Why 87% of Enterprises Missed Their 2025 Revenue Targets — and What the Predictive Revenue Movement Gets Right

One of the most-cited pieces of research in revenue-technology circles this spring is a Clari Labs finding that 87% of enterprises missed their 2025 revenue plans despite record AI spending across their go-to-market organizations. The number is the kind that headlines a vendor pitch, but it is also the kind that survives scrutiny. The 87% figure has been roughly consistent with what other sources, including Bain and Forrester revenue research, have reported in the same window.

The Clari thesis, which the company has been advancing aggressively since its merger with Salesloft closed last year and which now sits at the center of the combined company’s positioning under CEO Steve Cox, is that the underlying cause of the gap is architectural. Revenue technology has been built and bought as a stack of point tools — engagement here, intelligence there, forecasting somewhere else, enablement off to the side, deal acceleration in its own silo. Each tool produces signal about the revenue function. The signals do not aggregate into a coherent system of record. The forecast at the end of the chain is fundamentally guesswork dressed up as data.

This is the core argument behind what Clari and several adjacent vendors are now branding as the “predictive revenue system.” The pitch is that revenue technology should converge into a unified operating layer that integrates engagement, intelligence, forecasting, and execution under one roof — and that the convergence is what unlocks the AI capability that the unbundled stack cannot. The pitch is, in important ways, correct. The integration is the hard part of revenue technology, and the vendors who solve it will produce more durable customer outcomes than the vendors who do not.

The pitch is also, in other important ways, oversold. The convergence story has been told by every revenue technology vendor at some point in the last decade, and the convergence has been most successful when it served the customer’s actual operating model rather than the vendor’s go-to-market preference. Customers who have already built a coherent revenue stack around their existing tooling, with mature integrations and clean data flow, may not benefit from replacing it with a unified platform. Customers who have not — and who are spending real money chasing the kind of AI capabilities that require coherent data flow — almost certainly will.

The competitive context is fast-moving. Gong’s Mission Andromeda rollout in February — combining Gong Enable, Agent Studio, and AI Trainer in a single platform that crossed $500M ARR last month — is the most credible competing platform vision. Outreach’s Omni, launched in April, is making its own platform case. Highspot and Seismic, post-merger, are positioning the combined entity as the platform layer for sales enablement specifically. The market has, for the first time in roughly a decade, real competition among credible platform contenders.

For CROs in 2026 planning cycles: the platform-versus-point-tools question is now an active commercial decision, not a vendor talking point. The right answer for your organization depends on the maturity of your current stack, the willingness of your operations team to absorb a multi-quarter migration, and the credibility of the platform vendor on the workflows that matter most to your business. The 87% miss rate is real. The platform pitch may be the right response. It is not the only response, and the wrong choice between platform and point tools will cost as much as the original problem.