Raw

**Fieldbook: extend to North and Islands in Q1, subject to three prerequisites**

You asked for this review before the quarterly meeting. Fieldbook has been live for two quarters in Central, East, South and West, and the pattern across those four regions is not uniform: two paid back their running cost inside the second quarter, two are at break-even. Extension adds £13,500 a month and lands in the winter travel season, so the question is whether the North and Islands rollout would follow Central's curve or West's.

**Extend in Q1, but only after onboarding is rebuilt and the two integration gaps are assigned an owner.** Every shortfall to date traces to those three things; none traces to the platform, and all three have solutions already available to us.

**1. Where onboarding worked, Fieldbook paid for itself inside a quarter.** Completed jobs per technician per day rose from 4.6 to 5.3 across the four live regions and travel time between jobs fell 18%; overtime is down 12% quarter over quarter and fuel per job down 9%. Emergency insertion, formerly a round of phone calls, now takes a median of 4 minutes. Against £31,000 a month, Central and East cleared their running cost in the second quarter, both having reached full adoption within six weeks.

**2. The lag in South and West was a training-schedule decision we can reverse before Q1.** The two-day classroom course ran once per region, so anyone hired or absent after that date learned second-hand: 34 of West's 118 technicians attended no session, and West is still at 71% of orders in the system after nine weeks against South's eleven weeks to pass 80%. The consequences are all downstream of that single choice — 44% of the 1,940 support tickets are logins and password resets concentrated among untrained technicians, and improvised shadowing consumed roughly 380 technician-hours of route time in the two regions. Dispatchers, who had no fallback once the old spreadsheet was switched off, were at full use from week one across all four regions. A continuous onboarding model — the existing weekly webinar plus a scheduled session for every new starter — replaces the single classroom slot and removes the need to fly trainers into North and Islands in winter.

**3. Three known defects are eating the operational gains, and each is waiting on a Kestrel decision, not on vendor work.** No feed from the legacy asset register means the equipment-history panel is empty on every job screen: technicians call the back office or the customer, and in 9% of audited jobs replaced parts already replaced under warranty within the year. The request for a nightly export has sat in the asset register team's backlog since month two. Job-completion codes do not match the invoicing system, so the billing team re-keys about 300 records a week — a temporary clerk at £8,400 for two quarters, transcription errors reaching customer invoices, and a 2–4% weekly disagreement between Fieldbook and invoicing that finance reconciles by hand. The vendor says a configurable mapping table exists and takes about a day; it is unscheduled because no one owns the integration. Third, the search screen matches on exact postcode only, so technicians create duplicate customer records — 19% of all tickets, about 70 merges a week, and split job history until merged. The vendor's catalogue lists supported connectors for both the asset register and the invoicing import format; neither has been evaluated, because the program was staffed for deployment and training only.

**4. Without a route from field observation to program fix, North and Islands will repeat the same two quarters.** Every negative pattern in this review was visible by week four of the Central go-live, and each subsequent region reproduced it, because the program had no mechanism to turn early field reports into program-level changes. The support desk lead's summary: "We are doing the training program's job at ticket prices." The extension economics are favourable if Central and East's adoption pace is reproduced and negative if West's is — a repeat of West would leave both new regions underwater for at least two quarters, so the mechanism matters more than the licence cost.

**What I need at the quarterly meeting**

Approval to extend to North and Islands in Q1, conditional on three items closed before the first go-live date: a continuous onboarding model replacing the single classroom course; a named owner for the asset-register feed and the invoicing code mapping, with both connectors evaluated; and a standing weekly review in the first six weeks of each new region with authority to change the program, not just report on it. If any of the three is not in place by the decision date, my recommendation is to delay the North start rather than proceed — the same £13,500 a month against West's curve does not pay back.

One item from the source reports is excluded as not decision-bearing: the unpopular parts-ordering screen, which tests correctly and appears to be habit rather than a defect.