Coastal Freight: Benefits Realization on a Route-Optimization Platform
Coastal Freight spent $14 million on a route-optimization and dispatch platform intended to cut fuel and driver overtime. The approved business case projected $5.8 million in annual savings from the third year, with a benefits owner listed as the CIO. Nine months after go-live, only 22% of dispatchers use the recommended routes, realized savings are tracking at 31% of plan, and the vendor says the configuration is working as designed. Trailer-availability data is updated manually and is often four hours stale, so the optimizer produces routes dispatchers do not trust. Dispatchers are measured on on-time delivery, not fuel, and supervisors override plans when a customer calls. Training was a two-hour webinar. The CFO has asked whether the remaining $2.1 million of planned change funding should be released, and the audit committee wants to know why the benefits case was approved with an IT owner.
1. Diagnose why the projected benefits are not materializing. Separate technology causes from business-change causes.
Read the worked answer
Technology causes: stale trailer-availability data means the optimizer's recommendations are wrong often enough that dispatchers rationally ignore them. A two-hour webinar did not build the skills to interpret exceptions. The integration with dispatch may also be pushing recommendations rather than embedding them in the workflow.
Business-change causes are larger. Dispatchers are measured on on-time delivery, so overriding a route is rewarded. Supervisors override when a customer calls because the escalation path has no costed alternative. No one's performance objective includes fuel per mile or route adherence. The vendor's 'working as designed' statement is technically defensible and strategically irrelevant: the design assumed data and incentives that Coastal did not provide.
The benefits-owner error compounds all of this. Listing the CIO as owner made benefits an IT outcome, but the CIO cannot change dispatch incentives, customer promises, or data-entry discipline. The result is a classic benefits dependency failure: the technology was delivered, but the enabling changes were not funded or owned.
Diagnosis: roughly 30% technology and data, 70% operating-model and change. Releasing more money without fixing ownership and incentives would buy more training for a system people are correctly avoiding.
2. Who should own the benefits, and what should the revised benefits register and reporting cadence look like? Be specific about baselines and assumptions.
Read the worked answer
Owner: the Chief Operating Officer or VP of Operations, with the CIO accountable for platform capability. Each benefit needs a single named business owner who can change the behaviour it depends on.
Revised register - one line per benefit, with formula:
- Fuel per mile: baseline from the last 12 months, target, owner (fleet operations), dependency (data freshness under 30 minutes, route adherence above 70%).
- Driver overtime hours: baseline by depot, target, owner (depot managers).
- Dispatcher route adherence: leading indicator, currently 22%, target 70% by Q3.
- Customer on-time delivery: guardrail metric, to prove savings are not being taken from service.
Report leading indicators monthly to the operations leadership team and lagging savings quarterly to the CFO. Every savings figure must state its baseline, calculation method, and what was excluded. The CIO should not sign off benefits; the COO should, with finance validating the calculation.
Cadence: monthly adoption review for the first two quarters, then quarterly. If route adherence stays below 50% for two consecutive months, the register should flag the benefit as at risk rather than quietly annualizing a partial run rate. This is also the evidence the audit committee needs to see that ownership has genuinely moved.
3. Recommend whether Coastal should release the next funding tranche. State the decision criteria you would apply and what you would do if adoption does not improve by Q2.
Read the worked answer
Release it conditionally, not automatically. The remaining $2.1 million should be split into two gates tied to adoption, because the binding constraint is now behaviour and data, not software.
Decision criteria:
- Is there a named business owner with authority over dispatch incentives? If not, no release.
- Is there a funded plan to automate trailer-availability data to under 30 minutes? Without it, the optimizer remains untrustworthy.
- Have dispatcher and supervisor objectives been changed to include route adherence and fuel per mile, with on-time delivery as a guardrail?
- Is there a supervisor-level exception process that does not reward ad hoc overrides?
Release the first half against criteria 1 and 3; release the second half when leading indicators show adherence above 50% for two months and data freshness is achieved.
If adoption has not improved by Q2: stop the change spend, re-baseline the benefit case to the savings actually achievable at current adoption, and present the board with three options - fix the operating model, narrow scope to depots where adoption works, or retire the platform and take the write-down. Do not continue annualizing a plan the organization has already voted against with its behaviour. The audit committee needs a decision, not another status report.