Running a Diagnostic helps you understand what is contributing to a business outcome and decide what to do next. This guide covers how to choose your question, plan participation, and use the findings to make changes interviewees can see. For the product overview and access, see New User Onboarding for Running a Diagnostic.

Start by agreeing with the team leading the project on what you need to understand and how the answer will be used. You might investigate a performance gap, learn why one team succeeds, check a hypothesis, or understand a risk before committing resources.

Ask whether understanding the cause would materially change your next steps. A clear purpose helps you make good use of interviewees’ time.

For example, a category with a $1.3 million net sales gap against plan has a specific problem to explore: what is contributing to the shortfall, and what could help address it? A request to “drive better efficiency and communication” needs more detail about where the difficulty occurs and what you hope to change.

You can refine the objective through the setup conversation. It asks pointed questions to help you clarify the problem and develop useful interview questions. See Setting Up a Diagnostic.

Stay open to what the findings reveal. A performance problem may involve skills, processes, responsibilities, or several factors together.

Work with the group’s direct leader to introduce the Diagnostic and explain why their input matters. Arist’s rollout experience has seen approximately 70 to 77% participation with leader-led launches, compared with 5 to 10% through cross-functional outreach.

Agree on who will participate, your participation target, and whether participation is required or voluntary. As a planning guideline, findings below roughly 40 to 50% participation should be presented as directional. Higher participation still needs to be reviewed for gaps between groups.

Build the rollout around four things:

Allow around two weeks for interviews and several days for analysis and the readout. Explain anonymity accurately, based on your project setup, and share Taking an Arist Interview so interviewees know what to expect.

Test interviewee and administrator access using real accounts on your corporate network. Complete the experience, including starting an interview. See Diagnostic - Firewall Testing.

Keep each project focused on one business question. Around four questions per cohort is a useful starting point for a five to seven minute conversation.

Give the setup conversation relevant organizational data, such as strategy, previous research, benchmarks, or performance data. Ask it to suggest example questions, then review whether those questions will help you understand the issue.

“Walk me through the last sale that went really well” gives interviewees a specific experience to discuss. Use clear, neutral wording, explain unfamiliar terms, and avoid assuming something is wrong.

Keep core questions consistent where you want to compare groups, while adapting questions to their responsibilities. End by asking which one or two things interviewees could improve that would most help their performance.

Test the questions with 10 to 15 people across the groups before the wider launch. Have the person coordinating the project try every cohort’s interview.

Once available, choose the depth that fits your topic:

Depth is separate from interview length. A hard time limit caps the amount of follow-up available.

Read the shortest and incomplete transcripts during the first 48 to 72 hours. These can show where the introduction, questions, access, or expectations need attention.

Track invitations, starts, completions, completion rate, and interview duration by cohort. Review the substance of the answers too. Even full participation can produce limited insight if interviews are very short or answers lack detail.

Agree on how often to share updates based on the audience, the collection window, and participation. Use those updates to highlight whose input is still needed and coordinate any follow-up.

Keep collection open for the agreed window. Concluding interviews early can leave out interviewees who still intend to contribute. If participation remains below your agreed threshold, make the limits clear when presenting the findings.

See Collecting Diagnostic Interviews for guidance on distributing interviews and monitoring progress.

Review the report and its recommendations with relevant stakeholders. Discuss what the evidence supports, what context needs to be considered, and which actions could have the greatest impact.

Use that discussion to agree on:

Include strengths the organization can build on alongside areas that need attention. Present the findings live so the team can ask questions and agree on next steps. See Reviewing Diagnostic Findings and Reports.

Keep interviewees informed about what you learned and what happens as a result. Plan for an initial summary and later updates as actions progress.

Choose the timing, format, and delivery channel to suit your audience, and agree on who will coordinate these updates.