The approval process that costs you four hours per request.

The software estate at most enterprises is not understood. It is catalogued. There is a difference. A catalogue tells you what you own, when it renews, and how many seats were provisioned. It does not tell you what those products can do, at the feature level, for the specific workflow someone is trying to solve today. That gap is where the four hours goes.

The question that sounds simple

When a business unit requests a new tool, the question on paper is familiar: do we already own something like this? Procurement managers hear it constantly. And almost every time, answering it properly takes the best part of a morning.

The reason is that the question, as asked, is the wrong one. What the requesting team actually needs to know is whether existing tools cover their specific use case, and if not, what combination of products comes closest. That is a feature-level comparison across an estate that may run to thousands of products. No catalogue, no spreadsheet, and no SAM platform is built to answer it on the fly. The procurement manager ends up doing manually what should be handled structurally.

Why the catalogue cannot close this gap

A typical enterprise catalogue entry for Microsoft Teams reads something like: unified communications, 6,000 seats, renews September 2027. It does not say whether Teams covers the specific messaging and file-sharing workflow a newly acquired department relies on Slack to run, or whether the overlap is close enough to justify a migration rather than a parallel licence.

That gap is where duplicate spend accumulates. Not through negligence. Through a process that lacks the right information at the right moment. The procurement manager searches vendor comparison sites, reads whatever came up in the last review cycle, tries to recall whether something similar was discussed six months ago, and then writes a response the requesting team ignores because their tool is, in their view, totally different.

By the time a second platform is onboarded, provisioned across a department, and embedded in daily workflow, consolidation requires far more than a cancelled licence. It requires migration, retraining, and a negotiation with a vendor who now knows you are trying to leave.

The cost of compounding

The licence fee for a duplicate tool is rarely the number that matters. The real cost is what follows. One approved duplicate becomes a departmental standard. The departmental standard becomes a company-wide rollout. By the time rationalisation appears on the agenda, the estate holds three overlapping contracts for the same capability, each with its own renewal cycle, admin team, and migration risk.

Most enterprises accept this as a cost of scale. It is not inevitable. It is a governance failure that repeats once per request, indefinitely, until the approval process has something better than a senior person's memory and a comparison article from last year.

What the decision layer actually changes

A capable software decision workflow does not require months of integration work or a perfectly clean data estate. Sam works with imperfect data from day one, and returns a sourced answer in minutes. The ability to say what the estate owns, what it does at the feature level, and whether it covers the case in front of you, is what turns a four-hour investigation into a two-minute answer.

The ambition is not to make procurement faster for its own sake. It is to make the no credible. When a procurement team can point to a sourced, feature-level comparison, the requesting team listens. When they cannot, the request goes through. The tool gets approved. The duplicate gets added. The catalogue grows by one more entry nobody fully understands.

The people responsible for saying no do not have what they need to say it convincingly.

Four hours is a symptom

The four hours a procurement manager spends per request is not inefficiency. It is the correct amount of effort required to do the job with the tools currently available. The answer is not to ask procurement to work faster. It is to give them a decision layer that has already done the legwork, drawn on the estate, and returned a sourced recommendation before the meeting starts.

One avoided duplicate purchase typically covers the annual cost of doing this properly. One caught renewal covers considerably more. Enterprises that have gone through a portfolio rationalisation consistently find the savings were already sitting in the estate. They just had no way to see them fast enough to act on them.