Big (Easy) Data

September 17, 2014

Surveys and polls can be a great way to provide a springboard for your brand. Few are aware of how easy and cost effective a professionally conducted survey can be. Following the success of her survey on parenting stress  for the Clinton Foundation, Annabel Kelly provides pointers on how to design and report on an attention grabbing survey.

Annabel 006

After Viagra was approved by the FDA to treat erectile dysfunction in 1998, its release was boosted by headlines reporting that while older men the world over were aware that erectile dysfunction was a side effect of aging or stress, they did not appreciate that medical conditions such as hypertension or diabetes were also significant factors. The new brand also drew attention to the communication barrier between men and their medical practitioners when it came to discussing concerns about sexual function. An early example of a company mining its big data? No, the insights were generated through an impartial third party, Ipsos, through a survey of over 4,000 men aged 40 years and over across Asia, Europe, Central/Latin America, and Africa.

Polls and surveys can be a great way to provide a platform for your brand’s talking points and outreach activities especially if you lack the data to support your observations or you have the data to support a position but don’t want to risk alienating your customer base by releasing it. While many are familiar with the concept of polling, few are aware of how easy and cost effective they can be. Read on for tips on how to design and report on an attention grabbing poll.

Ensure a Robust Design

Polls that attract widespread media attention almost always boast a watertight design. Discerning journalists check press releases for a respected and objective fieldwork partner; a large enough sample size; fieldwork dates, and a robust methodology. While an invitation to respond to a question on your website or social media page can be engaging to passers by, the results of these do-it-yourself polls, at best, only reflect the views of a very limited audience and, at worse, are subject to gaming. Redditors, for example, regularly swarm polls to show support for their very distinct community’s agenda. Needless to say, the results of these polls are rarely picked up by the wider media.

Managing Costs

The  most cost effective public release polls are conducted through an omnibus, the research equivalent of a car pool where a variety of clients commission one or more proprietary questions in a survey but share the cost of recruiting, interviewing, and collecting and processing demographic data. Omnibus surveys are conducted by telephone or, increasingly, online. The gold standard for public release surveys used to be a telephone methodology in order to achieve a true random sample (i.e. everyone within the target population had an equal chance of being approached for interview) but now, with almost universal internet access (87% of American adults as identified by Pew Research in January 2014); the move away from landlines; and an increase in call screening; online surveys of a panel of respondents balanced to represent the population are now a commonly accepted approach.

Define Your Objectives

If you’re thinking about commissioning a poll, the first step is to define your objectives. A good way to nail these down is to develop your ideal press release complete with graphics. While results cannot be guaranteed, having a goal to communicate will help ensure that all your bases are covered in the research design. Make sure your topic is not too self-serving: if you’re a pizza company and your results show people eat pizza, you’re not going to surprise anyone. Also, before getting too far down the road, Google, Google, Google to make sure that your idea hasn’t already been covered by a reputable source. For a pointer that you’re moving in the right direction, look for aspects that have already been addressed and saw traction in comparable territories such as the UK or Canada.

Make Sure It’s Newsworthy

Readers of the general media are most interested in polls on health; the economy/jobs; food; crime/safety; and family/kids (Ipsos 2009). Business journalists, on the other hand, are interested in data that support your business model and strategic goals. It’s always a good idea to talk to your target journalists early on about what they think the information gaps are around your area of focus. A great way to get journalist buy-in is to offer them an exclusive and invite them to contribute to the research design in some way.

If you’re struggling for a subject matter, look to your social investments: would a survey be helpful to your charity or community partners? Many NGOs are desperate for data to support their goals and would willingly partner with a corporation to acquire and communicate it.

What Audience?

When deciding on what audience to interview, it is worth noting that surveys of elite audiences such as C-Suite, medical professionals, legislators etc. are extremely newsworthy. They are also very expensive. Cheaper is to survey people that meet certain demographic criteria – retail workers; parents of kids with allergies; or working moms for example –  through a number of waves of omnibus or specialized panel. Some of the global research companies offer international omnibus.

Provide Context

When designing your questions, make sure you provide context: if X% say they are positive about A, it’s helpful to be able to compare that with how they also feel about B, C and D. Also, don’t blow your entire budget on one wave of research. Providing regular updates on at least one of your key metrics will show which way attitudes or behavior is moving and help your brand command ownership of a topic. The big fieldwork providers offer omnibus on a regular basis so it’s possible to run questions over daily, weekly, monthly or annual waves.

Provide Depth

Give your survey depth by allowing respondents to elaborate on their point of view in their own words. Open-ended questions (that can be coded after interviewing) provide illustrative quotes, greater insights, and humanize the quantitative findings.  You can also add color and draw attention to your survey by inviting relevant personalities to respond to some of the key questions and report their point of view alongside the main findings.

Brand Your Survey

When releasing your results, giving your survey a brand will add credibility and gravitas. Some examples of survey brands include IBD/TIPP’s Economic Optimism Index; Sallie Mae’s How America Pays for College; and’s What did you call me?

Time Your Survey

If appropriate, try to time your survey so results are available for relevant events. For’s What did you call me?, a survey of American’s views on their first names, results were published and reported alongside the Social Security Administration’s annual release of the most popular baby names.

Increase Engagement

To boost engagement with and the shelf life of their survey results, some sponsors are embracing the current quantified self movement and inviting stakeholders to respond to the same survey questions before presenting their answers in the context of the wider population and demographic peers. For examples, see the quizzes developed by the Pew Research Center where visitors can compare their knowledge of and attitudes towards different topic areas with the general population.

Annabel Kelly is a 20 year veteran of opinion research.  Her surveys have received global coverage and have spanned topics such as financial literacy, breast implants, innovation, first names, asthma, the food business, the environment, corporate social responsibility, and leadership.  

Image via wikipedia