Survey says: gimme more AI Speaking of revenue growth, ~88% of enterprise decision-makers anticipate spending more on AI over the next 12 months, according to a recent Wharton Survey. That’s a 16pp increase from last year, or +22%. The biggest shifts in “substantial” spending increases are in IT, Operations, and Finance/Accounting, where ~25% of respondents expect a lot more to come. Legal remains the relative laggard, but even there, a majority of respondents expect AI spend to increase. Also of note: the number of respondents expecting to decrease spend is barely visible. That’s consistent with the wildly impressive retention curves, recently highlighted by Ramp’s spending data: AI retention curves for business customers have improved every year, with current 9-month retention just under 90%. This one pretty much speaks for itself. Sure, it’s only a survey, but expectations are that enterprises are going to keep investing in AI, and even more so than before (and basically no one said “less”). ROI may be hard to measure, but we’re two years in and businesses like what they see, more and more. We’re still in the very early innings of this thing. Read more here: https://lnkd.in/gpENSDcW
The data says it all — AI spending isn’t just growing, it’s compounding. We’re clearly moving from experimentation to enterprise-wide adoption. 🚀
Couldn’t agree more Andreessen Horowitz — the compounding effect of AI adoption is just beginning. 📈 Enterprises are moving from experimentation to measurable impact, and the retention data proves it’s sticking. The real story now is how fast AI spend translates into process redesign and new business models.
I'm curious to what extent the term AI is being used as a marketing phrase - putting a wrapper around existing functionality that either was already growing or needed the naming help of what's currently en vogue.
Everyone’s racing to spend more on AI — but the next wave isn’t about more models, it’s about verifiable outputs. Crebiliti calls this the Trust Layer — where every AI statement can be scored, verified, and traced back to evidence, not assumption. Enterprises won’t just buy AI; they’ll demand auditable truth streams. That’s when ROI stops being fuzzy and becomes measurable.
This reminds me starkly of the vast amount of CapEx moved into Y2K, Internet (with a capital 'I'), Internet of Things (or IOE) promises of trillion dollar market cap, EVs CapEx then losses for Big 3 auto mfrs, now hundreds of billions into the AI, chip Mfrs, grid expansion, "mini" / modular nuke reactors to cover gap energy needs ... TRILLIONS moving around. Money will always move to where it most efficiently will be utilized (in the end), but these promises have a hollow ring of verisimilitude to the callbacks to previous aspirational "new, new things."
Encouraging signal. Budgets rising in IT, Ops, and Finance is exactly where AI creates durable value in regulated sectors: QA/CAPA, labeling and claims review, and retailer-compliance workflows. The improving retention curves suggest the ROI is real even if it’s hard to measure upfront—cycle-time reduction and right-first-time rates are doing the work.
Serial Founder & CEO | Ex-Google | Engineering & Business Leader | EMBA | Exec Educator @UCBerkeley | Public Speaker | Hi-Tech: AI, Web3, WebAI | #ResponsibleAI
9hAbsolutely. A huge wave of spend is flowing into cost reduction, which pushes budgets toward Ops, IT, and Finance. These are classic left-brain functions where AI shines. Smart companies are realizing that the fastest ROI often comes from upgrading the analytical backbone first. h/t #ArthurBrooks