Commissioned in October 2023 across access, distribution and core, every intermediate and main distribution frame including cooling and power consumption, and all data centers — to set a manufacturer standard ahead of the coming refresh.
The methodology stated its reliability content plainly: each manufacturer would be assessed on how it determines availability, along with redundancy, mean time between failures and return on investment. The study was scoped to produce a quadrant ranking of manufacturers, a determination of where the existing design created single-manufacturer dependency, and a finding on whether a full retrofit or mixed-vendor coexistence carried the better value.
The instrument ran two matrices and twenty-five criteria at weights of one to four — the last descendant on this record of a scoring method first written in 2008 — with redundancy and survivability scored separately. The first asks whether a design absorbs a component failure; the second asks whether it keeps delivering service when it cannot. Keeping them apart is a reliability engineer’s distinction rather than a procurement one, and collapsing them is how a design that fails gracefully scores the same as one that merely fails slowly.
It earned itself in three weeks
A challenger’s reseller submitted a competing analysis claiming fifty to sixty percent savings on hardware and licensing. The figure had been priced against the wrong baseline.
A framework fixed in advance is what makes a substitution like that visible as a scoring failure rather than as a pricing dispute. Without one, the argument becomes whose number is right — which is an argument the incumbent usually loses, because the challenger chose the comparison.
Outcome
A framework fixed before the bids arrived, which is what made a substituted baseline visible as a scoring failure rather than a pricing dispute.