Scoping the First Release
Scope Meridian’s first release by weighing customer value, enforceable boundaries, and operating readiness. Interpret the deliberately selected 60-contact review without treating its category counts as population estimates. Propose representative-reviewed general billing-guidance chat drafts, conditional on approved guidance, staffing, fallback, measurement access, and accepted owners. Exclude account access, account changes, autonomous customer responses, and voice contacts. Define pilot eligibility, draft support, escalation, handling-time comparisons, and seven-day outcomes with denominators and unknowns. Record the recommendation, alternatives, exclusions, evidence, and unresolved dependencies before seeking approval to proceed. Keep proposed work separate from results.
We'll cover the following...
Which workflow should an agent take on first? The tempting answer is the biggest category in the data, and at Meridian that answer would put account access and plan changes ahead of the permissions needed to handle them safely. A good first release needs three things at once: a task worth doing, a boundary you can enforce, and people ready to run it. This lesson takes Meridian from the brief to a scoped first release, with a success definition the customer agrees to and a written list of what the release won’t do. You’ll record the decision in telecom-case-brief.md.
Which discovery questions change the plan?
The brief left the first workflow open. Meridian still has 12,000 contacts a month, 4,800 of them by chat, an eight-minute median handling time across phone and chat, and a support leader who wants 70% automation. None of those figures tells you how big each category is or how an agent would perform.
The FDE role includes scoping work, sequencing delivery, removing blockers, and adjusting scope, speed, and quality to protect delivery. Good discovery questions are the ones whose answers could change your recommendation:
Identity: What verification must happen before an automated system reads account-specific information? Without an accepted method, any account-specific automation is blocked.
Policy: Who approves the official answer when guidance conflicts? Without approved guidance, even a general answer may be wrong.
Interfaces and permissions: Which integrations are ready, and what may a service read or change? A console that representatives use doesn’t grant a service access.
Operations: Who reviews drafts, takes escalations, and handles failures? Available representatives are a good sign, but staffing and fallback need someone’s commitment.
Customer cost: Which mistakes hurt most? A wrong general explanation, showing one customer another customer’s bill, and an unauthorized plan change each call for a different boundary.
What did discovery find?
Discovery produced a review of 60 contacts: 28 account-specific billing contacts, 20 general billing-guidance contacts, and 12 plan-change contacts. Meridian picked them on purpose to show how workflows differ, so the counts aren’t estimates of the full contact population.
Four of the 20 general-guidance contacts depended on guidance that reviewers couldn’t reconcile. Automated identity verification and live integration access are still unconfirmed. Representatives can review drafts, but Meridian hasn’t committed staffing or a fallback.