you are probably in the 10% range. it is you have to pull prices just because it depends on the sample size, how long the poll is and speaking of which day away from long polls. it is -- you are not getting god information if you have somebody on the phone for 20-plus minutes. by my best guess you get people's attention for 10 or 12 minutes of like the heart of the poll. so, not talking to screeners, how do you think the direction of the country.

but when you want good information about 10 to 12 minutes is about their attention span then questions about the demographics, they don't have to think about that. i have done long campaigns, they are horrible and we really need to find out about this and we are going to poll and we should put that in there. at a certain point you are making the poll worse. i hate going over 15 minutes. 18 minutes is the max that i will go. i have put a 25-minute poll in the field. they are excruciatingly painful. and i'm weird, pollster ares are weird but if somebody calls me i take it. i put my watch on and see about when i start tuning out and it is about 10 to 12 minutes. and that is when somebody is interested in how they word the questions, what they are asking. so your afriverage person who ps

up the phone who doesn't have call are i.d. is not really giving you great information after they are bored and all they can think of is get me off the phone. how much longer? what you start to see as you go on is you get po more of those middle answers yes, middle one. if you are on a zero to 10 scale they keep saying five or somewhat satisfied. because they are not really listening and they are trying to do it quickly to get off the phone. so, try not to put the behemoth polls in the field. you are doing yourself a disservice. if you have to do two, you have to do two and sometimes that is a budget hit you may not want to take but if that is the case pick what is most important and call the poll that you have because you are not doing yourself a favor with a long

poll. >> i know our on c-span none of your clients will hold you to this. >> the industry standard for focus groups and these prices are the standard pricing. so, if you need to pick some outrageously hard group to find the price will go up. but you are looking at about $7,500 per focus group and you do two a night so about $15,000 a night for a night of focus groups. polling, again depending on your state, a soiled baseline poll, say you do a poll of 600 you are at the 15 to 18, you are probably somewhere in the mid to upper 20's. you may hit 30, depending on how long the poll is.

but something in that range. if you are in a smaller district and doing like 400, maybe 500 people and 15-minute poll you are probably in the high teens to low 20's. that is about where you are. i would say if you are looking at a baseline of say round numbers say 30,000, then a night of focus groups and you are like 45. midway tracker, let's put on like 18 for that, i just lost my math. where are we, 35? 45. ok. now we are ated a 63. now you have you are getting in the tracker. say for round numbers on a standard statewide i doubt you are getting away with less thn b -- i mean, i think $100,000 is a fairly shoestring budget and you

will probably do more than that. because you are not just going to do one night of focus groups. you could if you don't have the money but the ideal is you won't do, let's say, pennsylvania the state that i just did work in, you are not just going to do one night in philadelphia because you want to hear from voters in pittsburgh and you want to hear not just men and women but you need to talk to black men and black women and some combination of white men, hispanic men. where you are not just doing one night. you are spreading it out. you are talking to different areas. and it depends on how much media you are doing and the entire side of the campaign. and there are a lot of times also again with bigger states that you will do the baseline a couple of months later and

something will happen. some issue will hit the national scene, it was something that no one saw coming, you don't know the answers, you need to go back and figure out what people are thinking about this, what you need to say that you need to say to make sure you are bringing in the swing voters, independent voters and not upset the people that are against you. so these things tend to come up and you want the flexibility to be able to do that. because again the more information you have going into the fight the better you are going to do. and it is the same thing with military, sports, whatever sort of example. the more information you have going in the better you will do. the deeper understanding you have of your side and your opponent's side the more informed decision you can make.

you don't want to be in a position where you are on a conference call and you hear people talking about well, i think or i guess. what you want to be able to say we know that voters thing x. what does that mean for us on this policy or issue. what do we need to do to move forward and make sure we are putting ourselves in the best position. that is why try not to chintz on the research budget because it is what is giving you the building block to a successful campaign and if you don't have that understanding your mail isn't going to be as effective and the media won't be as effective because you just don't have that level of understanding that you should have. what i thought i would do now is talk about anning plan. i work better this way a little

less going through nuts and bolts an giving an example. i came up with a fictitious candidate in new jersey and this is fictitious. mr. smith here is a an independent so just to get people out of their sort of partisan thinking. we have a guy who is a self-made millionaire, made it in the tech world. started a nonprofit that focuses on raising the graduation rate in public high cools in new jersey. white, mid 40's. married to a hispanic woman. put out a good amount of money in but won't completely self-fund and two kids in private school. here is a green candidate. rich candidates. especially if they are willing to spend a lot of money. you colonel in, you are hired. you know he is running. you don't need to prove to him

or yourself that you should run because he has decided to do it. you are 10 months away from your election and you are about to go into your baseline poll. you need to know, you have someone who is probably not well known, hasn't been in the political sphere in the past. you have a different landscape than probably the democratic and republican who are running in 20 2010. a lot of things you will be doing is looking at issues that -- how the issues that will be favorable to your candidate and how they will match up with what people think is important. looking at this guy, business savvy, also the education piece.

these are all going to be things that you will want to flesh out. what do people think about this. and a lot of it won't be saying mr. smith started a education program that tries to get more kid to graduate from public school. do you think that is a positive or negative? not many or going to say that is a negative thing. you are much better taking it back a step and making it sort of a nonspecific to your candidate. but how important is the graduation rate from high school for you in public cools in new jersey? what that will tell you is is this something we should be focusing on or is this just sort like a piece of background puzzle that we are trying to put together, something that is good to bring to the table but not really what the focus of the campaign is. and again with good polling you will get those answers. what should be the front and

centerpiece of what we are doing and what is sort of the background and with a lot of other things. so, we have the issues. favorabilities, you will be testing your candidate, you will be testing the democrats that are running, testing the republicans that are running, you will be testing the new governor of new jersey. you will do the president. maybe you have some surrogates that will come in. you want to try to figure out do people know who the surrogates are. so, it really doesn't do a lot of good if joe smith's cousin says i'm for joe smith. if no one knows who they are. this is where you are going to sort of figure that out. the horse race, and this goes back to what i talked about in the beginning, no one is going to know your guy. you will get like 2% who will

vote for you because you always get 2%. it is very hard to get zero because somebody always thinks they know you and they are going to vote for you. if you say independent especially in the year coming up you may get more votes than you think. so that falls into the category of it is good to figure out where you stand but it is sort of the least important thing that you are doing. in the bio piece, putting up the description and asking people what you think you are not learning anything. so, this is where i would start really breaking this out and start talking about almost piece by piece what do you think about people -- and you will couch this in favorable ways to your own candidate -- self-made man

who was very successful in the tech industry, was that an important thing to bring to the table as a candidate. mall business experience an important thing? is the fact that he started a nonprofit. how does that resonate with people? then you get into, well, married to hispanic woman, is that getting you any traction not just with hispanic voters but is that your white guilty liberals who think marrying somebody who is not white is great, does that bring them on and you start fleshing out how you are going to see the narrative of how you talk about your campaign and your candidate. then what is a self-made person in is that a negative? is that something we need to address further down the line?

the fact that he started a nonprofit that talks about going from graduating from public school but his kids are in private school, how big a deal is that? is that a negative that we need to address? at the end of the day what you will get is a clear picture piece by piece of the candidate and what they think is important and what they bring to the table and you can figure out what are our biggest strengths, what are our biggest weaknesses and what do we need to do to address these things? and the one thing -- other -- another thing is don't worry about your strengths. that will come out. those tend to be things that you need to -- they help paint the picture but it is not something that you need to completely focus on to the point that you are ignoring krr weaknesses. and if you see weaknesses that you know are going to come up, you know some of this will come up. there is no way the fact that

his kids go to private school won't come pickup in some fashion. so don't put your head in the sand. start owning that issue. and the earlier you can get on top of these whether it is policy, issue, something that is good or bad in the bio, if you can beat everyone to the punch on framing the issue you are more likely to be successful in the end. and again, as an example of this, back to 2004 kerry and the swift boat being, they waited forever to talk about that and really take that on. they also invited the whole military piece into the campaign because it was a big focus in the early stage. i did a six-hour class on why that was not the best idea but it is one of those things where you know it is going to come up. do something about it and the earlier you can do about something about it the better.

and with god good -- with good baseline poll you will know how to set the table for this likely attack that come your away and start trying to use it to our advantage as much as possible. so you have the bio pieces, the policy and demographics. the baseline poll tends to ask more demographic questions. so, you have not only do they have kids, education, you might get into does anyone in their household a union member, military member, do at the own a firearm. depending on what state you are in. and there are also some in the policies there are sometimes basics of just getting a handle on what percent of your voter population is pro-choice. what percent is for gun reconviction. what percent are against.

which heplps paint the landscap and gives you a better understanding of where the voters are in your district and your baseline is where you are going to learn this and where you are going to start getting a handle and you might be from the state, you might have worked in the state before but keep in mind that you are one person, even if you worked in a couple of campaigns you worked on those specific campaigns it is always good to actually get more people no think from the people that you are trying to persuade to vote for your candidate as opposed to just thinking you know and going on that assumption. you are much better off knowing what the people are thinking as opposed to assuming what they think. and again if you are on a conference call and a lot of people are saying i think, i think, i think, you are probably

about to make a very bad decision. you are much better off having a conference call saying well, we know that and going from there. i'm sure we have all or some of i was have been on those conference calls. when you do the baseline poll you are starting to get into the flow of the campaign. now you have your midway poll. if you tested an issue two months ago, three months ago, unless something radically has changed in the public sphere, someone's view on abortion is not going to change in two months. someone's view on public education is probably not going to change in a couple of months. if there is something specific, a piece of legislation that has passed or about to pass, a big news story hit your state that you didn't talk about before,

you may want to put that -- that is what you want to put in the midway tracker. but retesting what you already know is wasting time and you are spending more money that you don't want to spend. not being a good business model here but don't reinvent the wheel if you don't have to. so if you have an understanding of where voter are don't retest it because you know what it is. favorabilities? you probably tested a lot of people in the first one because you wanted to get an understanding of a lot players. it is interesting to know where the governor stand and you want to know where your candidate stand so more people know, what do you think of the people you are running against. but don't waste your time if you tested unions or a group library the n.r.a. or plant parenthood. whatever group you tested unless they have been in the news the opinions haven't really changed all that much.

the horse race becomes more interesting. as you get close to the election voters tend to pay more attention. you start getting a clearer picture of where things are heading. i started probably at low point since you are working for a guy that no one has heard before and you want to see positive movement. but don't expect to be winning a huge race and you probably will still have at least 30% of the elect kwraet who says they are undecided at this time. it is interesting to find out who those people are and what they think about the issues. a lot of things you can flesh out are they really undecided or playing hard to get. new jersey is one of those states where voters are notoriously coy about who they are going to vote for. they just say i don't know who i'm going to vote for and up until this past election the numbers didn't change the last month but a lot of times in

presidential elections is new jersey turning red or are republicans going to win and usual the democrat comes back and wins without too much trouble. the past year is a different story but you figure out because a lot of voters get in voting habits of party politics and will vote for the democrat or the republican regardless of who else is on the ballot. but you can figure out who are the true undecideds are and who the independent voters are and a lot of times that is the main focus of your campaign. your bio piece, unless something changed or you have a specific question about something you didn't test in the first one, no reason to be relearning what you figured out in the first one. so no point wasting time going through your bio again to see if anything changed and what palmeiro were thinking about your -- what people were thinking about your bio.

you may want to get into new issues that hit the radar, hone in on one specific issue that you didn't have a chance to do in the first poll. here is a good chance to do it in your midway tracker. then moving into tracker polls, now you are looking at very short polls moving up to the election. at this point, unless something has completely pwraoeupbd sided you -- blind-sided you, what you need want to know is do people know you, how do they think about you and how has that changed as you moved toward the election and where the horse race is. and in some campaigns it is just the horse race. are you going to vote for candidate x or y. i work on some campaigns where we were seeing more movement in

descriptions of candidates, which tends to be more interesting but the horse race wasn't moving but we saw more movement in people saying the candidate i was working for understood the problems of day-to-day people in new jersey. and we started seeing that was positive movement for our candidate. so, sometimes there are other ways of getting at the movement. times it presents itself in the horse race, other times you need to find a different way to get there. but with the tracking poll that is pretty much it. at this point the policy proposals have been made, you have been arguing what they are, you are on tv. you are so the of losing the ability to redefine the race so you are trying to figure out where are my last mailerse to g out. what media markets do i need to be in? do i need to have to worry about a policy in hthis community.

do i need a surrogate to talk for me to get the last bump to get me over the finish line? the other things that you should remember with polling is that you should be polling because it is going to inform a decision. if you have made all of your decisions, if you have spent all of your money, if you have the final media bit in the can, if that is at the tv station, i'm going to put x dollars in this media market and the other in the other, polling is doing nothing for you at that point. again, except spending money and putting a little bit extra money in my pocket which i'm not against. but there is really no reason to poll if you are not going to make a decision based on the polls. so, if you hit the point where you are 100% comfortable or there is just no decision left to be made, don't poll.

>> [inaudible] >> often you will see a situation the baseline was january and we will do a midway in june and we will do tracking in august and we will say why. well, why not? and that is the answer. do a poll to inform a decision. >> you can drive yourself nuts on a campaign, and sometimes too much data is that. too much data. i worked on one campaign where going back and forth and i had recommended to not poll but then some other people wanted to get the poll in. and the one thing that you always have to remember about polls is that we you hear them talk about it at the talk about the margin of error at 95% effective rate is usually the way they talk about it. so if you did the same poll 100 times in the same district or same state, 95 times you will

get pretty much the same answers within whatever that margin of error is. a lot of times it is three or four points but it will be consistent. five times you are going to get ridiculously weird numbers. that is how statistics work. that is how it happens. in this case i'm specifically thinking about we we were deciding whether or not to poll my concern was we had too many numbers to begin with and all of this was going to do is freak out the candidate. then we got a truly bar poll back where we had been seeing like numbers changing by centimeters over the last two months and all of a sudden we did one poll that was done two days after the last poll that all of a sudden our candidate had dropped eight points and the other candidate had dropped five points and everyone was undecided and we were seeing weird things and it was just one of those five polls that happens. the candidate competely flipped out and just was a basket case

for 24 hours because they were convinced they were going to lose even though the last two months they were up by six points, and they won with 55%. it was one of those polls. so, as a campaign manager especially toward the end you know your candidate, you know what is going to make your campaign the most effective campaign and if you have no more questions that need answered, don't muddy the waters with more numbers that can just cause problems. so, get a clear sense -- and if this poll comes out, are we going to make a decision based on the poll? if the answer is no, then you probably shouldn't do the poll. make sense? questions? no. yes.

>> out of curiosity, if you have to have 500 people polled how many do you usually have to call in >> it is getting worse and worse every year. usually the way that the phone banks think about it is not how many -- it is done a little more of how many names you need to give them to get 500. these days you are looking somewhere usually about 30 to 35 for one. so the one call you need you need to give them 35 names off the voter file to get that one person. it used to be 20 to one and as we have gone on with more and more people just having cell phones, no land lines, caller i.d., it is getting harder to get people on the phone. that being said, it is a very odd thing where when you look at studies done on what is truly

stats can i correct research versus what is done for let's say political campaigns which i'm not going to argue is truly pure statistically correct, the difference between the two are so minimal that the amount of money that it would cost to do it perfectly correct is not worth it for the purposes of the campaign. and i know it doesn't seem right, seems odd bush administration it is true. what we have seen on some of the presidentials where you have a chance to look at this, when you introduced cell phone numbers into the sample, the difference really wasn't worth the effort. again, unless were doing a specific poll for, let's say, younger people. if i really want to know what under-5 people and i want to talk to 400 people i'm doing a lot of internet an cell phones. i'm not calling a lot of land

lines because it is not going to happen. and the other reason why you are hiring a pollster is they know the dep graphics of the area and how the calling works and they know how to make sure the calling house is getting a representative sample of the district that you are calling. these are things you don't need to worry about because that is what keeps us up at night we are making sure we are getting the correct spread geographically, demographically. we're getting the right age groups. this is what is the heaviest lift because if you just call people you will get old, white people who day at home a lot and like to talk on the phone. that is what we are going to get and you really need to -- our job is to make sure that doesn't happen. that you are getting a true representative sample of the district that you are working in.

>> you just briefly touched on for the first time internet polling. do you think that is less reliable than actually talking to somebody on the phone and doing the phone poll or do at the even out? >> it depends on what you are doing. like right now i would be 100% comfortable with a nationwide sample of voters on line. some states i would be comfortable in doing an online poll. once you get under the state level and even in some states i don't have a comfort level with the online studies. now, that is not to say you couldn't get a good panel. but a lot of times some states now -- most states are getting e-mail addresses we people register to vote. a lot of us have sort of that e-mail address that we give out on things that we don't really want to but you have to. we never really check it. phone polling isn't perfect, internet polling isn't perfect.

there are tradeoffs with both. for the most part i'm for an election, for a campaign i'm going to be relying more on phone polling right now than i will be with online. different kinds of outside of the political sphere it is much more online these days than phone polling. i think that is going to keep changing. we sort of see the lines crossing and we are probably a couple years out from almost going exclusively to online. but you -- there are problems with both. we are going to completely underrepresent lower income groups and voters that don't have regular access to the internet. you have problems reaching people who again you get the people who are at work on line all day, have a very easy time of doing a survey online mu.

not to say phone call polling is perfect but with the way the voter files are right now you are going to probably get better information from a phone poll than online. >> what information would i need to give you and what information would you like basically would i be writing the questions for you and giving you topics to ask or how would that work? >> good question. basically a good pollster and campaign will be a sitdown whether on a conference call or in person, especially for the baseline of talking about, ok, what do we think the big issues are? what do we need to know about the issues? what are the policies that we are going to be thinking about talking about in and one of the more important things is what is this candidate going to like to talk about? bus regardless of what you tell

tell to talk about they are going to talk about a what they want to talk about. and the way i look at it is you are going to talk about your two pet issues regardless of what we tell you. what is the right way to talk about it? figuring out from whether it is the candidate themselves or the campaign manager who has a sense of god they really love education and migrating birds, something ridiculous. ok, let's figure out000 tie it into something that means something to the voters because they are going to be that day where they are bored and go off on a tangent of this pet project of theirs and you would really hope they will talk about it in a way that can be tied back into meaningful policy or issue. for a campaign it is sort of getting that background information of here is what we know and what we would like to find out. th

then i will request back and with by colleagues and work out a draft poll. that would get sent out to the campaign. then there will be time for everyone to look at it and then there is a conference call ag n again. we would walk through and talk about it. then this is where people have different ideas of000 word a question or we forgot to put this in, add this, change this, et cetera, et cetera. and there are usually a couple of rounds of drafts and it gets to a point where everyone is satisfied with where they are and think we have the best poll going into the field. then the calling happens. then i farm that out to the a calling house. there are calling houses all around the country. these are things you don't need to worry about. but a good pollster will match the calling accents with the people they are calling. if i call into a southern house

i may use a southern calling house. if i call new york i am not using a southern calling house. you get weird answers and it is especially important if you have a very diverse population that white people calling black voters get different answers than black people calling black voters and vice versa. and again a lot of it goes into a good poll -- and this is what falls on my shoulders -- you need to make sure you are not bias i biassing the questions and asking questions where people think what is the right answer and they will think what is the person on the phone wanting to hear? if this is a white person and they are pretty sure the person talking to them is black and will is a question about does barack obama's race have any effect most people will say no anyway even if they are lying

but they are sure as hell not going to say it if they are pretty positive they are talking to a black person on the other end of the line. so there is making sure you have a good mix of callers that tend to reflect the demographics that you are calling into. now you can't make it perfect. it impossible to say you have these people calling these people but you try to get it as perfect as possible. then we have the deliverables on the background. usually the day after the calling ends, depending on how big the sample, three nights or four nights, then you will get the top line results, midday and the next day. then you usually get the breakouts, maybe later that day or possibly a couple of days later because that takes a little more time. and if you are doing a baseline poll will is not that huge rush. with trackers we work at night, 8:00 a.m. today we go over the

numbers, we have this ready. it is important to make snap decisions on what the tracking poll says. but you are 10 months oupbt from an -- 10 months out getting the results the day after is not the most important thing. so as much as possible i like to take a couple of days to go through the data to get a picture so i'm talking intelligently to the campaign as opposed to just opening up the book and looking at pages and saying that looks like it could be interesting and making decisions based on i have a hunch that ended up being incorrect. and it happens where a lot of times the first reaction to something in the data will turn out to be not founded, or then i look at something deep ever and say i'm seeing this trend in women, i think. 10 you start looking at it and it turns out it is more an education based trend of what is going on. that is what my job is it to

figure that out. and then there is usually a presentation of findings where it is really going through everything. it like these are the issue that are important, this is how voters view these issues. these are your target voters. the voters who are urn decided, these are independents. these are the voters there is no way in hell you will win regardless of what you do. don't bother wasting money reaching out to them. they won't be in your camp regardless of what you do. so that is the biggest piece. really, what you are bringing in from a campaign titan is this is what we do know, this is what we need to know, then the rest is done by the pollster with input from the campaign and other consultan consultants. media should be weighing in because they have different ways of looking at things.

and i'm just one person and someone else may have a good idea of how to word a question or an issue that might be important that i'm not thinking of. you have to eat my humble pie and take their advice and go with it. but the it is a collaborative effort and you don't want that your bottle do not have things answered with the poll because they were not included. you want to make sure everyone who is on board is part of the process to make sure everyone is getting their questions answered. >> would you do the baseline poll before somebody declares their candidacy or afterwards? and along the sam lines, how do you get -- how do you find your clients and how do potential clients find you in is it a networking thing or you develop

a reputation? >> sometimes candidates at all into your -- sometimes candidates fall into your lap. sometimes you hear about somebody who is considering a run and you find a connection. it sort of works in a lot of ways. a lot of times existing relationships with with other people who work on campaigns. the campaign managers i have worked with in the past will get on a new campaign and say got a new campaign, would like you to do the polling. sometimes -- a lot of times a pollster tends to be one hired from the consulting class. but a lot of times a media consultant was with the consultant and they recommend that the candidate or campaign manager give me a call. so it is a little bit of reputation, sometimes past experience and working relationships. for the announcement, again, it all depends on what is happening

in the campaign. if you have a candidate that is 100% going to run for senate in new jersey regardless of what happens but you have made a decision you are not going to make an official announcement for three months because you want to wait until the papers get the signatures and you think that is a better time to get more media attention, whatever the decision that there is a lag between now and when the official announcement will be made, again i would say as long as the money is there this is information that you need to know and the one thing you can never get back ever on a campaign is time. and i guarantee you that on every campaign you work on you will be a week out from the election and you will say i wish we had another week and then you are going to think back to that time in january or february when were not doing things that you maybe could have been doing and you will pull your hair out and what you wouldn't give for that time.

so, the earlier you can get an understanding of what is going on, where you stand, what you need to talk about the better. and that also applies because maybe in your ideal world transaction day was the perfect day it roll it out because they are thinking and we think the beer connection is great because the guy used to own a brewery but two weeks beforehand a news item hits and people know you are thinking about running and reporters are running up to the campaign what does your campaign think about x or they are talking about the candidate, what do you think about this and you have no idea what you should be saying or what people think, how could you be framing the message in a way that will resonate with the largest number of voters so you will i give an answer that may or may not be effective. and if you had done the research and had that understanding you will be able to adapt to what happens on the ground because no

campaign ever works the way you want it to. they are always going to have hills and valleys and that understanding will sort of let you get through those changes. and the earlier you get an understanding of where you are and where you want to go the earlier you can make the decisio decisions. is this an important issue or is this something that can help us, hurt us or just ignore and it will go away in a day or two. so if you don't have that information you are going on assumptions an guesses which could be right but at the end of the day i would rather be making my decisions based on facts and knowledge versus assumptions. other questions? my time is up.

>> [inaudible] pan span .