Documentation

Studies

Studies allow you to send targeted surveys to specific audiences in order to gather rich, focused feedback on particular topics—whether it’s evaluating a recent initiative, validating a hypothesis, or understanding team sentiment in more detail. Use Studies to go beyond your regular snapshots and engage the right people with the right questions, at the right time.

Unlike traditional survey tools, Studies are delivered as personal direct messages from you—not from a bot. This design drives significantly higher response rates and helps DevEx teams build authentic relationships with developers, rather than adding to notification noise.

Study results

Creating a study

To create a new study:

Navigate to DX → Tools → Studies.
Click the + Study button to begin.

The study builder will walk you through 5 setup steps:

Step 1: Basic study details

Fill in the foundational information:

Study Name: Internal label used for tracking and organizing your studies.
Sender: The person the study will be sent from as a direct message (usually you). Studies are delivered as personal DMs because developers are far more likely to engage with a message from a real colleague than from a bot.
Message: A short introduction or prompt that accompanies the study in Slack/Teams. Keep it conversational and relevant—this is the first thing recipients will see.

Step 2: Add questions

Choose the questions you want to ask:

Add as many questions as needed (see supported types below).
You can reorder or edit questions at any time before sending.
Optionally, configure a reminder to nudge non-responders after a few days.
Click Save & Continue when done.

Supported question types

Question Type	Description
Short text	Single line text input
Long text	Multi-line text input
Multiple choice	Single selection from options
Rating scale	Numeric rating selection
Multi-select	Multiple selections from options

Dynamic question visibility

Studies supports dynamic question visibility based on previous answers. To dynamically show questions, click the “dynamic visibility” icon on the right of the question you want to display conditionally, and select the question you want to depend on. From there, you will be able to select the condition you want to trigger question visibility with.

Step 3: Configure target audience

There are two options for choosing who receives the study:

Dynamic Targeting (recommended)

Automatically target users based on filters:

Teams: Choose all teams or specific ones.
Attributes: Filter by role, location, tenure, etc.
Link to Snapshot Segment: Use filters or segments from your latest snapshot.
Coding Activity: Filter by users who write code, don’t write code, or both.

Manual Targeting

Upload a list of recipient email addresses (CSV format). This is useful for targeting specific individuals across teams or roles.

Targeting Options (applies to both methods)

Exclude recent participants: Avoid survey fatigue by skipping users who recently completed a study.
Exclude users marked Out of Office (OOO): Ensure you’re reaching active participants.

Step 4: Limit audience

To reduce fatigue and improve response quality, you can limit how many people receive the study:

Sample users (recommended): DX will randomly select a representative sample of your target audience.
Do not limit: Send the study to everyone in the target audience (may reduce effectiveness if overused).

Step 5: Finalize study

Review your study setup. When you’re ready:

Choose Send now to immediately deliver the study.
Choose Schedule send to save the study and send at a future date automatically.
Or select Save & exit to save without sending or scheduling.

Once sent, you’ll be able to monitor responses and download results from the Responses tab.

Designing high quality studies

Follow these practical guidelines to maximize the effectiveness of your studies and ensure high-quality, actionable responses.

Step 1: Choose your approach

Pick the survey method that fits your goal best:

Qualitative (open-text): Detailed stories, “why/how” questions.
- Audience: Start small (≈ 6–12); add more only if you’re not hearing new things.
- Question types: Short text, Long text (optionally add 1–2 multiple-choice items for context).
- Analysis: Read for themes; pull quotes; summarize top 3–5 patterns.
Quantitative (ratings/choice): Measure or compare trends numerically.
- Audience: Use a larger sample; consider sending to your full target group.
- Question types: Rating scale, Multiple choice, Multi-select.
- Analysis: Look at distributions/averages and compare segments (e.g., team, role, tenure).
Hybrid: Combine quantitative measurement with brief qualitative context.

Step 2: Set your audience size

Based on your chosen survey type:

Method	Recommended participants
Qualitative or Hybrid	Start with 6–12, expand only if necessary
Quantitative	Larger sample or full target group

If targeting multiple groups (e.g., teams or roles), aim for 6–8 per subgroup, or run multiple smaller studies rather than one large one.

Start small and expand if necessary

When asking open-ended (qualitative) questions, begin with a smaller group and only add more participants if necessary:

Start with around 6–12 participants for most topics.
Review responses for useful insights.
Expand gradually only if you need more detailed information.

Study Type	Recommended Participants
Focused topic or single-team feedback	6–12
Broader team or organization-wide insights	16–24+

These numbers are a general guide to get useful insights without overwhelming participants. Adjust as needed based on your organization’s size.

Step 3: Design clear, effective questions

Keep studies brief: aim for 5 or fewer questions.
Ask simple, single-focus questions.
Favor shorter formats (multiple-choice, short text, or rating scales) for higher response rates.

Step 4: Send thoughtfully and boost responses

Send studies during local work hours (Tue–Thu, late morning or early afternoon).
Write a brief, friendly introduction clearly explaining the purpose.
Send one gentle reminder a few days later if needed.

Step 5: Avoid over-surveying

Space studies out—avoid repeatedly surveying the same participants.
Use DX’s sampling and exclusion options to manage your studies frequency effectively.

Common questions

Why are Studies sent as direct messages instead of from a bot?

Studies are intentionally delivered as direct messages from you—not from a bot or app—because this design makes them significantly more effective.

When developers receive a DM from a real person on their DevEx team, it signals a genuine conversation rather than automated noise. This personal approach:

Drives higher response rates: People are far more likely to respond to a message from a colleague than from a bot notification.
Establishes authentic connection: DevEx teams need direct lines of communication with developers. Studies help build that relationship.
Avoids survey fatigue: Bot-delivered surveys quickly become background noise that gets ignored or dismissed as spam.

This is a core design principle of Studies—not a limitation. Effective developer experience work requires DevEx practitioners to be comfortable reaching out directly to developers, and Studies supports that by making those conversations feel natural and personal.