In the workplace dictionary, the three letters CYA stand for an unprintable mantra about covering one’s own back — its meaning is simple: protect the person who made the decision. Those three letters explain the most stable pattern in enterprise procurement. Given the same requirements, the option that costs half as much and does everything asked of it loses; the expensive one wins. This is not stupidity. It is rationality — with the beneficiary being the buyer’s career, not the company.
Big Vendors Are Stable, and Stability Doesn’t Explain the Gap
Some say the big brands cost more for good reasons: stability, ecosystem, support — all real, so choosing a big vendor is prudence, not cowardice. Fair. But that logic explains nothing about the proposal that is half the price, meets every requirement, and gets rejected without a trial. The reason for rejection is not on the spec sheet. The spec sheet has no column titled: when this breaks, who answers for it. The buyer keeps a different ledger. Choosing Microsoft or IBM is beyond criticism. Choosing an unknown brand — even one that satisfies every requirement — means that the first time anything goes wrong, the question will not be “what broke” but “why on earth did we pick these guys.”
Insurance Is Billed Per Incident, and One Incident Is Enough
Others say: buy the cheap option, and if it fails, swap it out — the cost of a trial is low. The ledger is wrong. What burns on the day of failure is not the price difference; it is the buyer’s credibility. The first outage, the meeting room has a single agenda item: who signed off on this. The budget saved by the cheaper proposal cannot cover one accountability session. This insurance is billed per incident, and one incident collects in full — what it takes is not money but the person who chose the cheap option.
Nobody Is Obligated to Be the First Martyr
There is a third line of argument: this is a culture of cowards, and good companies should reward correct-but-cheap decisions. They should. But until the incentive system changes, the individual cannot beat the organization’s arithmetic: an organization sees the errors that happen, never the errors that were avoided. Buy the big brand, nothing breaks, the product gets the credit. Buy the small brand, nothing breaks, nobody says a word; break once, and the name is on file forever. The person who picked the big vendor has not necessarily decided wrongly — that person traded an unprovable decision for an insurance policy that can be produced on demand. Before the culture changes, nobody is obligated to be the first martyr.
As for the buyer who took the small brand and walked into the meeting room after the first failure — the choice was not necessarily wrong; it merely paid full price for a kind of correctness the organization never encouraged.
Building In-House Doesn’t Remove Risk, It Relabels It
Then build it in-house — keep the functionality and the data. Sounds the toughest. But the three things in-house projects promise — budget, timeline, scope — have never been estimated accurately by IT: say a million dollars, land at one and a half to two million; say six months, take twelve. Building in-house swaps immunity for liability: buy an off-the-shelf product and failures get escalated to the vendor; build it and failures lead back to the person who approved the build. That is why ready-made solutions are the default: the code is not in one’s own hands, and neither is the blame. Rights to use the functionality remain; ownership of the outcome passes on too.
The Other Half of the Premium Is Paid to the Meeting Room
Of the gap on the invoice, part buys functionality. The rest is paid to the meeting room after something breaks — paid so that no one asks, at the moment of failure, who made the original choice. This insurance is not worth buying everywhere: reversible purchases, spending one’s own money, trials that are genuinely cheap — pick on merit there. Heavy premiums are charged in exactly one territory: organizational procurement where decisions are irreversible and accountability lands on a named person. To decide whether to pay, ask one question first: if this choice goes wrong, who sits in the defendant’s chair.
