La connectivité est un problème d’exploitation, pas une ligne budgétaire d’achat.
Traitez le réseau comme une ligne budgétaire, et chaque panne devient le problème de quelqu’un d’autre.
Procurement buys connectivity the way it buys stationery. Specify the circuit, compare the quotes, sign for the lowest price that meets the spec, file it. Done, until the site stops working. A network is not stationery. It is the operating system of a business. When it degrades, tills stall, calls drop, shifts are lost. And because it was bought as a line item, nobody owns the degradation. The vendor met the spec. Procurement met the budget. The site still can’t trade. Buying connectivity as a procurement line optimises for the moment of purchase and ignores every moment after it, which is where all the value and all the risk actually live. The cheapest circuit that meets the spec is not the same thing as the cheapest site that works. An operating problem is owned differently. You measure whether it carries the work, not just whether it was delivered. You fix degradation before it becomes an outage. You hold one accountable line from order to evidence, instead of a folder of contracts and a diagram of who to blame. It costs less that way too, consolidating a multi-supplier stack typically runs around 20% cheaper. But cost is the smaller point. A site that works is an outcome you can own. A circuit that met its spec is not.