Microsoft's move to consumption pricing invites the one question the bundle was designed to prevent

The bundled enterprise software subscription was, among other things, a device for making price scrutiny inconvenient. A single annual invoice, negotiated once and forgotten, is not an invitation to ask whether each component earns its share. Usage-based pricing ends that arrangement. Patrick O’Shaughnessy, investor and founder of Positive Sum, has made a specific, checkable call about what follows: when enterprise customers start seeing itemized monthly bills tied to actual consumption, they begin evaluating Microsoft’s products one by one.

O’Shaughnessy’s framing is worth quoting directly, because the mechanism he identifies is behavioral as much as financial. “If you’re every month looking at your Microsoft bill and how much do I use, you start thinking about what am I paying for. How good is each of these products? Should I actually just start thinking about and spraying this out?” The question embedded in “spraying this out” is the operative one. It is a description of disaggregation: replacing a bundle with a portfolio of best-in-class point solutions, each chosen on its own merits.

The timing of the call matters because Microsoft has moved its Copilot offering toward consumption-based billing. That is a structural change, not a marginal one. The predictability that defined enterprise software contracts for two decades is being traded for a meter, and the meter changes what customers see and when they see it.

If you're every month you're looking at your Microsoft bill and how much do I use? You start thinking about what am I paying for? how good is each of these products? Should I just start thinking about and spraying this out? Patrick O'Shaughnessy

The case O’Shaughnessy is making is not that Microsoft’s products are bad. It is that the bundle insulated those products from the kind of monthly scrutiny that a consumption bill now forces. A per-seat fee, paid annually, is sunk. A metered fee, visible each month, is a standing question. The cognitive difference between the two is where his prediction lives. Once customers are already reading the meter, the mental work of comparison-shopping is mostly done.

What would have to be true for the call to land? Enterprise customers would need to act on the scrutiny that monthly billing invites, not merely feel it. That is a meaningful hurdle. Switching costs inside a Microsoft ecosystem are real: Active Directory, Teams, SharePoint, and the broader Microsoft 365 stack create dependencies that a per-product evaluation does not dissolve overnight. A sophisticated procurement team might scrutinize the bill every month and still renew the bundle, having decided the switching cost is not worth the savings.

The counter-scenario is also worth naming. Microsoft could find that usage-based pricing increases revenue rather than threatening the bundle, because customers who pay only for what they use may be willing to use more. The metering model does not automatically weaken Microsoft’s position. It only changes the question customers ask. O’Shaughnessy’s prediction requires the answer to that question, in enough cases, to be “less Microsoft.”

The horizon on this call is not stated with precision, which is a limitation of the prediction. What it does offer is a falsifiable claim about direction: that the shift to consumption pricing is a threat to Microsoft’s bundled enterprise model, not merely a pricing adjustment. In two or three enterprise procurement cycles, it will be possible to see whether customers are actually disaggregating, whether renewal rates on bundle tiers are softening, or whether the monthly bill turns out to be something finance departments glance at and file. Until then, the call stands as a specific and serious bet on how procurement psychology responds to a change that has already happened.