The truth isn’t always pretty. Any market researcher who has gone back to a client with bad or unexpected news knows this all too well.

‘Our job is to give them the good news and the bad news – not to figure out some way for them to look good,’ says Trish Simmie, VP, Camelford Graham Research Group in Toronto.

While researchers admit that there are clients who just don’t get it – who want research just to shed a positive light on a brand or campaign – most marketers do understand that blatantly leading questions can come back to haunt them. Even so, all marketers have preconceived notions of what they’ll get back from a survey, and results contradicting those notions can be hard to swallow.

Here, then, is a list of considerations that marketers should keep in mind if they want market research results that reflect reality.

Do’s

* Do offer an ‘other.’

Aided or multiple choice questions (where a sampling of brands is used, for example) should always include an ‘other’ for the most accurate results, according to Simmie. ‘Don’t force them to choose one of three answers when, in fact, number four is correct.’

* Do commission both quantitative and qualitative research.

‘The quantitative is the framework, the structure, the skeleton,’ says Marion Plunkett, president of Plunkett Communications in Toronto. ‘The qualitative is the flesh on the bone.’

* Do aim for balance.

When you design a scale, try to reflect the full range of opinion, says Simmie. For example, the ‘excellent, very good, good, fair, poor’ scale, while widely used, is not great because ‘good’ is actually a relatively low rating. A better scale would be ‘very good, good, fair, poor,’ she says.

* Do be prepared for findings you didn’t expect.

Have an action plan in place that will deal with the good, the bad and the ugly.

Don’ts

* Don’t prejudge.

‘If you really, truly want to know how consumers feel in order to build a strong brand blueprint, then listen to what they have to say,’ says Plunkett.

For example, a marketer watching a focus group is tempted to zero in only on that information supporting his or her presupposed notions, she says. ‘It’s human nature.’

* Don’t ask people to comment on things they know nothing about.

Asking the general public ‘Is Canada in a recession?’ makes absolutely no sense. In the same way, says Simmie, asking people about products or services they have no experience with can lead to questionable results.

* Don’t assume that how they felt yesterday is how they feel today.

While past research is always useful, it shouldn’t be given too much weight in the current research, advises Simmie.

* Don’t play dumb.

Recognize that everyone has a tendency to seek data that supports a particular viewpoint, says Alan Middleton, marketing professor at York University’s Schulich School of Business.

‘What you include from the research will always be the data that fits with the argument you’re making,’ he says. ‘You will probably weed out the data that contradicts or raises doubts.’ To avoid this, put in your own checks and balances. If you are senior management, ask to see the raw results.

* Don’t provide a leading lead-in.

For example, prefacing a question with ‘Many people are concerned about limiting their fat intake…’ is just telling people to respond that they are concerned about their fat intake.

* Don’t rationalize.

When a marketer is under pressure to get results, he or she might rationalize a finding that is just under the acceptable level, blaming a degree of error, or even basically ignoring the bad news in order to move forward, says Middleton.

For example, when deciding whether to introduce a new product, you might set an action standard of 60%. If you get 55%, what do you do? It’s a form of creeping commitment – and it can be risky. ‘You get all this kind of rationalization that goes on,’ he says. The onus is on the marketer to stick to the original action plan.

* Don’t be a client from hell.

Stay involved – but not so involved that you’re trying to lead questions. It comes down to trusting your research company.

‘The only way you’re going to get any validity is to do your very best to do unbiased research and understand that you may win or you may lose, but if you try to bias it, you’re going to lose – and look like an idiot,’ says Simmie.