Discovery Call Questions for Automation Projects (54 Questions by Category)

A practical field guide from Automation Ace.

The short answer

A good automation discovery call covers eight areas: process, data, people, tools, volume and timing, exceptions, compliance, and success metrics. Below are 54 questions grouped by category. You will not ask them all on every call; pick the ones that fit, listen for workarounds and exceptions, and leave with enough to map the process and score its readiness.

Useful for consultants, forward deployed engineers, and in-house ops teams. Part of the automation planning playbook.

How to use this question bank

• Send a short pre-call form with the basics (process name, tools, volume) so the call goes deeper. The scoping worksheet works well.
• Ask open questions first, then specific ones.
• Ask to see the work on screen; showing beats describing.
• Record answers straight into your exception inventory, systems inventory, and metrics baseline.

Process questions

1. What starts this process, and what does “done” look like?
2. Walk me through the last time you did it, step by step.
3. Which steps are manual today?
4. Where does work wait, and for how long?
5. Who hands work to whom?
6. Which steps require judgment rather than rules?
7. Is the process documented? Is the documentation accurate?
8. How often does the process change?

Data questions

1. What information does each step need, and where does it come from?
2. Which fields are often missing, wrong, or inconsistent?
3. Is there a unique ID that matches records across systems?
4. Which system is the source of truth for each record type?
5. Are there spreadsheets acting as databases?
6. What formats are dates, currencies, and names in?
7. Do you have historical data to backfill or migrate?

People questions

1. Who owns the process and can approve changes?
2. Who does the work day to day?
3. Who receives the output, and what do they do with it?
4. Who needs to be notified, and how (email, Slack, SMS)?
5. Who handles exceptions today?
6. How comfortable is the team with change?
7. Who will own the automation after launch?

Tools and systems questions

1. Which apps are involved, and who administers them?
2. What plan tier is each on? Do they include API or webhook access?
3. Do native Zapier, Make, or other integrations exist for these apps?
4. Are any apps shared across multiple accounts or workspaces?
5. What automations already exist, and who built them?
6. How do you authenticate to each system?
7. Are there rate limits or usage caps we should know about?

Volume and timing questions

1. How many times does this happen per day, week, or month?
2. Are there peaks (month end, campaigns, seasons)?
3. How fast must the automation respond?
4. Are there business hours or time zone constraints?
5. How long can a step wait (hours, days, weeks)?
6. What happens if it runs late?

Exceptions questions

1. What goes wrong most often?
2. Tell me about the last time this went badly.
3. What happens when data is missing?
4. Can the same item arrive twice?
5. Do people manually override what the system does?
6. What partial or conditional approvals happen?
7. Which exceptions must a human always handle?

Compliance and risk questions

1. Does the process involve personal, financial, or health data?
2. Which regulations or contracts apply?
3. Who may see which data?
4. How long must data be retained, and where?
5. Are audit logs required?
6. Can AI tools process this data under your policies?

Success metrics questions

1. How long does this take today, per item and in total?
2. What is the current error or rework rate?
3. What is the current turnaround time?
4. What does this cost today in hours or money?
5. What would make this project a clear success in 90 days?
6. How will we measure it, and who will report it?

After the call

1. Map the process: process mapping workshop.
2. Score readiness: readiness assessment.
3. Inventory systems: systems and data inventory.
4. Log exceptions: exception inventory.
5. Capture the baseline: success metrics.
6. Prioritize and spec: impact vs effort and the requirements doc.

What to listen for

• “Usually” and “it depends”: hidden exceptions.
• “I just copy it over”: prime automation targets.
• “Only Sam knows how”: undocumented rules and key-person risk.
• “We tried automating this before”: ask why it failed. See why automations break.

Frequently asked questions

What questions should I ask on an automation discovery call?

Cover process, data, people, tools and systems, volume and timing, exceptions, compliance and risk, and success metrics. Ask how work is done today, where it waits, what goes wrong, which systems are involved, and how success will be measured.

How do I prepare for an automation discovery call?

Send a short pre-call form with the process name, tools, and volume, pick the most relevant questions, and ask the client to share their screen to show the real process.

What should I listen for during automation discovery?

Phrases like 'usually' or 'it depends' signal exceptions, 'I just copy it over' signals automation targets, 'only one person knows' signals key-person risk, and past failed attempts reveal pitfalls.

What happens after a discovery call?

Map the process, score readiness, inventory systems and data, log exceptions, capture a baseline, prioritize candidates, and write a requirements doc for sign-off.