The Regional Rollout That Outran Its Evidence
Northwind Utilities runs a 14-person customer service team handling 900 calls a day on a 12-year-old CRM. The operations director wants a £1.2m replacement approved within six weeks to align with a corporate 'digital first' announcement. Two data points exist: average call handling time is 7.4 minutes, and a staff survey shows 62% dissatisfaction with the current system. No baseline exists for first-call resolution, repeat contact rate, or cost per contact. The finance business partner will only sign off if the case shows measurable benefit within 18 months. The vendor has offered a 9% discount if the contract is signed this quarter. The CEO has publicly committed to the project at an industry conference.
1. Identify the gaps that prevent a defensible business case and explain why each matters to the approval decision.
Read the worked answer
The case rests on two weak signals: handling time and satisfaction. Neither is a benefit measure.
Key gaps:
- No cost-per-contact baseline. Without it, you cannot express savings in money, only in minutes. Finance cannot validate an 18-month payback.
- No first-call resolution or repeat contact data. These drive demand volume. If FCR improves, call volume falls and the benefit is real; if not, the new CRM is a like-for-like swap.
- No root-cause analysis of the 62% dissatisfaction. Dissatisfaction may stem from policy, training or staffing, not the tool. A new system that leaves those untouched will not move the number.
- No options considered. A single vendor with a discount deadline is not an appraisal.
The discount and the public commitment are decision pressures, not evidence. The analyst should flag them explicitly to the sponsor and recommend a short baseline exercise (4–6 weeks of FCR, repeat contact and cost-per-contact measurement) before the case goes to finance.
2. The sponsor says the six-week deadline is immovable. Recommend how to proceed without fabricating evidence.
Read the worked answer
Proceed, but change what is being approved.
- Split the decision. Ask for approval of a funded discovery and options stage (a small fraction of £1.2m), not the full implementation. This respects the deadline without committing capital to an unproven case.
- State assumptions explicitly. Where baselines are missing, present a range with the source of the assumption and a sensitivity test: 'if FCR improves by 5–15 points, annual benefit is £X–£Y.' Finance can challenge a range; it cannot challenge a fabricated point estimate.
- Name the risks on the page. Discount expiry is a commercial risk, not a benefit. Public commitment is a reputational risk if the project fails, which argues for rigour, not speed.
- Set a gate. Define the evidence needed before the implementation tranche is released.
This is the professional response: deliver a decision on time, but make it the right decision.