These prompts help researchers use AI at specific stages of service blueprinting: synthesizing employee interviews into process maps, aligning customer experience data with backstage operations, analyzing support ticket data for fail point patterns, and drafting the final blueprint document. Each prompt is designed to be pasted into ChatGPT, Claude, or another LLM with your data.
Prompt 1: Employee interview synthesis for blueprinting
I have conducted [N] interviews with employees involved in delivering [service name]. The roles interviewed include [list roles: support agent, operations manager, billing specialist, etc.].
[Paste transcripts or key excerpts, labeled by role]
Extract service blueprint data:
1. For each role, list:
- Process steps they perform (in sequence)
- Systems and tools they use at each step
- Information they need and where they get it
- Handoff points: who do they pass work to, and how?
- Pain points: where does the process break, slow down, or require workarounds?
2. Identify steps that are visible to the customer (frontstage) vs. invisible (backstage)
3. Map dependencies between roles — which steps cannot start until another role completes their work?
4. Flag any contradictions between employees' descriptions of the same process
5. Note any informal workarounds that are not documented in official procedures
Format as a table: Step | Role | Frontstage/Backstage | Systems Used | Handoff To | Pain PointsPrompt 2: Customer-to-operations alignment
I have two datasets for the same service:
1. Customer journey data showing touchpoints and pain points
2. Employee process data showing backstage operations
Customer journey:
[Paste journey stages, touchpoints, and pain points]
Employee process data:
[Paste the extraction from Prompt 1]
Align these two views:
1. For each customer touchpoint, identify the backstage process that supports it
2. For each customer pain point, trace the operational root cause: which backstage step, handoff, or system limitation produces the frontstage friction?
3. Identify "orphan" backstage processes — work that employees do that doesn't connect to any customer touchpoint (potential waste)
4. Identify "unsupported" customer touchpoints — moments where the customer expects service but no backstage process exists to deliver it
5. For wait points: where the customer waits, what is happening backstage during that wait?
Format the alignment as a service blueprint table: Physical Evidence | Customer Action | Frontstage Action | Backstage Action | Support ProcessPrompt 3: Fail point analysis from support data
Here is support ticket data for [service name] over the past [time period]:
[Paste support ticket categories, frequencies, and example tickets — or describe the data structure]
Analyze this data for service blueprint fail points:
1. Group tickets by the service process step where the failure originates (not where the customer reports it — trace back to the root cause if possible)
2. For each failure cluster, estimate: frequency (tickets per week/month), severity (service blocked vs. degraded vs. cosmetic), and customer impact (churn risk, revenue loss, satisfaction drop)
3. Identify systemic patterns: failures that recur despite being "fixed," failures that spike at specific times (month-end, releases, holidays), failures that cluster around specific handoff points
4. Rank the top 10 fail points by combined frequency × severity
5. For each top fail point, suggest where it belongs in the blueprint: which backstage process or support system is the root cause?Prompt 4: Blueprint documentation drafting
I ran a service blueprinting workshop for [service name]. Here are the workshop outputs:
[Paste or describe: customer actions, frontstage actions, backstage actions, support processes, fail points, wait points — from whiteboard photos, notes, or recordings]
Draft a clean service blueprint document:
1. Organize into five swim lanes: Physical Evidence | Customer Actions | Frontstage Employee Actions | Backstage Employee Actions | Support Processes
2. Draw three boundary lines: Line of Interaction, Line of Visibility, Line of Internal Interaction
3. For each column (process step), fill all five rows
4. Mark fail points with [F] and include: description, root cause, frequency, impact
5. Mark wait points with [W] and include: customer wait duration, what happens backstage during the wait
6. List all handoffs between roles or systems, with the information passed and the risk of information loss
7. Summarize: top 5 fail points ranked by impact, top 3 wait points, and recommended priority improvements
Format as a structured document that can be converted into a visual diagram.