One of the great dilemmas in marketing is whether to trust your instincts over what the research is showing you. Both the agency (“we know how people think”) and the client (“we know our clients”) have important insights that might not show up clearly in the research, but research can often reveal hidden truths that neither the agency nor the client knew existed.

Long ago, I did some consulting for an Arizona bank that did some research and found, much to their surprise, that the majority of their customers were Hispanic. Somehow this information hadn’t bubbled up to the executive level, and it took them by surprise. The data they’d been looking at had revealed this before, and they quickly realized that some of their customer communication problems were due to their complete misunderstanding of their customer base. Without the research, they’d have continued to drift along cluelessly.

On the other hand, the majority of market research is junk. It’s hard to measure this stuff, and even when you can measure it, it’s hard to know how to interpret it. Surveys lie. Focus groups lie. Interviews lie. Financial reports lie. It’s good to get whatever data you can, but it takes a seasoned professional to know how to wade through that information and decide what’s relevant and what should be thrown out.

If you’re a parent, you might be familiar with the experience of getting an ultrasound of your unborn baby. You’re looking at bunch of fuzzy nothingness on the screen, while the technician points out “Look, it’s a boy!” and “Ooh, strong arms!” and “He’s gonna be a tall one.” Meanwhile, you just feel a little depressed that what you thought was a face turned out to be a butt cheek, and you could swear you saw Elvis in there at one point too. The expert has learned to see things that could easily mislead or confuse the untrained eye. Market research works the same way.

Here are some quick tips for sorting out the useful from the irrelevant:

• Don’t discount instinct: The human brain can absorb and interpret information faster than any computer can. Both the client and the agency will have important gut instincts that should be considered seriously.
• Question data sources: If an agency tells you that 75% of your customers are female, ask them how they know. Where did they get that data? Is it a reliable source? What research methods were used?
• Use ethnography: Ethnography is the act of observing people in their natural environment, which can provide much more accurate data surveys or interviews can
• Work with experts: Market research requires a subtle touch, and you should find someone who has the experience to interpret results effectively, and the clarity to explain the findings in plain English.
• Be ready to change your assumptions: If you think your customers are one thing, but the data consistently points somewhere else, be ready to give up your assumptions. It can be hard to change your worldview, but sometimes that’s the only thing keeping you from making an important change.
• Iterate: You might not know everything up front, so test some ideas and see how they work. When in doubt, do something.