Read
Structured data and llms.txt agents can parse
Volve Technology
Agent Readiness for VTEX
Audits, fixes, and monitoring for VTEX IO and FastStore
When your customer's agent comes to buy, will your store be ready?
Pass/fail audits on VTEX IO and FastStore. No vague score to chase.
The shift
More store visits come from agents, not people
Human
72%
Agent
28%
Illustrative traffic mix — directionally shifting
How agents shop
Same signals, every assistant
ChatGPT
Claude
Perplexity
Gemini
Shopping bots
| Layer | What agents consume | Most stores | We implement |
|---|---|---|---|
| llms.txt | Site summary at domain root | Missing | Core |
| Structured data | JSON-LD on Home, PLP, PDP | Sparse | Core |
| A11y tree | Named search, cart, filters | Gaps | Core |
| CLS | Stable layout on load | Mixed | Core |
| WebMCP | Callable search, cart, order tools | Absent | Optional |
We audit your store. We do not score individual AI brands.
13.3
Agentic Browsing is in Lighthouse's default config
Experimental
Default config
May 2026
May 2026
PSI rollout
(gradual)
(gradual)
The gap
Most sites fail on first run
Fail audits
Pass
Directional figure. Almost nobody in VTEX owns this yet.
What we sell
AI agent readiness for VTEX commerce
Understand
Named controls on search, cart, and filters
Transact
Optional WebMCP tools when the spec is ready
What we fix
Five technical fronts
llms.txt
Machine-readable summary at your domain root
Low riskQuick win
Structured data
JSON-LD on Home, PLP, and PDP
Low riskSEO aligned
Accessibility tree
Named search, cart, and filters
Medium effortHelps users too
WebMCP
Tools agents call directly
OptionalHigh value
Layout stability
CLS on key templates
Low riskCore Web Vitals
Core package
✓
✓
✓
✓
What we implement first on most projects
llms.txt
Agents learn your catalog from one file
Structured data
Products and nav in JSON-LD
Accessibility tree
Named controls on key interactions
Layout stability
CLS hygiene we already know well
WebMCP is not part of the starting package. Covered on the next slide if relevant.
Optional later
WebMCP when you are ready — never required to start
WebMCP
Search, cart, and order lookup as callable tools for agents.
OptionalSpec still moving
We do not bundle WebMCP into the initial audit or core implementation.
If the audit shows you need it and the spec has settled, we scope it separately.
No pressure to buy it on day one. This is not a hidden upsell.
Delivery
How we work
1
Audit
Agentic Browsing on Home, PLP, PDP, cart, search
2
Scope
Map failures to fronts; quote by platform
3
Implement
FastStore kit or VTEX IO app
4
Verify
Re-run audit; show what changed
5
Monitor
Scheduled re-runs as the spec moves
Your store today
2 / 5 pass
→
After implementation
5 / 5 pass
Example audit delta — actual numbers come from your store
Platforms
FastStore vs VTEX IO
| FastStore | VTEX IO | |
|---|---|---|
| Stack | Next.js, open | Store Framework, closed |
| Delivery speed | Fast |
Slower |
| Repeatability | High (drop-in kit) | Medium (custom app) |
| Typical ticket | Lower | Higher |
| Best for | Rollout across stores | Premium, defensible work |
Same outcome. Different delivery model.
Engagement
Tier 1
Tier 2
Tier 3
Three tiers
Audit
Report and roadmap. Low cost, easy yes.
Implementation
Fix the fronts you need, scoped by platform.
Retainer
Keep passing as Lighthouse checks change.
Our promise
We will not oversell this
Standards still moving
WebMCP and llms.txt are proposals. We build in modules so pieces can swap when the spec changes.
Pass or fail, per check
No vanity score to chase. Each item is something we can show you fixed.
Why we sell a retainer
The checks will change. A one-off project gets you passing today. The retainer keeps you passing after the next release.
Why Volve
VTEX is what we do
VTEX IO + FastStore
Both stacks in production weekly
Performance
Core Web Vitals already in our wheelhouse
Brazil & LatAm
Active VTEX clients in the region
Automated audits
Reproducible Lighthouse runs every time