Making Your Opinion Count – Guest Post

Making Your Opinion Count

Submitted by Paul DesBarres, President of Nova Insights Market Research & Consulting in Kentville, Nova Scotia.

When many people think of market research, they think “polling” and about those supper-time telephone calls. However, it is much more than that. Yes, it includes polling, and it can include telephone calls at supper-time. However, many do not distinguish between market research and telemarketing, but they are very different. Telemarketing is sales, and that’s the only goal of the person speaking to you on the phone. Market research interviewers are after nothing more than your opinion. Your telephone number was likely randomly generated to be a representative of the general population or some segment within the population. You get to provide your opinion to influence businesses and other organizations in their decision-making. I am “in the business,” so I may be a bit biased, but I’ve always viewed this as a privilege to have my opinion count toward helping an organization better serve the market.

My view of market research may not be “representative” of the population, as many dread the ringing of the telephone for a survey, and many will not answer the phone if they see a number they don’t recognize. With the development of call screening tools (e.g., caller id) that have helped reduce response rates of telephone surveys, the reduction in the proportion of landline phones, and the growing prevalence and comfort with the internet, the market research industry’s techniques for collecting opinions have had to change.

The home landline is no longer necessarily the best way to reach a representative sample of the general public. Fully 84% of Canadians and 81% of Nova Scotians access the internet. According to a recent Nova Insights survey, 7% of Nova Scotians do not have a landline in their home. This increases to 13% among males, and 12% among 18-34 year olds. By trying to interview by telephone, we would necessarily be under-representing important segments of the population, and likely ones that have a different perspective than others on some topics.

Online research can be answered 24/7 at the convenience of the respondent. Respondents are not interrupted on the spot and put into a hostile mood for the interview; they have time to be thoughtful in their responses. Sensitive topics are answered more honestly since there is no “live” interviewer to provide perceived judgment. Plus, online research can be much more cost effective and results can be gathered more quickly since it does not rely on live interviewers.

I have an uncle that is aged 83 and has spent his entire career in market research, watching it progress from door-to-door to telephone using paper questionnaires, and then to Computer-Assisted-Telephone-Interviewing (CATI). Some traditional researchers like this ask about the traditional value of random telephone sampling?

Random Digit Dialing (RDD) has been a staple in the foundation of research for many years. The theory is that this random approach will provide a representative sample, not influenced by any other selection method, or even those choosing to list or not list their telephone number. What has happened in the last several years is that those having a landline phone, then choosing to answer that landline phone from unknown callers, and completing a survey on the spot has become much less representative of most folks out there. While this has been happening, computers have become increasingly prevalent (especially among key segments like 18-34 year olds that are difficult to reach on a landline) and so has high-speed, “always on” internet.

Most general population online research is conducted among panels of consumers (or professionals) recruited to be opinion-providers. Those panels are designed and managed to be demographically balanced. The other important aspect of these panels is that they are used for research only (not sales/spamming), so they include willing participants that can complete the survey at their convenience. These panel members have agreed to provide their opinions when it works for their schedule (no commitments or obligations), and they are compensated for each survey (usually about $1 per survey).

The overall panel is balanced, as mentioned, and it is large–usually 200,000-400,000 Canadians. Invitations are sent to replicates within this large sample. A replicate is a randomly selected sample within the overall panel, and due to this random selection, it is also balanced. Multiple replicates may be used throughout the data collection process.

What has happened is that telephone research has become less representative, and online has become more representative over time. They have essentially begun to meet in the middle.

Telephone research certainly has its place, and I still use it extensively. For example, it is the only option for some lower incidence studies for which there is not sufficient online sample available. The availability of opt-in online sample is growing quickly, but is still limited in many respects. However, when there is sufficient availability, it works well and is cost effective. My firm, Nova Insights, used an online methodology that allowed us to be the first to project an NDP majority in Nova Scotia, and the results turned out to be extremely accurate.

As a side note, a common strategy in the U.S. (and growing in Canada) is to use a telephone methodology that incorporates a wireless phone component to attempt to reach that group. Those on the wireless phone must be compensated, and there is a high refusal rate, so this is a very expensive method. Its success in overcoming the developing industry challenges is uncertain at this time. I am not convinced.

I understand the hesitation some have in accepting online survey results. Some view it as those pop-up polls on websites. However, scientific research conducted by professional firms use methodologies that are much more complex, structured, and reliable. There needs to be more education around the changes taking place in the market, and therefore in our industry. Those in our industry have, for the most part, embraced the changes, but we are still trying to educate our clients and the media.

The following two tabs change content below.

Aydin Mirzaee

Aydin is currently the Co-CEO of Fluidware where he spends his time focused on product strategy, marketing and growth for the company's FluidSurveys ( and FluidReview ( products. Aydin co-founded Fluidware in 2008 and since then it has grown to become a market leader with customers in over 80 countries.
FluidSurveys Presents

Free Survey Q&A

Join our survey & research expert Rick Penwarden as he answers all of your questions every Wednesday at 1PM EST!

Comments are closed here.