Tight budget constraints limit the kind of research and engagement political campaigns (or non-profits, or unions) are able to do. The most common question we get is: how much will it cost me to run a poll with 500 or 1000 completed responses?

The answer: it depends. The cost is almost entirely based on the response rate - how many people complete your survey, vs. how many hang up or don’t answer. The higher your response rate, the fewer calls you have to make, so the less the cost.

Organizations contacting their members tend to get response rates between 10-25%. We have seen political polls to the general public get response rates in the 3-7% range, though some have gone as high as 10%.

So how do you boost the response rate and make every call count? It turns out to be more of an art than a science because it depends on so many human factors. Based on what has worked well for our clients, we have compiled a list of the top 10 tips for improving your response rate. If you have experience with these approaches, or suggestions for others, please share them in the comments.

1. Use a local caller-id
   The caller-id is the first thing someone sees when you call them. If it’s from a local number, they are much more likely to pick up the phone. Precision Polling lets you select your own caller-id.
2. Call in the early evening
   If they aren’t home, they aren’t going to answer. The best time to call on weekdays is 6-8:30pm. Friday and Saturday evenings aren’t great, though you can call at almost any time on Sunday.
3. Keep your intro snappy
   Get straight to why you’re calling in the simplest words possible, or you may risk them hanging up. The best intros squeeze in what you are calling about; why the participant should care; who you are; and how long it will take. Of course, you have to also make sure that you meet any disclosure requirements (some states require an organization name and phone number, for example).
4. Use a recognizable voice
   If polling the general public, try to get a celebrity to record your poll (e.g. TV anchor, football coach), or failing that, use a female voice with a neutral accent. If you’re polling your own members, ask someone that they will recognize to record the poll, like your organization’s president.
5. Pump up the volume
   Be sure to test your poll to make sure it’s easy to hear. Recording with a loud and clear voice gives the best results.
6. Keep your poll short
   Automated polls work best when you’re asking no more than 10 questions (2-3 minutes). Any longer, and people start to hang up. As a rough rule of thumb, for every 10 additional questions, your response rate will drop by 1%. One way around this is to structure your survey so the important questions are at the front; and then define the survey to be considered as “complete” if the respondent finishes these important questions, with the remaining questions as optional (though useful) data.
7. Use simple sentences
   Simple language and short sentences make it easier to follow questions, especially when listening to them over the phone.
8. Keep it relevant
   Some of the highest response rates we’ve seen are for targeted local issues (e.g. mayor of your town, or pollution in a small neighborhood), or widely discussed topics (e.g. healthcare). Introduce your survey in a way to make it relevant, perhaps by mentioning a hot-button issue you know the respondent will have an opinion on.
9. Call likely voters
   If possible for your survey, call those people who are most likely to be actively engaged in the political process (look for good voting histories) because they generally are more willing to pick up the phone and share their opinions. Outside of politics, looks for other signs of engagement, like member attendance. Of course, only do this if it won’t bias your desired results.
10. Speak their language
    We’ve had customers record polls in dozens of different languages: Spanish, Mandarin, Vietnamese, and Russian to name a few. You can do the same by recording questions in multiple languages and starting your poll by asking which language they prefer. The response rates tend to be much higher for non-English-speaking groups receiving a phone call in their native language than it is for English-speakers receiving a phone call in English.