Quick Introductions — who do we need to wow (and why)?

• Who should be our main point of contact for coordinating the pilot? Options: VP Contact Center Operations, Head of Customer Care Technology, Director of Quality, IT/Integration Lead, Procurement Lead, Other (please name)
• Who else will be directly involved day-to-day (list names, roles, and title/department)?
• Which of these best describes your role in the decision? (pick one) Options: Primary decision maker, Influencer / evaluator, Technical approver, Budget approver, End-user representative, Other
• How have you typically run pilot evaluations in the past? Options: Short technical POC, Operational pilot with agents live, Vendor-led pilot with metrics, Proof-of-value with finance sign-off, We haven't run pilots before, Other
• If you've run pilots before, what one thing usually makes them succeed or fail? Tell a quick example.

Why are we still tolerating manual workarounds?

• When audits find agents skipping required disclosures at 18% (or similar), what's your immediate emotional response—annoyed, worried, resigned, motivated to act? Options: Worried (regulatory risk), Angry (process failure), Resigned (accepted cost), Motivated to change, Other
• How often do those compliance or escalation faults occur across your contact center? Options: <5% of calls, 5–10% of calls, 10–20% of calls, 20–30% of calls, >30% of calls, Don't have reliable data
• Give a recent example of a call or interaction where the manual workflow led to a failure or long handle time—what happened and what was the impact?
• Which parts of the existing agent workflow generate the most manual work or errors? Options: Post-call documentation, Knowledge base lookup, Policy exceptions, Call routing mismatches, CRM data entry, Other
• How much of a problem is limited quality monitoring (e.g., <2% sample) for you—is it 'we need better visibility' or 'we're blind to risk'? Options: Need better visibility, We're blind to material risk, We rely on other controls, Not a current concern

Who's really holding the keys (and the veto)?

• If someone said 'go' on this pilot today, who would need to sign off before work actually starts? Options: Operations leader, IT/integration, Security/compliance, Legal, Finance/Procurement, HR (training), Other
• What are the non-negotiable concerns those approvers will raise (for example: data residency, SLA levels, auditability)? Options: Data residency, Privacy consent, SLA for intervention latency, Audit trail granularity, Third-party vendor review, No production CRM changes, Other
• Who owns the pilot budget and how is funding typically released for a 60–90 day operational pilot? Options: Central IT budget, Operations budget (contact center), Innovation / transformation fund, Line of business P&L, Vendor credit/PO model, Unknown / need to discover
• What evidence or artifact will convince each stakeholder (e.g., legal needs DPIA, Ops needs AHT drop, Finance needs ROI model)? Please map stakeholder → proof required.
• What's your normal procurement timing—from statement of work to execution—once stakeholders agree? Options: <2 weeks, 2–4 weeks, 1–2 months, 2–3 months, >3 months, Varies widely

If the pilot moved the needle, what would actually change for your team?

• Which of these outcomes would make you count the pilot as a clear win? Options: AHT reduction, Resolution rate improvement, Quality score improvement, Reduction in compliance failures, Supervisor review time reduced, Agent satisfaction increase, Other
• For each outcome you selected, what's a realistic target for a 60–90 day pilot (give percentages or absolute targets)?
• Which KPIs are already tracked and trusted in your dashboards (and which are currently guessed)? Options: AHT, First Contact Resolution, Quality Score, Compliance failure rate, Supervisor review time, NPS/CSAT, Other
• What data sources will we rely on to measure those KPIs (telephony, CRM, QA tool, workforce management)? Options: Telephony provider, CRM (name), Third-party QA tool, Workforce management, Analytics warehouse, Other
• How will you decide to expand beyond the pilot—who signs off to scale, and what threshold must be met? Options: Operations leader approval, Finance ROI threshold, Legal/compliance sign-off, All of the above, Other

What will break us before we even start?

• What's been the hardest technical or organizational blocker in prior integrations (e.g., legacy telephony, restricted CRM APIs, no real-time hooks)? Options: Legacy PBX with no API, CRM without softphone integration, Restricted security policies, Lack of dedicated IT resource, Data residency constraints, Other
• Do you currently have a process for real-time data exchange (websocket/CTI) or would that need to be created? Options: Existing real-time integration, Partial capabilities (batch only), None — needs build, Don't know
• How sensitive are the calls/records involved—PII, PHI, financial data—and what controls must be in place to use them in a pilot? Options: PII (names/phones), PHI (health info), Financial data, No sensitive data, Unknown
• Agent trust can kill adoption. What are your biggest fears about agents feeling overridden or penalized by real-time prompts? Options: Agents ignore prompts, Prompts increase cognitive load, Managers use flags punitively, False positives create more work, Other
• What controls or guardrails would you insist on to protect agent autonomy and prevent false positive noise? Options: Supervisor review workflow, Confidence thresholds, Silent mode monitoring first, Agent feedback loop, Rollback option, Other

Pilot team, playbook, and who actually does the work

• If we ran the pilot as a production slice, who will be the pilot manager responsible for day-to-day decisions? Options: Operations manager, QA manager, Project manager, IT integration lead, Vendor partnership manager, Other
• Which of these roles can commit time during the 60–90 day window (select all that apply)? Options: Dedicated agent cohort, Team supervisor(s), QA reviewers, IT/integration engineer, Analytics owner, Training lead
• What training or enablement would make agents feel supported (choose all that apply)? Options: Live walkthroughs, Short micro-learning modules, Shadowing sessions, In-call coach support, Incentives for participation, Other
• Who will own monitoring and daily dashboard reviews during the pilot? Options: Ops leader, QA team, Analytics team, Vendor-managed dashboard, Shared responsibility
• Please outline the three most important runbook items or escalation steps we must include before going live (roles, timing, and triggers).

Commitment and timeline — can we actually move fast enough?

• If we proposed a 60–90 day operational pilot that starts with one team and one use case, how ready is your organization to begin in the next 30 days? Options: Ready now, Ready in 2–4 weeks, Ready in 1–2 months, Not ready this quarter, Unsure
• What are the non-negotiable legal or procurement steps that will determine your earliest possible start date? Options: Standard PO, Master services agreement, Data processing addendum, Security questionnaire, Third-party risk review, Other
• Which of these pre-deployment items are already available or can be provisioned quickly? Options: CTI access, CRM test tenant, QA scoring rubric, Sample call recordings, Agent cohort list, None of the above
• Who must be present in a weekly pilot steering meeting and how long can they commit each week? Options: 15 minutes, 30 minutes, 60 minutes, Only ad-hoc
• What would cause you to pause or stop the pilot early? List hard stop conditions (metrics or events).

Lessons from past pilots — what would you change if you could rewind?

• Think about the last pilot you ran: what surprised you in a way that changed your view of pilots overall?
• Which of these common pilot failures have you seen before? Options: Insufficient measurement plan, Stakeholder fade-out, Integration took too long, Agents rejected the experience, Data quality issues, No plan to scale
• If you could design a pilot that guaranteed learnings even if it didn't improve KPIs, what would that look like?
• What does success look like from an organizational change perspective (not just metrics)—for example, 'ops trusts AI-assisted prompts' or 'supervisors save X hours/week'?
• What follow-up cadence would keep stakeholders engaged after the pilot—demo, weekly metrics, executive summary, or something else? Options: Weekly metrics review, Bi-weekly steering committee, Mid-pilot demo, Final executive readout, Ad-hoc as issues arise, Other

Let's map a day in the life of your agents

• Walk me through a typical agent’s shift from first login to final wrap-up—what tasks take up the most time?
• Which desktop applications does an agent actively use during an average call? Options: CRM (Salesforce, Zendesk, etc.), Telephony softphone, Knowledge base, WFM/scheduler, Case management, Email clients, Proprietary tools
• On average, how many different screens or systems must an agent use to fully resolve a typical customer issue? Options: 1, 2–3, 4–6, 7+
• How much of an agent’s after-call work (ACW) is manual data entry versus automated documentation? Options: Mostly manual (>75%), About half/half, Mostly automated (>75%), We don’t track this
• Who today is responsible for updating the CRM and closing the loop after a call (agent, supervisor, centralized team)? Options: Agent, Supervisor, Centralized back-office, Automated process, Varies by team
• Tell me about a recent call where the agent clearly struggled—what systems or steps contributed to that struggle?

Is the process itself quietly causing compliance lapses?

• Where in your current call flow do required disclosures or compliance checks most often get missed? Options: Call opening, Product explanation, Escalation/transfer, Closure/consent, Post-call documentation, Other
• How often do your audits find agents skipping required disclosures or steps? Options: <1% of reviewed calls, 1–5%, 6–10%, 11–20%, >20%
• What are the top three root causes you see when compliance fails (e.g., too many steps, poor prompts, knowledge gaps)? Options: Cognitive overload / too many steps, Unclear scripts/policies, Poor KB search, Tool latency or crashes, Inadequate training, Agent shortcuts to save time, Other
• When a compliance failure is discovered, how is it typically documented and who reviews it? Options: QA team documents; supervisors review, Automated audit flags; QA validates, Manual notes in CRM; no formal review, Compliance/legal team reviews, Other
• Can you give a specific example of a compliance audit finding and what you discovered caused it?

Who pays the price when things go wrong—and how badly?

• What regulatory, financial, or customer-experience consequences have arisen from missed compliance items in the past 12 months? Options: Regulatory fines, Customer churn, Escalations to legal, Reputational damage, Internal retraining costs, None observed
• How often do compliance misses trigger formal escalations (legal, leadership, regulators)? Options: Never, Rarely (quarterly or less), Occasionally (monthly), Frequently (weekly)
• When an incident occurs, which teams must coordinate to respond and how long does that coordination typically take?
• How does a single compliance incident translate to operational cost in your view—supervisor time, rework, fines, customer impact?
• Who internally becomes the de facto owner of mitigating future incidents—QA, operations, compliance, or IT—and how effective has that been? Options: QA, Operations, Compliance/legal, IT/engineering, Cross-functional task force

How much of the truth does your quality program actually see?

• What percentage of interactions are sampled for quality today and how was that sampling rate decided? Options: <1%, 1–2%, 3–5%, 6–10%, >10%
• Do you use rule-based sampling, risk-based targeting, 100% automated QA, or a mix? Tell me which and why. Options: Random sampling, Risk-based (escalations, transfers), Keyword/trigger-based, 100% automated metrics, Mixed approach
• How long after an interaction do QA findings typically return to the agent for coaching? Options: Within the same day, 1–3 days, 4–7 days, More than a week, No formal feedback loop
• What proportion of QA flags are false positives or require manual disambiguation? Options: <5%, 5–15%, 16–30%, >30%, Unknown
• Describe a recurring example of a QA blind spot (e.g., tone, multi-step policy exceptions) and how it impacts decisions.

If a nudge could appear during the call, would agents welcome it—or resist it?

• How do agents currently receive prompts or coaching during live calls (desktop pop-ups, supervisor whisper, none)? Options: Desktop prompts, Supervisor whisper, Post-call coaching only, Chat messages, None
• Have you run any experiments with real-time guidance before? What happened and how did agents react?
• What percentage of agents trust automated suggestions today, and what would increase or decrease that trust? Options: Moderate trust, High trust, Low trust, No trust measure
• How often do supervisors override an automated or suggested action because it was wrong? Options: Never, Rarely, Sometimes, Often, Don’t track
• What would an acceptable false-positive rate look like for you in a pilot focused on compliance prompts? Options: <1%, 1–5%, 6–10%, 11–20%, Unsure

Where do your systems create friction instead of flow?

• Which backend systems are in scope for integration during a pilot (telephony vendor, CRM, KB, workforce management)? Options: Avaya/Genesys/Other telephony, Salesforce, Zendesk, Custom CRM, Knowledge base (Confluence/KB), WFM, Other
• Do those systems expose APIs or real-time events we can rely on? If some do and some don’t, which are which? Options: All have real-time APIs, Partial APIs (some do, some don’t), Mostly batch-only interfaces, Legacy systems with no APIs
• How long does an integration project typically take and who is required to support it (internal team, vendor, consultants)? Options: <2 weeks, 2–6 weeks, 6–12 weeks, >12 weeks
• Where have you hit the hardest technical barriers—field mapping, single sign-on, latency, or something else? Options: Field mapping/CRM schema, SSO/authentication, Real-time latency, Call metadata availability, Security/compliance approvals
• What change-control or security approvals would a vendor need to complete before a pilot could access production telephony or CRM data? Options: Security review, Data processing agreement, Pen test, Privacy review, Legal/compliance signoff, None required

Are leaders ready to trade saved time for reduced risk?

• Who are the decision-makers and influencers for a pilot like this (titles and functions)?
• Which KPIs do those stakeholders care about most when approving projects (AHT, NPS, quality score, FCR, compliance rate)? Options: Average Handle Time (AHT), First Contact Resolution (FCR), Quality score, Compliance rate, Customer Satisfaction/NPS, Operational cost
• What thresholds or improvements would be sufficient to justify moving from pilot to wider rollout?
• How quickly does procurement and legal typically move on vendor pilots—weeks or months—and what are the usual blockers? Options: <2 weeks, 2–6 weeks, 6–12 weeks, >12 weeks
• Which internal team would own the pilot once launched (operations, IT, vendor-managed), and who would be the day-to-day contact? Options: Operations, IT/Engineering, QA/Quality, Vendor-managed, Cross-functional

Which metrics keep you awake at night—and which ones are realistic to change quickly?

• What are your current baseline values for AHT, FCR, quality score, and compliance rate?
• Which of these metrics is business-critical in the next 60–90 days? Options: AHT, FCR, Quality score, Compliance rate, Customer satisfaction (CSAT/NPS), Cost per contact
• How do you currently attribute improvements to a specific change versus other seasonal factors? Options: A/B testing, Pre/post comparison, Regression analysis, We don’t have a formal method
• What minimal absolute improvement on a chosen KPI would convince leadership to expand a pilot? Options: 1–3% improvement, 4–7% improvement, 8–15% improvement, >15% improvement, Unsure
• Which reporting cadence and dashboard views are required for stakeholders during a 60–90 day pilot? Options: Daily summary, Weekly deep-dive, Executive monthly, Real-time dashboard, Ad hoc on request

What would a sensible, low-risk pilot actually look like?

• If we scope a pilot to one team and one use case, which team and use case would you pick first? Options: Inbound billing support, Technical troubleshooting, Retention/Winback, New product onboarding, Claims/eligibility
• How many agents would you be comfortable including in an initial pilot? Options: 5–10, 11–25, 26–50, 51–100
• What are the non-negotiable acceptance criteria for the pilot to be considered successful?
• Which internal resources can you commit to support the pilot (integration engineer, QA reviewer, ops lead, trainer)? Options: Integration engineer, QA reviewer, Ops lead, Trainer, Data analyst
• What minimal data or sample do we need up front to run a meaningful 60–90 day test? Options: Call recordings + metadata, CRM records + fields, Knowledge-base articles, Agent desktop screenshots, All of the above

From discovery to decision: what would make this easy?

• Reflecting on everything, what single barrier would you want us to remove to make a pilot possible within the next quarter?
• How would you prefer we demonstrate quick value—through a short AHT win, a compliance improvement, or QA workload reduction? Options: AHT reduction, Compliance improvement, QA efficiency / fewer false positives, Agent satisfaction
• What would success look and feel like to agents, supervisors, and executives at the end of a 90-day pilot?
• Who should be in the room for a pilot kickoff so decisions can be made quickly? Options: Ops manager, QA lead, IT/Integration, Security/compliance, Vendor lead, Training lead
• Finally, what concerns—technical, cultural, or political—should we address up front to avoid surprises?

Quick Check-In: Where This Pilot Fits

• Which team and use case are we piloting first? Options: Customer service voice (inbound), Customer service voice (outbound), Collections, Technical support, Sales enablement, Omnichannel (voice + digital), Other
• How many agents will be in the pilot group? Options: 1-10, 11-50, 51-100, 101-250, 250+
• Who is the executive sponsor and who will be the day-to-day owner from your side?
• Which single metric matters most to you for go/no-go at the end of the pilot? Options: Average Handle Time (AHT), Resolution Rate / First Contact Resolution, Quality Score / Compliance, Agent Adoption / Trust, After-Call Work time, Other
• When would you realistically like the pilot to be live (month/quarter)? Options: Within 2 weeks, Within 1 month, 1–2 months, 2–3 months, Unsure

Why Hit ‘Deploy’ When the Foundations Feel Wobbly?

• What internal pressures are driving the chosen pilot timeline—compliance audit windows, executive deadlines, cost targets, or something else? Options: Compliance audit, Executive mandate, Cost savings target, Product launch support, Vendor contract milestone, Other
• Have you attempted a similar integration or pilot in the past and run into unexpected blockers? Tell us one story.
• If we had to pause the pilot mid-flight, what would be the single worst operational consequence for your organization? Options: Regulatory exposure, Customer experience degradation, Agent morale drop, Wasted budget, Escalation to legal/execs, Other
• Who across your organization needs to agree before we start and how aligned are they today? Options: Everyone aligned, Mostly aligned with a few questions, Significant misalignment, We haven't engaged all stakeholders
• If you had to rank the pilot risks in order of pain—integrations, privacy, agent trust, false positives—how would you order them and why? Options: Integrations first, Privacy first, Agent trust first, False positives first, Other order

Can Your Telephony and CRM Actually Speak Fast Enough?

• Which telephony platform(s) handle the agent desktops and CTI today? Options: Avaya, Genesys, Cisco UCCE/CCE, Five9, Amazon Connect, In-house/Custom, Other
• Which CRM(s) will need live integration for the pilot? Options: Salesforce, Zendesk, ServiceNow, Microsoft Dynamics, Oracle CX, Custom/Proprietary, Other
• How are integrations currently implemented between telephony and CRM (CTI, APIs, screen-scrape, middleware)? Options: Modern APIs/CTI, Middleware / ESB, Screen-scrape/browser plug-in, Batch sync only, Not sure / needs discovery
• What latency or performance limits would break the agent experience (e.g., >500ms, >1s, >3s)? Options: <250ms, 250–500ms, 500ms–1s, 1–3s, >3s
• Who owns telephony and CRM configuration on your side and who will be the integration point-of-contact?

Where Is Your Data Leaving Gaps or Raising Eyebrows?

• Which specific data fields are critical for real-time mapping (e.g., customer ID, product codes, account status, consent flags)?
• Do you have an existing data mapping or canonical field dictionary we can use as a starting point? Options: Yes, complete and current, Yes, but incomplete/outdated, No, but available ad-hoc, No formal mapping exists
• How would you describe the cleanliness and reliability of the key data feeds we’ll need? Options: Very clean, Mostly clean with some edge cases, Messy with frequent exceptions, Unknown / needs assessment
• Are there known data sources that will require normalization or enrichment before they can be used in real time? Give examples.
• Which identifiers can we use to stitch voice events to CRM records (e.g., ANI/CLI, session ID, user login)? Options: ANI/CLI, Session ID, Agent desktop ID, Customer email/phone, Custom correlation field, Not sure

Who Decides What Is a Compliance ‘Hit’ — And Why Do Supervisors Roll Their Eyes?

• How does your current QA/compliance process flag interactions today (keyword rules, manual sampling, post-call review)? Options: Keyword/rule-based, Manual sampling, Scorecard-driven, Third-party audit, Other
• What percent of interactions are currently reviewed for quality/compliance? Options: <1%, 1–2%, 3–10%, 10–25%, >25%
• When supervisors receive a flagged item, what is the end-to-end review workflow and how long does it typically take?
• Tell us about a recent false positive—what triggered it, how much supervisor time did it consume, and what was the outcome?
• How tolerant are supervisors to initial false positives during tuning—do they prefer conservative settings or aggressive detection that we will iterate? Options: Prefer conservative (fewer false positives), Prefer aggressive (capture more, then tune), Undecided / need guidance

Is Privacy and Data Protection a Checklist or a Strategy?

• Are there regulatory regimes or internal policies that will restrict what we can capture or store for the pilot (GDPR, CCPA, HIPAA, PCI)? Options: GDPR, CCPA, HIPAA, PCI, Other, None / Not sure
• Do you require data to remain in a specific region or on-premises for the pilot? Options: Yes — specific region, Yes — on-premises only, No regional restriction, Not sure
• What are your expectations for data retention, masking, and redaction during the pilot?
• Who is the designated security/privacy contact (CISO, DPO, legal) and how quickly can they review technical controls?
• Which certifications or controls would reduce your friction most (SOC 2, ISO 27001, HIPAA BAAs, penetration test report)? Options: SOC 2, ISO 27001, HIPAA BAA, Pen test report, SaaS security questionnaire, Other

What Will Make Agents Stop Distrusting the ‘AI Button’?

• How were agents involved in the evaluation so far—were they shown scripts, demos, or invited to provide feedback? Options: Heavily involved, Some demos and feedback, Shown demos only, Not involved
• What are the top three agent objections you expect (e.g., loss of autonomy, accuracy concerns, speed changes)?
• How do you want agents to surface disagreement with the system during the pilot (silent override, feedback button, supervisor escalation)? Options: Silent override, Feedback button, Immediate supervisor escalation, Post-call feedback survey, Other
• Would you be open to a phased rollout where agents see recommendations but can opt out until confidence grows? Options: Yes — phased, Maybe — depends on conditions, Prefer immediate full experience, Not open
• What training or change-management support will make agents feel respected and prepared? Options: Role-based training, Job aids / cheat sheets, Quick in-call tips, Peer champions, Incentives/recognition, Other

Who Will Run the Show After We Leave the Room?

• If the pilot meets success criteria, which team will own ongoing operations (CC Ops, IT, Automation Center of Excellence, Vendor-managed)? Options: Contact Center Ops, IT/Platform, Automation COE, Vendor-managed service, Other
• Do you already have an SLA or runbook template that defines alerting, incident response, and escalation paths? Options: Yes, comprehensive, Yes, partial, No, but we can create one, No runbook exists
• What dashboard metrics and refresh cadence would you need to operationalize this (real-time alerts, hourly dashboard, daily report)? Options: Real-time alerts, Hourly dashboard, Daily report, Weekly summary, Ad-hoc queries
• Who will be responsible for tuning detection thresholds and reviewing false positives after go-live?
• What’s an acceptable response time for critical issues once the pilot is live (minutes, hours)? Options: Within minutes, Within 1 hour, Within 4 hours, Next business day

If We Built a Readiness Runbook Together, What Would Be Non-Negotiable?

• Which of the following acceptance tests must pass before the pilot is allowed to take live traffic? Options: Telephony connectivity and CTI handoff, CRM record correlation accuracy, Privacy/consent enforcement, Supervisor review workflow functional, Agent UI performance under load
• Who on your side must sign off on the runbook and final go/no-go (names/roles)?
• If an acceptance test fails, what is the preferred remediation workflow (pause pilot, partial rollback, fix & continue)? Options: Pause pilot, Partial rollback, Fix and continue, Escalate to execs
• How much buffer time should we plan between final signoff and flipping the switch to account for last-minute fixes? Options: Same day, 24–48 hours, 3–7 days, 1–2 weeks
• List the top three technical or organizational blockers you believe could prevent signoff.

Putting a Number on Readiness: Be Brutally Honest

• On a scale of 0–10, how ready is your org to run this pilot end-to-end (0 = not close, 10 = fully ready)? Options: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10
• Which single shortfall keeps your score from being a 9 or 10?
• Which internal resource (role/team) could remove that shortfall fastest if we give them the right playbook?
• If we allocated two full days with your key integrators and supervisors, which three outcomes would you want us to guarantee by the end?
• Would you be willing to set a tentative go/no-go decision point now (date) and commit to the joint readiness checklist? Options: Yes — propose date, Maybe — need more discovery, No — timeline not set

Quick Snapshot: Who's on this Mission?

• To get us started, what best describes the size of the contact center team we’d be piloting with? Options: ≤ 25 agents, 26–100 agents, 101–250 agents, 251–1,000 agents, > 1,000 agents
• Which industry best matches your operation for this pilot? Options: Telecom, Financial services, Healthcare payer, Utilities, Retail/ecommerce, Other
• What specifically triggered interest in a pilot right now (pick the primary reason)? Options: Failed compliance audit, Sudden AHT increase, New product launch, High QA failure rate, Automation initiative, Other
• Who are the core decision-makers and day-to-day owners for a pilot (list roles/titles)?
• What’s your target timeline to start the pilot? Options: Within 2 weeks, Within 1 month, 1–3 months, 3–6 months, Unsure

Are You Comfortable Letting 1 in 5 Calls Slip?

• When your audit showed ~18% missed disclosures (or similar), what was your immediate reaction—surprised, annoyed, resigned, or something else? Options: Surprised, Annoyed, Resigned/used to it, Alarmed/urgent, Other
• How often do compliance failures like this surface—quarterly audits, ad-hoc complaints, regulator outreach, or other cadence? Options: Quarterly audit, Monthly sampling, Ad-hoc complaint-driven, Regulatory notification, Other/unsure
• Walk me through one concrete example of a compliance failure you want us to understand—what happened on the call, and why was the disclosure missed?
• What are the downstream consequences when disclosures or required steps are missed (fines, rework, lost revenue, customer churn, brand risk)? Options: Fines/regulatory risk, Supervisor rework, Customer escalations, Revenue loss, Brand/reputation impact, Other
• How worried are you about agent pushback if an AI-driven assist flags compliance in real time? Options: Very worried, Somewhat worried, Neutral, Open/optimistic, Already supportive

Where the Day-to-Day Gets Messy

• If I sat with an agent for an hour, what three tasks would I see stealing the most of their time?
• How do agents currently find policy, scripts, or exception rules—search KB, ask a peer, supervisor interrupt, or something else? Options: Search knowledge base, Ask peers/supervisor, Rely on memory, Scripts in CRM, Other
• What parts of the CRM/telephony interaction are manual and repeatedly slow down handling (e.g., callbacks, multi-screen lookups, post-call notes)?
• What’s your current average handle time and what change would you view as meaningful in a pilot? Options: < 4 min, 4–8 min, 8–12 min, > 12 min, Unsure
• How much of your QA is sampled vs. automated today (percentage or description)? Options: <1% sampled, 1–2% sampled, 2–5% sampled, Mostly manual reviews, Some automated analytics
• Tell us about one integration that worries you most—telephony vendor, CRM, authentication system, or legacy middleware—and why.

Why Supervisors Keep Hitting Snooze

• Do supervisors trust the signals that existing tools provide, or do they feel overwhelmed by false positives and noise? Options: Generally trust, Mixed trust, Distrust due to noise, Don't use existing signals, Unsure
• How long does it typically take a supervisor to investigate a flagged compliance issue from detection to resolution? Options: < 30 minutes, 30–60 minutes, 1–4 hours, 1+ days, Varies widely
• What percentage of flagged items are false positives today, and how does that affect supervisor bandwidth? Options: <10%, 10–25%, 26–50%, >50%, Don't know
• Share a story where a false positive either created a problem or caused leadership to ignore alerts—what happened and what was the fallout?
• Which supervisor workflows would need to change to triage real-time flags efficiently (e.g., new dashboards, escalation queues, QA-automation handoff)?

If Your Agents Could Wave a Magic Wand

• Imagine the pilot succeeds—what three things look and feel different at the agent level?
• What single KPI change would make leadership say the pilot was a clear win (choose one)? Options: AHT reduction (%), Resolution rate increase (%), Quality score improvement (pts), Fewer compliance misses (%), Agent satisfaction
• How would agent trust show up in measurable ways—adoption rate of suggestions, overrides, NPS, or other signals? Options: Suggestion adoption rate, Override frequency, Agent CSAT/NPS, Supervisor interventions, Other
• If we had to define 'no-regrets' success for 60–90 days, what minimal improvements would justify expanding the rollout?
• What cultural or incentive levers would help agents accept AI nudges rather than resist them?

What Would Make You Confident to Pilot?

• What specific telemetry and dashboards will you need to see during the pilot to feel comfortable (real-time AHT, compliance hit rate, false-positive rate, agent adoption)? Options: Real-time AHT, Compliance hit rate, False-positive rate, Agent adoption metrics, QA score trends, Other
• Which KPIs should carry the most weight in the pilot success scorecard? Options: Handle time, Resolution rate, Quality score, Compliance miss rate, After-call work time
• What is an acceptable false-positive rate during the pilot before you’d consider retraining/rule changes? Options: <5%, 5–10%, 11–20%, >20%, No tolerance
• How many agents and how many weeks do you believe are needed to produce statistically meaningful results for your KPIs?
• Who in your org must sign off on pilot success (operations, compliance, IT, finance), and how will they evaluate the outcome?

Integration & Practicalities: What's Realistic?

• Do you control the telephony and CRM environments, or are they managed by a different team/vendor? Who will grant access? Options: We control both, Telephony managed by another team, CRM managed by another team, Third-party vendor
• Which of these technical items are ready for a pilot today, require small changes, or are significant blockers? Options: Telephony integration, CRM access, User provisioning, Data mapping, Privacy/consent controls, None/unsure
• Describe any regulatory or privacy constraints (HIPAA, PCI, GDPR) that we must design around for this pilot.
• Estimate how long your IT team would need to provide the necessary access and test integrations: days, weeks, or months? Options: A few business days, 1–2 weeks, 3–4 weeks, 1–3 months, Longer/unsure
• Who will be the primary technical contact on your side for integration and troubleshooting?

Next Steps, Commitments & Red Flags

• If the pilot delivers the expected KPI improvements, what is the typical approval path and timeframe for scaling (immediate, phased over months, budget cycle)? Options: Immediate approval, Phased over months, Requires next budget cycle, Depends on exec sponsor, Unsure
• What budget or commercial constraints should we be aware of that could stop the pilot even if performance looks good?
• What would make you pull the plug on a pilot early—false positives, agent backlash, integration failures, negative customer impact, or other causes? Options: High false positives, Strong agent resistance, Integration failures, Customer complaints, Budget/contract issues, Other
• Who on your team will be our executive sponsor and who will be the day-to-day champion to keep momentum?
• Before we leave this session, what single concern do you want us to prioritize resolving in the pilot plan?