Founder EASTWEST PR and Host of the SPEAK|pr Podcast
Organisations, governments, companies, and political parties all use polls to drive conversations and create compelling content, which is one of the main elements of public relations. Polls give an opportunity to establish credibility and to be a thought leader on any topic. The caveat, though, is that in order to be a thought leader, a poll or a survey must be seen to be independent and have enough scale to be meaningful.
Polls are an avenue to showcase one’s expertise. Through it, you can tell a story or create a narrative around a product or service that can be justified by the data from a poll. It can also be used as part of substantiation for news that’s going to be in a press release, article, or white paper. Polls can be used at events like (virtual) press conferences or product launches as well as in brochures, in advertising, speeches, and annual reports. There are many different ways that this content can be used and repurposed.
How to make polls that will produce relevant results
To create a poll, first begin with the end in mind. Think about the story or initiative to be established in the media and the data needed to do that. Next, think about the content. Issuing a poll that no one is interested in will not get you anywhere. Simply look at the current media or talk to people and see what’s trending and what questions are not being answered. That’s the void to be filled with information discovered from polling. Once there is a topic, think about questions that will produce meaningful answers.
One important point is that the questions themselves should be of critical importance, because how the questions are crafted will determine the way they are answered. Avoid questions that sound biased, such as if you craft a question this way, "All people believe this is wrong and it shouldn’t happen. Do you agree: yes or no?" The question is loaded with a negative bias, which may cause apprehension and so some may therefore choose not answer it. Instead, ask questions that are even-handed and will elicit an accurate response. The media may even want to see the questions to understand the data, so if the questions are self-serving, then the data itself will lack credibility.
The next step is to think about the format of the question. It can be a yes or no question, or asking people to rate something on a scale from 1 to a certain number. There can also be a blank space for the answer. It’s good to have a mix of these kinds of questions, because the yes or no questions create a simple chart, while the scaling creates a sense of relativity, and the open text creates a narrative. It gives your respondents the chance to provide answers you may not have thought about. Afterwards, think about the headlines and the stories that will be told, as the format of the questions can be geared towards that. For instance, for a research-heavy white paper, you’ll want quantitative data, so including yes and no as well as scale questions could be useful. If the poll is for a blog where you want to showcase a narrative, the questions could be more of the open text type, as the goal would be to gather thoughts and foster involvement from the participants in the survey.
Thinking about the structure of the content is important, because not only will it determine the structure of the questions, but the structure of the questionnaire itself as well, how many questions to ask of which type, in which section they’re going to be in, and ultimately, how long it will take to answer the poll. The number of questions is worth thinking about in order to have significant results. However, do remember that a quick poll may give you simple answers that might not produce meaningful data that would have been subsequently translated into content.
Incentivize people to answer, but not in a way that would skew the results. This is an effective way to get answers, but always look at the cost of what will be given away. There could be a reward for every nth person who answers the form or a lucky draw among the respondents. Next, consider the size of the sample. To be statistically significant, the sample size needs to be credible. For an internal survey, the aim would be to ask everybody, but 70&-80% is also reasonable enough. If it’s an external survey, how many people do you need in order to turn the survey into something that you could actually take to the media and say this is credible?
One view is that a sample size needs to be a minimum of 1,000 respondents, but that might still not be representative of the population. The general rule would be to take into consideration the overall universe of whom your questions are about, and then identify a percentage that would make the results relevant. A sample survey could first be send out to a small percentage of the target audience just to proof test the questions. This will help determine if you’ll be able to get the narratives you’re looking for. Bear in mind that there will be a margin of error, so the sample size needs to accommodate a margin of error of up to 5%.
Once you have all of these, you can use a platform to create the poll or you can use an a third party agency. The benefit of a third party agency is that you will have the credibility of the results that are produced by a third party, because research is big business. There are companies like Nielsen, which has revenue of over $6 billion, Kantar which has $3.85 billion worth of research, IQVIA (formerly Quintiles and IMS Health, Inc.) with $3.5 billion, and then there’s Ipsos which has over $1.9 billion in revenue. So, if you’ve got the budget and scale that requires global research, there are companies that can do that, but there are also smaller independent agencies as well that can create surveys.
Quantitative market research where public opinion or business opinions are gathered can be done in house or with third party agencies. Third party agencies also have the benefit of understanding how questions should be written, which is also really important and could be slightly difficult to create on your own, as it does take some skill to write questions that are balanced, comprehensive, and tease out interesting facts and figures from people. PollDaddy, SurveyMonkey, and even Facebook are some tools you can use if you want to create the poll on your own. Poll.app.do is a Facebook-specific app that creates polls, and its pricing is between $8-$28 a month to survey 1,000 to an unlimited number of participants per poll, so if you’re looking for a polling tool that specifically works on Facebook, this app will do that for you.
LinkedIn also has its own free and simple polls. PollDaddy and SurveyMonkey are two of the major platforms for polls, and they produce great output. The questions are simply placed into their system, and then they’re provided with the database of respondents to which the surveys will be sent out to. The charts and the narratives from these polls will then create the opportunity for media coverage. To create coverage, show facts that can be delivered as a press release and also as an infographic which is the combination of information and illustration. If you can create contrarian results or evoke shock when people wouldn’t have expected or believed something to be true, then that’s guaranteed to get you interest from the media, especially if it’s accompanied by a well-written press release, article, or an invitation to an interview.
Simplicity is the next key element of a visual. People don’t have the time to read through all the results. They want a graphic, possibly a picture showing the key results. With this, think about who’s going to be reading the infographic. Where are they? Are they looking at it at work or at home? On a mobile device or on a big screen? Would you also want to make this a graphic that they could share and use in their own presentations? Because if something is novel and easy to read, then it’s shareable just from the quality of your content. Then think about fonts, colors, and design. There are platforms like Canva with infographic templates (or you can even build your own) and Visme. These will help you turn data into something appealing.
This article is based on a transcript from my Podcast SPEAK|pr, you can listen here.
Cover Photo from ArtDivision