Today we have a guest post about a topic we haven’t talked about before: PR surveys. Do you do PR surveys? Let us know your thoughts in the comments after this post.
What is a PR Survey?
What constitutes a PR Survey is not universal. For some people, any survey which collects data which is then used for PR purposes is a PR survey. It could, for example, include a survey to explore key audience messages, or to support the promotion of a brand, or create new content, including an ebook or info graphic. But for most people, a PR Survey is a poll. It could be a flash poll (designed to pick up on a very current story in the news) or a survey to set the agenda with new insights.
Flash polls and the idea of “newsjacking” merits posts in their own right. But here are 10 tips to help create an effective poll.
1) Keep the questions short
It’s generally best to keep both individual questions and the questionnaire as short as possible, always concentrating on the key objectives of the survey and the potential headlines you want to generate.
2) Use closed questions
For most surveys closed questions will generate the results you need. Open ended questions can generate some credible quotes, but in general are of limited use for media content. Ultimately if you need to report % and mean scores, there is little point in asking open questions.
3) Credible stories
Quirky or fun stories make good headlines, but unrealistic surveys do not. Whatever the angle, the survey must be seen to deliver something objective and credible.
4) Forced answers
Too many PR surveys are still based on leading questionnaires. For example, limiting the realistic choice of answers or excluding doesn’t know. This approach will increase your chances of getting the “right” answer but it also risks being found out by more inquisitive journalists, or delivering skewed answers.
5) Sample size
Nationwide surveys (nationally representative or gen rep) typically should be 1,000 interviews. Some polls are 2,000, although in reality 1,000 interviews often will suffice. The sample size for niche groups, such as young mums, or Business Decision Makers can be as few as several hundred to be credible, although larger surveys are best if the budget is available.
6) Think international
Conducting international surveys is more affordable online than it used to be. Consider a pan European or surveys with a mix of countries to represent different regions.
7) National or regional?
National surveys are popular but regional or even city based surveys can be used to focus on the regionalpress. Be careful not to combine national and regional surveys without a sufficient sample size. 1,000 interviews provide scope to compare London with the North East, but often not enough interviews to compare London with Birmingham. Consider 2,000 interviews or a national poll of 1,000 but topped up with additional interviews in the regions you are more focused on.
8) Do you really want rank order questions?
Ranking a question in order can sound like a good idea, but sometimes PR execs struggle with how to report the results. Are you interested in the rank order, or do you just want to know what % mentioned x,y and z.
9) Single or multicode questions
Think carefully whether each multiple choice question should have a single code or multicode answers. If you only have a few answer options, a single code is better. If you have a long list of answers, offering multicode is likely to result in a greater spread of answers (e.g. each with lower %). Consider limiting multicode to the top three or five answers.
10) Be realistic with B2B surveys
You are not going to reach C suite respondents in large companies via an online panel. Consider whether you need to reach the most senior job titles, or whether other Business Decision Makers are not only more realistic to reach via a panel, but actually may know more about the subject (e.g. internet security is something an IT network manager will know about rather than the CIO!)
A final thought on one of the most important issues of being effective, in terms of outcome. Data quality doesn’t make the top 10 tips because it’s not an easy tip to explain, but it is important. Data quality is often not seen as important until the data is shown to be wanting.
There are so many new ways to gather straw poll data, including online services that intercept respondents via websites. I have a big concern with this type of service and data. There are questions to be asked about whom the respondents are and the quality of response. Polling companies are concerned with issues like having a robust sample, representativeness and removing poor quality data. If they do their job properly, data quality doesn’t raise its head above the parapet. Unseen, but still important.
Neil Cary is an experienced market research consultant. He founded the online survey tool, Surveygoo.com, and the network of online consumer panels, Opini. He also established the online research company, Asia Opinions. He contributes to the Surveygoo blog.
Elena is founder of a technology PR agency that works with startups to billion-dollar companies. She is passionate about helping marketers and small business owners with practical publicity strategies, which she's also using for her own bling flip flop company.
Did you enjoy this post? Subscribe now and get new posts either via RSS feed or your email. Please click on the RSS button below or type your primary email address below and click "Subscribe."