With any big project, it’s easy to lose sight of the end goal and accidentally go off on a tangent. Whether your end product lacks continuity or you’re kicking yourself for wasting time, it’s definitely something you’re going to want to avoid.

Luckily, the solution is simple. Before you start, set yourself a SMART goal. It must be:

• Specific

• Measurable

• Attainable

• Relevant

• Timely

For instance, you might decide that your goal is—in two weeks time, I need to know which websites my customers go to on a daily basis, so I can help the marketing department distribute its advertising spend most effectively.

As long as it’s SMART, it’ll keep you on track. Write it somewhere visible in your office so you and anyone else involved in the project are constantly referring back to it when making decisions about the market research project.

Keeping surveys short and sweet is key to a high response rate. When it comes to the final draft, you’ll need to be brutal—chopping out any questions that are even slightly unnecessary. You can call yourself Edward Surveyhands.

Defining what you do know is as important as what you don’t know.

By getting lots of different people in a room to brainstorm, you can pool your knowledge about your ideal customers. Having people from different departments will give you a more rounded, true representation of who you’re talking to when you’re communicating via a market research survey.

Here are the questions you’ll need to answer before you begin writing your market research questions:

• Who is our ideal customer (and is it your current customer)?

• What do they struggle with?

• What do they really want?

• What sets us apart from our competitors?

• What benefits do our customers think we can give them?

Some people call this buyer persona development.

We just call it getting to know people.

Once you’ve written down what you do know, it’ll spark off lots of thoughts about the gaps in your knowledge.

Because of this, it’s likely that you’ll come out of the brainstorm with a list of questions as long as your arm.

But as your final survey should be as short as possible, now is the time to get brutal. Splitting your questions into topics will help you decipher which questions essentially provide you with the same information, while also helping you get as much information from as few questions as possible.

A general rule of thumb in market research is to start general and get more specific later on. Cover the biggest, broadest question first, before zooming in on more detailed questions.

You could also consider doing multiple surveys in this way. If you still had a long list of questions that needed answering, you could send out weekly or monthly surveys.

If you do this, make the first one as broad as possible. After that, you can pick and choose who to send which detailed question to.

Think of your survey as a one-to-one dialogue.

Once you’ve got your final draft, read it out loud to another person.

Do the questions seem self-serving? Have you used conversational language? If someone asked you this question in real life, how would you feel?

To get the most valuable information, you’ll need to avoid any kind of leading language. At the same time, to inspire the most thoughtful responses, you’ll want to give them an emotive reason to answer the question.

This can be a tricky balance—but working with another person helps to ensure that your survey is just part of the bigger conversation you’re having with your customers.

A lot of the time, people assume that all of their customers would ideally take part in their market research survey. That’s certainly not the case.

It may sound counterproductive, but you should actually send your survey to as few people as possible. That way, you’ll end up with the most hyper-focused think-group possible—allowing for some truly insightful information from key questions that have the power to drive effective change in your organization.

If a customer hasn’t purchased your product or interacted with product features in a long time, they’re ideal for a survey that aims to re-engage previous customers in order to increase customer satisfaction. They won’t be quite so suited to a survey that looks to get feedback from a recent tweak in the product.

Similarly, think outside your immediate audience. If you’re hoping to get insight on how to reach a new demographic, there’s not a whole lot you can gain from sending a customer survey that targets your existing demographic—you want potential customers, not existing ones.