How to ask a question: conducting research for your startup

TOMER SHARON: Hi, my name is Tomer Sharon,

and I'm a Google Search User Experience Researcher

and the author of "Lean User Research."

I'm here to help you ask your customers, users,

and potential customers better questions

and get even better answers.

"Converse like a talk show host, think like a writer,

understand subtext like a psychiatrist,

and have an ear like a musician."

This beautiful thing was said about interviewing people.

An interview is gathering information

through direct dialogue.

Before you start an interview, though,

ask yourself why before you get into the

how to phrase the questions.

Think about the reason you're there.

What is it that you're trying to learn?

A key aspect of interviewing people

is getting familiar with the phenomenon

that is called rationalization.

Rationalization is when people, when

asked about a certain situation that happened in the past,

they tend to change reality a little bit

to be perceived as good people, to be

perceived as smart people.

The example I have for this is if I have ask you,

would you throw a banana peel on the floor if nobody's looking?

You would probably say no, I don't do that.

I'm not that kind of a person.

But given the opportunity that no one is looking,

two weeks from now, you might do that.

And it's not because you're a bad person.

It's not because you don't believe in what you say

or you're trying to lie to me.

You rationalize.

You want to be perceived as a good person,

as a helpful person.

When you interview people, when you ask people questions,

look for the story.

This is the most important thing in an interview.

Ask about stories, about things that happen.

Ask about behavior.

For example, so in that situation,

you keep quiet and see what people say.

Better yet, if possible, ask to observe behavior.

If you're asking how they read email,

how they check their email, ask them to show you that.

Show me how you do that.

Ask about perceptions.

Not because you're interested in them,

but because they will help you, again,

to pull out these stories.

For example, what is the difference between X and Y?

Not a very interesting question, but it

might cause them to remember something

about something interesting that happened.

The most important thing in an interview

is something that doesn't look like the most important thing--

follow-up questions.

These will tell you, give you, a lot of insight

into what really happened.

Why do you roll your eyes when you say that?

What do you mean?

Why do you call your phone "my third arm"?

These are extremely helpful in understanding behavior.

A few don'ts.

Don't ask people about the future.

Three questions product developers tend

to ask their interviewees are, would you use the product?

Would you pay for it?

And how much you'd pay for it.

These are all very, very bad questions.


Because they ask people to predict the future.

The problem is, people have no idea.

The even bigger problem is that they

think they do have an idea, and they answer.

Another don't is about leading the witness.

Don't insert your opinion into the questions you're asking,

and then biasing your interviewees

into understanding what you're looking for

and then give you the answers that you want, maybe.

Don't intimidate.

If you stand over the shoulder of someone who's sitting

and then ask them a question, that

is an intimidating body language.

Don't do that.

That will affect how they answer your question.

Don't explain the question.

Shut up and see what happens.

When you ask a question, sometimes there's

silence from the other side.

Maybe the person is thinking.

Maybe the person is trying to remember something.

Maybe they don't understand the question.

Keep quiet.

Your intuition would be to explain the question.

Don't do that.

Just wait a few seconds and see what happens.

People will fill in the gap.

And finally, don't ask for feedback.

This is another very popular thing

product developers do during interviews.

They describe the product or show off the product

and then ask for feedback.

If you have the product, ask people to use it.

Even if you have a mock of it, ask people to use it.

Use a pencil and use it.

To sum it up, here is a little exercise

about leading questions.

I'm going to show you a few questions,

and let's talk about whether they're leading or not.

First one.

"Would you rather use the current version

or this new, improved one?"

The word "improved" is a way for you

to insert your opinion and biasing people into the answer

that you're looking for.

"How does this compare to the way Netflix works now?"

Now you're biasing people with their opinions about Netflix.

Good, bad-- you're biasing them into an answer

that is not neutral.

Try and ask about the thing without comparing it

to anything else.

"What did you think about that?"

When you ask that, you are forcing people

to think about something they might not

have thought about if you didn't ask.

Don't do that.

"Would you click here to log in?"

It's exactly as saying, click here to login.

This is an extremely biased question.

"What's wrong with this?"

Again, inserting your thought that this is wrong,

and then expecting people to cooperate and say

that it's wrong.

Here are three resources I recommend

for learning more about how to ask questions--

a book, an article, and a video.

The book is called "Interviewing Users," by Steve Portigal.

The article is "16 Interviewing Tips,"

by Michael Margolis from Google Ventures.

And the video is "What people Are Really

Doing," by IIT's Institute of Design.

Here's a quick recap.

Look for the story.

Ask about behavior.

Better yet, ask to observe behavior.

Ask follow-up questions, extremely important.

Don't ask about the future.

Don't lead the witness.

And don't ask for feedback.

But do ask great questions.

Thank you.