The 12-month purchase of a 6-week feature

A $500K enterprise software deal takes six to eighteen months to close, and some run past two years. Roughly 60 to 70% of that clock is evaluation and procurement - not building anything. So when the deliverable is a loyalty feature a small team could ship in six weeks, you are spending a year buying weeks of work. The machinery is sized for a platform bet, then pointed uniformly at everything that passes through it.

Why does buying cost more than building?

A complex B2B purchase is not one decision; it is roughly eleven people each making a smaller one, in sequence, with the standing right to send the whole thing back a step. Gartner's research finds 77% of buyers describe their most recent purchase as complex or difficult - and that is the buyers talking about a process they themselves run. The two functions that reliably add the most delay are legal and procurement, which is to say the parts of the cycle that never touch the product. Of the months a deal takes to close, the majority is spent producing confidence, not working software.

Every row on that table is defensible in isolation. Read down the column, though, and you are looking at a fixed overhead applied to a purchase before anyone has asked how big the purchase actually is. A committee assembled to weigh a five-year platform will convene for a one-season feature just the same, because the intake process cannot tell the two apart. The cost of the cycle is decided by the shape of the process, not the size of the thing moving through it.

You are pricing a seven-year bet, not a feature

None of that machinery is irrational on its own terms. A platform you will run for five to seven years deserves a security review, a total-cost model, data-residency answers and a credible exit clause. The diligence is proportionate to the horizon. The failure it exists to prevent is real - a seven-year commitment is exactly the kind of bet that can end with a go-live suing its integrator. The trouble is that the same template gets pointed at everything entering the funnel, including purchases whose horizon is a single season.

A returnables reconciliation rule, a promo mechanic, a delivery-window optimiser: these are features, not platforms. Modern teams ship them in weeks, and the honest numbers on AI-augmented delivery have made the build faster still. Running a feature through platform-grade procurement means paying seven-year diligence costs on a six-week deliverable. The unit you buy - a multi-year platform commitment - and the unit that carries the value - one dated outcome - have come apart, and procurement prices the larger one every time.

What does the delay actually cost?

Here the mismatch stops being an accounting curiosity and starts costing money. Suppose the business case for a loyalty suite rests on the coming promotional calendar: the summer push, the festive window, the sponsored tournament. Those windows are the entire justification. Run the RFP at enterprise pace and, by the time it closes, the promo calendar that justified it has passed twice. You did not merely pay for the delay - you bought a thing whose reason for existing expired before it arrived.

In beverage and CPG, the value of most features is dated. It is tied to a season, a launch, a peak, a route change. A dated asset delivered late is not a late asset; it is a different and smaller one, because the demand it was shaped around has moved on. The carrying cost of a twelve-month cycle is rarely the interest on the money. It is the compounding decay of a business case that was only ever valid inside a window.

Does signing the contract end the wait?

Signature is not delivery, either. The organisation that took a year to buy is usually the same one that meters what it built - the quarterly release train that guarantees quarterly features. So the feature that took six weeks to build and twelve months to buy then waits again for a slot on the calendar. The buying cycle and the delivery cadence are two separate queues, and the same small artifact sits in both, ageing. By the time it reaches a user, the promo it was for is a case study in someone else's retrospective.

The fix direction is not faster procurement of the same thing; it is buying a different thing. Buy an outcome with exit criteria - a scoped pilot with a named success metric and a real off-ramp - rather than a seven-year platform bet wearing the costume of an urgent feature. Then the diligence matches the horizon, the committee shrinks to the people who actually own the outcome, and the window you were aiming at is still open when you arrive.

The gap is widening from both ends. Building keeps getting cheaper and quicker; enterprise buying, governed by committee size and legal throughput, does not. Left alone, the ratio between the twelve-month purchase and the six-week feature only gets worse - which suggests the next real advantage in route-to-market goes less to the brewer with the best platform and more to the one that learns to buy in seasons instead of epochs.