Options paper · warehouse management

Extend the current system; revisit in eighteen months

Replacement scores highest on capability and worst on everything that determines whether we survive the transition. The recommendation is the option we can reverse.

Prepared 11 Sep 2026
3 options · 7 criteria
Decision due: 30 Sep 2026
Recommended
Extend
reversible, lowest carry
Cost over 3 years
€1.4m
vs €4.8m to replace
Reversal cost
€0.2m
if we change course in FY28
01

The comparison

Exhibit 1

Options against weighted criteria

Replace wins on capability and loses on risk, disruption and reversibility. Those three are what a failed WMS programme is actually made of — and the highest raw score belongs to an option that is not viable at all.

Criterion Weight Extend current Replace Do nothing
Weighted total
Passes the support gate
"Do nothing" scores highest and is still not the recommendation. Vendor support ends in Q3 FY28, which is a gate, not a criterion — no weighting of the other six makes an unpatched warehouse system acceptable. A weighted score ranks the options that remain viable; it does not decide which ones are.

Scored 1–5 against each criterion by the programme board, 04 Sep 2026. Weights set before the options were scored, which matters — setting them afterwards lets the preferred answer choose its own scoring. Cost figures are three-year totals including internal effort. The support gate is binary and was agreed before scoring began.

02

What each option costs to get wrong

Exhibit 2

Score and asymmetry side by side

Replace leads on weighted score only until you read the right-hand column. The cost of being wrong is not symmetric.

Exhibit 2a

Weighted score

Exhibit 2b

The asymmetry

OptionIf rightIf wrong
ExtendBuys 18 months for €1.4m Sunk €1.4m, replace anyway in FY28
ReplaceModern platform, 3-year horizon Peak-season outage during cutover; no way back
Do nothingZero spend Vendor support ends Q3 FY28 — fails the gate, not on the table

A cutover in peak season is the failure we cannot buy our way out of, which is why reversibility is weighted at 15%.

03

What would change the recommendation

Exhibit 3

The path, and the condition that invalidates it

Recommending a reversible option is only honest if you also state the review point and what would force the other choice.

  1. Extend, and start the discovery
    €1.4m over three years. Discovery runs in parallel so a future replacement starts from evidence rather than a blank page.
  2. Review point
    If throughput has grown past 40k lines a day, or the vendor confirms an end-of-support date inside 24 months, replacement moves from optional to necessary.
This recommendation fails if peak volume grows faster than 12% a year. Above that the extended system runs out of headroom before the review point, and we would be choosing under pressure rather than on evidence. Current run rate is 7%.
Why "do nothing" is on the list at all

Because it is the honest baseline, and because scoring it is what exposed the gate. Every options paper that omits it quietly assumes the spend is already approved. Scoring it and then ruling it out on a stated gate is more defensible than never showing it.

Who was consulted

Operations (three sites), IT infrastructure, finance, and the incumbent vendor. The vendor's input is recorded but not weighted — they are not a neutral party on replacement.