Bottleneck Hunter

Workflow Constraint Review — Customer Onboarding

Evidence status: The figures are hypothetical and based on 20 cases. Treat the result as an initial hypothesis, not a statistically reliable diagnosis. Confirm how calendar days and stage boundaries were measured.

Observed pattern

Reported end-to-end median is 12 calendar days. Median waits are sales handoff 2 days, security review 6 days, setup 1 day, and customer scheduling 3 days. The listed stage waits total 12 days, but confirm whether they are sequential, overlap, or omit other touch time and queues.

Leading constraint hypotheses

Security review queue: largest reported stage wait; investigate capacity, batch cadence, required information, policy gates, and rework. A long wait alone does not prove that this stage limits sustainable throughput.
2.

Customer scheduling: 3-day wait may create demand-driven delays; check response time, time zones, availability, and outreach cadence.
3.

Sales handoff: missing or inconsistent information could cause downstream rework even though the wait is shorter.

What 20 cases cannot establish

The sample may not represent different customer segments, complexity, seasonality, or exceptions. Medians conceal variation and dependencies. Throughput, touch time, queue size, rework, and resource capacity are not supplied. Do not use Little's Law without compatible system boundaries and stable averages.

Data to collect next

For each case, record start/end timestamps per stage, active touch time, queue-entry time, customer segment, complexity, completeness at handoff, approval/rework loops, security reviewer capacity, weekly arrivals/completions, and exception reason. Preserve privacy and required controls.

Safe improvement tests

• Pilot a required-information checklist at sales handoff; measure missing fields and rework.
• Pre-triage security submissions for completeness without waiving review; measure queue age and rejection loops.
• Offer structured scheduling windows and reminders; compare response delay.
• Run one change at a time, keep a baseline, and watch quality, risk, compliance, and downstream queues.

Decision rule: Scale only if lead time or throughput improves without increased defects, risk, or work shifted to another stage. Re-measure after each test; the constraint may move.