Agentic AI spending: $944B now, $3.35T by 2030 — read it right
A new PHD/WARC forecast puts agent-facilitated consumer spending at $944B in 2026, rising to $3.35T by 2030. That's 3.8% of consumer spend. Here's the operator read.
A number is circulating this week that will end up in a lot of sales decks: agentic AI spending by consumers hits $944 billion in 2026 and $3.35 trillion by 2030. Both figures are real, both come from the same forecast, and both are being quoted without the denominator that makes them useful. Here it is: that $3.35 trillion is 3.8% of global consumer spending. Anyone telling you to rebuild your storefront for AI agents this quarter is quoting the numerator.
What actually happened
The forecast comes from From Abundance to Agents, published July 21 by Omnicom's PHD in partnership with WARC. MediaPost's write-up has the headline math: agent-facilitated consumer spending goes from roughly $944 billion in 2026 — 1.3% of global consumer expenditure — to $3.35 trillion, or 3.8%, in 2030. Tripling in four years, from a small base to a still-small base.
The category breakdown is the part worth your attention. Per ITBrief's summary of the report, the steepest growth is telecoms and utilities — up 611.9%, from $57.6 billion to $410.3 billion — driven by recurring billing cycles and price comparison. Financial services reaches $237.9 billion (up 235.3%). Travel and transport goes from $78.1 billion to $275.6 billion. Geography concentrates hard: the US takes $1.1 trillion of the 2030 total (31.9%), China $505.8 billion, the UK $131.2 billion, and ten markets account for about 68% of everything.
Read the winners and the pattern is obvious. Agents take over purchases that are repetitive, comparable, and boring — the ones where a human was never really deciding, just re-authorizing. They do not take over the purchase where someone wants to see the thing first.
Why agentic AI spending matters for your business
If you sell a considered purchase — custom work, high-touch services, anything with a conversation before the invoice — this forecast is not a threat and it is not a mandate. What it is: a five-year notice that a slice of demand will arrive through software that never loads your website.
So do the cheap work and skip the expensive work. The cheap work is making your data machine-readable: structured product and service data, accurate pricing and availability, a clean feed, schema markup that isn't lying about stock. That is a weekend of engineering for most catalogs and it pays off in ordinary search too. The expensive work — replatforming your commerce stack around an agent standard that will get rewritten twice before 2030 — is the thing to refuse for now.
And treat the forecast as a forecast. Four-year projections of a technology this young are directional, not bankable. The directional part is credible: delegation is coming for the routine end of your order book first. Plan for the recurring, reorderable, price-comparable part of your revenue to get mediated by something that isn't a person. Plan for the rest to keep working exactly the way it does today.
Key takeaways
- PHD and WARC forecast agent-facilitated consumer spending at ~$944B in 2026 (1.3% of global consumer spend), reaching $3.35T (3.8%) by 2030
- Telecoms and utilities grow fastest — up 611.9% to $410.3B — because recurring billing and price comparison are exactly what agents are good at
- The US alone accounts for $1.1T of the 2030 figure; ten markets make up roughly 68% of the total
- Agents take the repetitive, comparable purchases first. Considered, high-touch sales are largely untouched by this curve
- Operator move: make your catalog and pricing machine-readable now. Don't replatform for a standard that will change twice before 2030
Want your catalog readable by agents without rebuilding your store? We do the unglamorous version: clean product data, accurate feeds, real availability APIs, structured markup that matches reality. No replatform required. See what we build or send us your catalog and we'll tell you what's broken.
Sources: MediaPost, ITBrief Asia.
- #agentic-commerce
- #ai-agents
- #retail
- #product-data
- #forecasting
Tommy Rush — Founder, Rush Commerce
Operator turned builder. 15+ years running operations — now shipping the systems businesses run on. More
Get The Rush Report weekly — one email, zero fluff.
Keep reading
Instacart buys Arpalus: who owns your shelf data?
Instacart acquired computer-vision startup Arpalus to turn 600,000 shoppers into a shelf-intelligence sensor network. If they run in your stores, they know your inventory better than you do.
Read it68% of Google searches end without a click. Now what?
New SparkToro and Similarweb data puts zero-click Google searches at 68% and AI Mode visits at 279M a month. What that means for retailers who rent their traffic.
Read it