It's important to have some context explaining what the survey is that people are going to fill out, even if they were directed here from an email.

This text gives the reader a reminder about the goals of the research and sets expectations about the amount of time that this should take.

I learned this naming convention during my time at Slack from Nicole Zeng, who ran our Quantitative Research Team and built our initial tracking surveys.

This naming convention was helpful because we did the analysis in R and combined survey question responses with usage data. The answers to these questions were coded with the variables named here so that we could reference them.

Q stands for question, the second part is the shorthand for the question (i.e. name), and oe refers to open-ended response.

I think we're asking a few too many questions in this screener overall, but because we were recruiting people for whom we had little internal data to go off of, we tried to get visibility into other signals so that we could evaluate their fit for the conversation.

If we had better internal data, we probably would not have had to ask so many questions here.

One of the reasons that we asked this as a closed ended question is because we wanted to be able to programmatically filter out people who chose specific responses. If we had done this as an open-ended question, it means we would have had to read and process people's responses before filtering them out, which would have added unnecessary time to the process.

There are a lot of different ways to cut up company size segmentations. Slack had a much different one that grouped everything less than 250 employees as SMB (small and medium businesses) and had more divisions in the thousands of employees.

For our purposes, these distinctions were meaningful up until a certain point (over 100 employees), so that's what you see here.

Having a page break provides a cognitive break for your participants so they don't have an endless scroll (though they shouldn't have to answer that many questions anyway).

One of the ways to determine where page breaks are appropriate is identifying clusters of questions that are thematically related OR places where you may drop certain people who are not a good fit for the study.

Because we were not interested in talking to people who were unemployed, students, or retired, this page break allowed us to pass them to a message that said something that indicated that we appreciated their interest but they probably were not a good fit for the study at that point.

From a job role perspective, this could have been a much longer list. We only included job roles we thought were relevant to the study and the people we were reaching out to, though we acknowledged that some people may have different roles which is why we added the "other" option.

This is an example of a perceived behavior question. It's not written in a way that there is a right or wrong answer, and it's not something that we had data for.

But something we cared about was recruiting people who spent different amounts of time writing code, so this was meant to ask about their perception of their behaviors (not the actual number of hours — who tracks that?).

This question is an example of a unipolar scale — a scale that goes from nothing to something. This is in contrast to a bipolar scale which goes in two dimensions (typically very negative to very positive).

Unipolar scales are best used when you are asking about something that does not need to be expressed in multiple directions (i.e. “How many emails did you send this morning?” because you cannot send negative emails). Bipolar scales are best used when you want to evaluate the full spectrum of emotion and understand the degree to which something adds or detracts from a product, experience, etc (i.e. “How satisfied or dissatisfied were you with your recent stay at the hotel?” where the scales go from “very dissatisfied” to “very satisfied”).

The reason this question includes bolded formatting is because it looks VERY similar to the last one, and if someone is reading quickly they may not even notice. However, the bolding draws attention to that question, where the participant ideally recognizes that the previous question was about them and this question is about their team.

Just because they filled out all of the above, doesn't mean that they want to participate in research with you.

In some cases, you could use this question to have them provide times that they would be available or indicate their availability from a set of options.

Because we wanted to set the expectation that we would be reviewing their answers and then reaching out, we did not have a way for them to indicate availability here.

This can be a very helpful question to ask because it opens up the space for the participant to share preferences, needs, concerns, etc that you should take into account.

Behzod Sirjani

I wrote this screening survey for an early stage company who was conducting research on developer workflows to inform a tool they were building.

[PAGE DESCRIPTION TEXT] Our team is conducting interviews with people who write code and build software as a part of their professional responsibilities, even if you are not an engineer. We'd love to learn more about you to determine if you are a fit for this research. The following questions should take less than 2 minutes to complete.

[q_email_oe] What's the best email to contact you:

 Share a link to your Twitter account (if you have one):

[q_github_oe] Share a link to your Github account (if you have one):

[q_linkedin_oe] Share a link to your LinkedIn account (if you have one):

Which of the following best describes your current employment

I own a business and/or am self-employed.

I do freelance or contract work for one or multiple employers.

About how many people work at your company?

[q_money] What is the primary way your company makes money?

Selling technical and/or strategic services

[PAGE DESCRIPTION TEXT] Tell us a bit about the work you do at your company and the software tools you use.

Which of the following best describes your role?

About how frequently have you written code in the last few months?

[q_tool_oe] When you're writing code, what tool are you using the most?

In general, how much influence, if any, do you have in choosing the tools that you use for work?

In general, how much influence, if any, do you have in choosing the tools that your team uses for work?

[PAGE DESCRIPTION TEXT] Please let us know if you are interested in sharing more feedback with us.

 We are looking to speak with a handful of people about the way that they write code and/or build software, so that we can improve our own products.

Are you interested in a member of [REDACTED] reaching out to you to participate in an interview in the coming weeks?

User Interviews Screening Survey at Stealth Startup

[LANDING PAGE]

[START BUTTON]

Sign up to see more