What the research says
The InfoBrief — Build for What's Next: Open-Source Architecture and the Future of AI Commerce, authored by IDC's Heather Hershey and sponsored by WooCommerce — synthesizes IDC's 2026 predictions and spending data into one commercial claim: agent-mediated shopping is no longer a forecast, and merchant preparedness lags well behind it. The framing that stuck with us is the second buyer. A human buyer responds to photography, copy and layout. The agent buyer parses fields. Where a field is missing, wrong or stale, the agent does not squint and guess generously the way a person might — the product simply drops out of consideration.
The sharpest line in the research is about trust: an agent will not recommend a brand it cannot verify. That sentence deserves more attention than the traffic projections around it, because it names the actual mechanism of exclusion. Verification is not a vibe. It is the ability to check a claim — this price, this stock state, this seller — against an authoritative source at decision time.
The gap, made concrete
"Not ready" is abstract until you look at what an agent actually encounters on a typical WooCommerce store. Three failures repeat:
- No machine-readable catalog surface. The products live in HTML built for human eyes. An agent is left to scrape, and scraped data is exactly the kind it cannot verify.
- No live answers. Feeds and snapshots age. Price and availability are boundary truths that are only trustworthy when they come from the store's own database at request time.
- No provenance. Even when data exists somewhere — a marketplace copy, an aggregator feed — the agent cannot establish that the merchant stands behind it, consented to its publication, or still agrees with it.
None of these is solved by better product photography or a faster theme. That is why we call it a data gap: the missing layer is structured, live, attributable product data served from the domain that owns it.
What readiness looks like when it ships
Our answer to the second buyer is deliberately boring: publish the catalog the way machines read, from the store itself. KaliCart Bridge puts discovery documents, bounded read-only search and honest price and stock blocks on the merchant's own domain — live WooCommerce data, not a copy. The federated catalog adds the cross-store layer with consent and provenance attached, so an agent that asks "which store has what I need" gets answers a merchant actually stands behind. And because readiness is measurable, ready.kalicart.com will scan a domain and show which of these layers exist on it right now.
We have also started publishing what happens when the second buyer shows up: one week of real agent traffic on the federated catalog — five hundred searches by an agent nobody prepared for, navigating governed taxonomy without help. The research says that buyer is coming. Our logs say it has visited.
Where we would push back
Two cautions, in fairness. First, this is sponsored research: WooCommerce commissioned a study whose conclusion favors investing in commerce infrastructure, and while the mechanism it describes matches what we observe, adoption projections in sponsored work deserve the usual discount. Second, readiness is necessary, not sufficient. A perfectly readable catalog still competes for agent attention through distribution — registries, discovery signals, network membership. Fixing the data gap makes a store eligible. It does not make it chosen. Merchants should hear the first claim as urgent and the second as honest.
Sources
- AI agents are already shopping and most merchants aren't ready — press release (August 2026)
- IDC InfoBrief, sponsored by WooCommerce: Build for What's Next: Open-Source Architecture and the Future of AI Commerce, Doc. #US54662026-IB, July 2026 — available from WooCommerce
- WooCommerce: Agentic Commerce hub
- Related: Before the wallet, there is the catalog · AI-Ready Commerce for WooCommerce