TL;DR
- Core definition: Agentic commerce is trade mediated by autonomous AI agents that discover, evaluate, negotiate, and complete purchases on a shopper’s behalf, which means your store now has to sell to software, not just to people.
- The shift that matters: The hard part is not that an agent can read your catalog; it is whether it can actually complete a checkout, and in our experience those two things fail independently all the time.
- What to do first: Treat machine-readability as infrastructure, not marketing. Publish a conformant UCP manifest, test a real agent-driven purchase end to end, and measure agent conversion as a first-class KPI.
We got a call last spring from a merchant who was convinced agentic commerce was hype. His storefront had a clean product feed, a well-reviewed Shopify theme, and strong human conversion. Then an AI shopping agent tried to buy a $140 pair of boots from him on behalf of a customer, and it got as far as adding the item to cart before silently giving up at a custom address-validation step that a human would have breezed past. The sale did not fail with an error anyone saw. It just evaporated. No abandoned-cart email, no analytics event, nothing. That quiet failure is the clearest agentic commerce definition we can offer: a world where your best-looking store can still be invisible or unusable to the buyers that matter most, and you will not even know it is happening unless you go looking.
This article is our attempt to give you the working definition we actually use with clients, not the dictionary version. We build Universal Commerce Protocol (UCP) infrastructure and implement agentic commerce for real merchants every week, so we are going to walk through the ten things we think every merchant must understand about what agentic commerce is, how it works, how it differs from the ecommerce you already run, and what it takes to win in it. We ordered these roughly by impact, strongest first. By the end you should have a far more precise mental model than “AI does the shopping now,” plus a concrete framework and a measurement plan you can run this quarter.
1. Agentic Commerce Is Trade Where an Autonomous Agent Is the Buyer
Start with the cleanest version of the agentic commerce definition: it is commerce in which an autonomous AI agent, acting under a human’s delegated authority, carries out the buying process end to end. The human sets the intent (“find me running shoes under $120 that ship by Friday”), and the agent does the discovery, comparison, decision, and transaction. The agent is not a search box and it is not a chatbot that hands you a link. It is a software buyer with a budget, a goal, and the ability to act.
What makes this different from every automation wave before it is autonomy over the full funnel. Predictive AI recommended products to a human who still clicked buy. Agentic AI clicks buy. We have written before about this progression in the third wave from predictive to agentic AI in ecommerce, and the practical consequence is blunt: your conversion optimization work, your persuasive copy, your urgency banners, most of it was built to move a human nervous system. An agent does not feel urgency. It reads structured data, evaluates against constraints, and transacts or moves on.
Best for understanding the stakes: think of the agent as a procurement officer who evaluates thousands of vendors in milliseconds and has zero brand loyalty unless your data earns it.
- Delegated authority: An agent acts for a human, which means trust and verification move to the center of the transaction.
- Full-funnel autonomy: Discovery through checkout happens without a human in the loop for most steps.
- No emotional levers: Scarcity and persuasion tactics lose most of their power against a constraint-solver.
- Data as storefront: Your machine-readable data becomes the thing being shopped, not your visual design.
- Silent outcomes: Wins and losses often leave no trace in human-oriented analytics.
2. “Agentic” Means Goal-Directed Autonomy, Not Just Automation
Merchants ask us constantly what “agentic” actually means in e-commerce, usually because the word gets slapped onto anything with a model behind it. Here is the distinction we tell clients: automation follows a fixed script, while an agent pursues a goal and chooses its own steps to get there. A rules-based reorder bot is automation. A system that is told “keep this customer’s pantry stocked within a $200 monthly budget,” then decides what to buy, when, and from whom, is agentic.
The agentic part is the reasoning loop. The agent observes the available options, plans, acts, observes the result, and adjusts. When a checkout step fails, a real agent may retry, substitute a product, or switch merchants entirely. That adaptability is exactly why the boots sale we opened with was so dangerous: the agent did not crash, it simply decided our merchant’s friction was not worth solving and quietly chose a different store. Automation breaks loudly. Agents reroute quietly.
Standout feature: goal-directed substitution. If your store presents any ambiguity the agent cannot resolve, a competitor who presents clean, complete, machine-readable data wins the order by default.
- Script versus goal: Automation executes predefined steps; agents choose steps to satisfy an objective.
- Reasoning loop: Observe, plan, act, observe, adapt, repeated until the goal is met or abandoned.
- Graceful rerouting: Friction does not stop an agent; it redirects it to a cleaner competitor.
- Constraint satisfaction: Budget, deadline, specs, and preferences are hard filters, not suggestions.
3. It Works Through a Discovery, Evaluation, and Transaction Pipeline
To answer “how does agentic commerce work” honestly, you have to look at the pipeline the agent runs, because each stage can succeed or fail independently. The first stage is discovery: the agent needs to find that you exist and understand what you sell in a structured way. The second is evaluation: it compares your offer against alternatives on price, availability, shipping, policies, and trust signals. The third is transaction: it actually completes the purchase, including payment and fulfillment details, on behalf of its human.
Most merchants we audit are decent at stage one and quietly broken at stage three. A product feed or a schema markup block gets you discovered. It does nothing to guarantee that an agent can push a real order through your checkout with a delegated payment method and a validated address. This is why we built our implementation practice around the whole pipeline rather than just the catalog. The discovery layer matters, and we go deep on it in our piece on the Universal Commerce Protocol well-known discovery layer, but discovery without a completable checkout is a beautiful shop window bolted to a locked door.
What this achieves when you map it: you stop guessing where agent sales leak and start instrumenting each stage so you can see exactly where an order dies.
- Discovery: Can an agent find you and parse your catalog without a human interpreting the page?
- Evaluation: Are price, stock, shipping, and returns exposed as structured, comparable data?
- Transaction: Can an agent complete payment and fulfillment end to end, not just reach the cart?
- Verification: Does your store confirm the agent’s authority and the human’s identity cleanly?
- Confirmation: Does the agent receive machine-readable proof the order actually went through?
4. The Buyer You Optimize For Is Now Software, Not a Person
This is the reframe that unsettles most merchants, and it should. For twenty years, conversion rate optimization has meant understanding human psychology: color, trust badges, social proof, loading speed as it feels to a person. In agentic commerce, the entity deciding whether to buy is a model reading data. A conformant, complete, unambiguous data layer is the new above-the-fold hero image.
We measure this directly now, and we treat agent conversion rate as its own number separate from human conversion. The dynamics are different enough that the two can diverge sharply. We broke down how we track and improve this in agentic commerce conversion rate with UCP, and the short version is that a store can have a strong human conversion rate and a near-zero agent conversion rate at the same time. The agent is not confused by your layout; it is blocked by your missing return policy field or your human-only CAPTCHA.
Best for CRO teams: stop A/B testing button colors for a quarter and instead test whether a scripted agent can complete three representative purchases on your site. That single test usually surfaces more revenue risk than a month of human-funnel tweaks.
- New buyer profile: A model reading structured data, not a human reading a page.
- Data over design: Completeness and disambiguation beat aesthetics for the agent audience.
- Separate KPI: Agent conversion rate deserves its own dashboard tile, tracked independently.
- Divergence risk: Strong human conversion can mask a broken agent funnel entirely.
5. Agentic Commerce Differs From Traditional Ecommerce in Who Holds Context
The seed question we hear most is how agentic commerce is different from traditional ecommerce, and the deepest difference is where the context lives. In traditional ecommerce, the human carries the context: they know their size, their budget, their deadline, and they apply judgment to fuzzy information on your pages. In agentic commerce, the agent holds and enforces that context, and it will not tolerate the fuzziness a human forgives.
A human sees “ships in 1-2 weeks” and shrugs. An agent with a hard “must arrive by Friday” constraint treats that same string as a disqualifier unless your data lets it compute a real delivery date. A human reads a sizing paragraph and estimates. An agent needs structured size and fit data or it will not commit. The move from human-tolerated ambiguity to agent-enforced precision is the single biggest operational gap we fix for clients. It is also why we think the distinction between doing this properly versus bolting on a feed is so consequential, which we argue in detail in UCP versus manual implementation.
Standout feature of the shift: zero tolerance for ambiguity. Every field a human would have guessed at becomes a field you must specify or lose the sale.
- Context holder: Humans carry context in their heads; agents enforce it from your data.
- Ambiguity cost: Vague copy that humans forgive becomes a hard disqualifier for agents.
- Judgment versus rules: Human judgment fills gaps; agents require the gaps be filled in data.
- Precision requirement: Delivery dates, sizing, and policies must be computable, not described.
6. UCP Is the Protocol That Makes Agent Transactions Actually Work
If agentic commerce is the “what,” the Universal Commerce Protocol is a big part of the “how it works in practice.” UCP is an open standard that gives agents a consistent, machine-readable way to discover a merchant, understand the catalog and policies, verify trust, and complete a transaction. Without a shared protocol, every agent-to-merchant interaction is a bespoke integration, which does not scale to a web of millions of stores and thousands of agents.
We are opinionated here because we build on it: a protocol-level approach beats per-platform hacks, and the specifics matter. We walk through the full specification in our agentic commerce protocol specification deep dive, and we also track the standards competition honestly in UCP versus ACP: which standard will rule the agentic web in 2026. Our take is that the winning merchants will adopt the emerging open standard early rather than waiting for the dust to settle, because the cost of being invisible to agents compounds.
The agentic commerce definition that matters to a merchant is not philosophical; it is operational: can a real agent complete a real checkout on your store today, or not?
For adoption context, UCP Checker, which independently monitors more than 20,120 storefronts, reports that roughly 81% pass full UCP validation, with 16,376 verified. Read that carefully, though, because it skews heavily toward Shopify stores and does not mean 81% of all ecommerce has UCP. More importantly, a conformant UCP manifest is not the same as an agent being able to complete a real checkout, which is exactly the gap we spend most of our implementation time closing.
- Shared standard: UCP replaces bespoke per-agent integrations with one consistent interface.
- Protocol over hacks: Standards-based adoption scales where platform-specific patches do not.
- Early mover advantage: Being agent-visible early compounds while competitors wait.
- Validation caveat: Passing validation is necessary but not sufficient for a completable checkout.
The Agent-Ready Store Framework
Here is the framework we run with clients when they move from “we read an article about agentic commerce” to “we want agents buying from us this quarter.” It has four steps, strongest-leverage first.
Step one, Expose a conformant discovery layer. What this achieves: it makes your store findable and parseable by agents in the first place, which is the precondition for everything else. Publish a valid UCP manifest at your well-known endpoint, with complete catalog, pricing, availability, shipping, and return policy data expressed as structured fields rather than prose. We detail the store-level mechanics for Shopify merchants in our Shopify UCP guide for 2026.
Step two, Prove a completable transaction. What this achieves: it confirms that an agent can get all the way through checkout, not just to the cart, which is where most stores silently fail. Run a scripted agent through three representative purchases, including one with a shipping deadline and one with a return-policy-sensitive item, and fix every point where the agent stalls.
Step three, Harden the trust layer. What this achieves: it lets your store verify the agent’s delegated authority and the buyer’s identity without introducing human-only friction like image CAPTCHAs. We go deep on this in UCP security: building the trust layer for agentic commerce, because trust failures are the most expensive kind: they block otherwise-ready transactions.
Step four, Instrument and iterate. What this achieves: it turns agent commerce from a black box into a measured funnel so you can improve agent conversion deliberately rather than hoping. Log every agent interaction by pipeline stage, watch discovery-to-transaction drop-off, and treat the agent funnel like the paid-funnel you already optimize.
- Discovery first: A conformant manifest at your well-known endpoint is step zero, not optional.
- Checkout proof: A completed test purchase by an agent beats any validation badge.
- Trust without friction: Replace human-only gates with agent-verifiable authority checks.
- Measured iteration: Log by stage so you can see and fix drop-off, not guess at it.
Win the Agent Shelf With UCPhub
If this framework makes it obvious your store is not yet agent-ready, that is the normal starting point, and it is exactly the gap we close. At UCPhub we build and operate the Universal Commerce Protocol infrastructure that lets AI agents discover, trust, and buy from your store without the silent failures we described, and we instrument the whole funnel so you can see agent conversion as a real number. The merchants who move now get to be the clean, completable option an agent chooses by default while competitors are still debating whether any of this is real. Talk to our team about an agent-readiness audit and we will run a real agent through your checkout and show you exactly where it stalls.
7. Trust and Verification Become First-Class Commerce Problems
In human commerce, trust is mostly ambient: reviews, a familiar logo, an SSL padlock. In agentic commerce, trust has to be explicit and machine-verifiable, because two pieces of software are transacting and each needs to know the other is legitimate. The agent needs to prove it holds real delegated authority from a real human with real payment capacity. The merchant needs to confirm that without leaking data or blocking the transaction.
We have found that trust is where otherwise-perfect implementations fail in production. A store can have a flawless catalog and still reject every agent at checkout because its fraud rules were tuned for human behavioral signals that agents do not produce, like mouse movement and dwell time. The fix is a verification layer designed for machine actors, which is a core reason protocol-level security matters so much. Getting this wrong does not just lose a sale; it can get your store deprioritized by the agent platforms that remember friction.
Best for risk and fraud teams: audit your bot-detection rules before you touch anything else, because the same defenses that stop bad bots will stop the good buying agents you now want.
- Explicit trust: Ambient human trust signals do not translate; verification must be machine-readable.
- Delegated authority: Agents must prove they act for a real, paying human.
- Fraud rule conflict: Human behavioral fraud checks routinely block legitimate agents.
- Deprioritization risk: Agent platforms route around stores that produce friction.
8. The Economics Shift From Attention to Availability
Traditional ecommerce economics run on attention: capture it with ads, hold it with design, convert it with persuasion. Agentic commerce economics run on availability and fit. The agent is not browsing; it is matching constraints against a candidate set and choosing the best objective fit. You do not win by being the loudest; you win by being present, complete, and computably the best answer to the constraint the human handed their agent.
This rewires budget priorities. Spend that used to go entirely to capturing human attention starts to flow toward being the structured, trustworthy, completable option in the agent’s candidate set. We think this is one of the most underappreciated strategic shifts of the next few years, and we lay out the longer arc in the future of UCP and agentic commerce in 2026 and beyond. The merchants who internalize it early will quietly accumulate agent-mediated orders while attention-first competitors keep bidding up ad auctions for human eyeballs that increasingly delegate the buying to software.
Standout feature of the new economics: compounding presence. Being the clean default answer for a given constraint pays off every time that constraint recurs across thousands of agent queries.
- Attention to fit: Winning shifts from loudest to best objective match.
- Budget reallocation: Dollars move from attention capture to agent-readiness and completability.
- Candidate set presence: You must be in the agent’s shortlist, which requires complete data.
- Compounding returns: Being the default answer pays repeatedly across recurring queries.
9. Human and Agent Channels Will Run in Parallel, Not Sequence
A mistake we see merchants make is treating agentic commerce as a replacement that lets them tear down the human storefront. It is not. For the foreseeable future you run both channels at once, and they behave differently enough that you manage them with different playbooks. Your human channel keeps needing great design, fast load times, and persuasive merchandising. Your agent channel needs structured data, completable checkout, and machine-verifiable trust. Starving either one costs you money.
The operational challenge is keeping both in sync from one source of truth. When your inventory, pricing, and policies live in one authoritative place and flow out to both the human-facing storefront and the agent-facing protocol layer, you avoid the nightmare where an agent sells something your human store shows as out of stock. We think of this as dual-channel integrity, and it is a big part of why a protocol approach beats manually maintaining a separate feed that drifts out of date. For the broader planning view, we map this out in our agentic commerce roadmap for 2026.
Best for operations leads: your single biggest risk in a two-channel world is data drift, so invest in one source of truth before you optimize either channel.
- Parallel channels: Human and agent commerce coexist and require different playbooks.
- Single source of truth: One authoritative data layer feeds both channels to prevent drift.
- Dual-channel integrity: Inventory and pricing must match across human and agent views.
- No teardown: Agent-readiness augments the human storefront; it does not replace it.
10. Early, Deliberate Adoption Beats Waiting for the Standard to Settle
The last thing every merchant must understand is a strategic judgment call, and we will be direct about ours: waiting is more expensive than it looks. Merchants tell us they want to hold off until there is one obvious winning standard. We understand the instinct, and we track the competition carefully, including in UCP versus ACP: the battle for the agentic commerce standard. But the cost of being invisible to agents accrues silently, exactly like our boots merchant’s lost sale, and you cannot recover the orders that quietly went elsewhere while you waited.
Our take is that a well-built, protocol-based foundation is largely portable across the leading standards, so the marginal risk of moving now is low and the upside of being an agent-preferred store early is high. The strategic roadmap we use to sequence this with merchants is laid out in the Universal Commerce Protocol 2026 strategic roadmap, and the throughline is simple: build the foundation now, instrument it, and expand coverage as the standards mature. The merchants who do this are already seeing agent-mediated orders that their waiting competitors are structurally unable to capture.
Standout feature of early adoption: option value. A clean protocol foundation keeps your options open across standards instead of locking you into a bet you made too late.
- Silent cost of waiting: Lost agent sales leave no trace, so inaction feels free but is not.
- Portability: A protocol-based foundation transfers across leading standards.
- Low marginal risk: Building now costs little relative to the compounding downside of absence.
- Option value: Early, standards-neutral foundations preserve strategic flexibility.
Measuring Success: 30, 60, and 90 Day Outcomes
A definition is only useful if you can tell whether you are winning. Here is how we structure measurement for merchants moving into agentic commerce, with concrete targets by milestone. Instrument agent conversion rate separately from human conversion from day one; we explain why in agentic conversion rate with UCP.
- Day 30 discovery: A conformant UCP manifest is live at your well-known endpoint and passes validation, with 100% of active SKUs exposed as structured data.
- Day 30 baseline: At least three scripted agent purchases run end to end, and you have a documented list of every stall point found.
- Day 60 completability: Agent checkout completion on your test suite reaches 100% for standard orders, including one deadline-constrained and one return-sensitive case.
- Day 60 trust: Zero legitimate test agents blocked by fraud or bot-detection rules, verified against a fixed test matrix.
- Day 90 conversion: Agent conversion rate is tracked on its own dashboard tile and trending up week over week, with stage-by-stage drop-off visible.
- Day 90 integrity: Inventory and pricing drift between human and agent channels measured and held at zero across daily checks.
- Day 90 revenue signal: First agent-mediated orders attributed and reconciled, proving the funnel produces real money, not just passing badges.
If you are just getting started, prioritize discovery and a single completable test checkout above everything else, because a store that cannot be found or transacted with has nothing else worth optimizing. If instead you are auditing something that already exists, start at the trust and completability layers, since most stores we review discover and display fine but quietly reject agents at checkout, which is the expensive failure that hides behind a passing validation badge. The deeper strategic context for either path lives in our strategic guide to AI-mediated trade in 2026 and our overview of unlocking the Shopify agentic plan in 2026.
Next Steps:
- Run a real agent through your own checkout this week and write down the exact step where it stalls.
- Publish or validate your UCP manifest and confirm 100% of active SKUs are exposed as structured data.
- Stand up an agent conversion rate dashboard tile so the number exists before you try to improve it.
Frequently Asked Questions
What is agentic commerce and how does it work?
Agentic commerce is trade in which an autonomous AI agent, acting on delegated authority from a human, carries out the buying process from discovery through transaction. The human provides intent and constraints, such as a product need, a budget, and a deadline, and the agent handles finding candidate merchants, comparing offers, and completing the purchase. It works by running a pipeline of discovery, evaluation, and transaction against machine-readable data that merchants expose.
In practice, the agent reads structured information about your catalog, pricing, availability, shipping, and policies, then evaluates whether you are the best fit for its human’s constraints. If you are, it attempts to complete the checkout on their behalf, including payment and fulfillment. The critical detail we emphasize with every client is that each stage can fail independently, so a store that is easy to discover can still be impossible to buy from.
What does agentic mean in e-commerce?
In e-commerce, agentic means goal-directed autonomy rather than simple automation. An automated system follows a fixed script; an agentic system is given an objective and decides for itself what steps to take to achieve it, adapting when something does not work. The agent observes, plans, acts, observes the result, and adjusts, which is why it can substitute a product or switch merchants when it hits friction.
This distinction matters commercially because agents reroute quietly around problems rather than failing loudly. A human shopper who hits a confusing checkout step might email support; an agent simply picks a cleaner competitor. The word agentic, then, is a warning as much as a feature: your store is now being evaluated by something that will leave without a word if your data or checkout is ambiguous.
How is agentic commerce different from traditional e-commerce?
The deepest difference is where context and tolerance live. In traditional e-commerce, the human holds the context and tolerates ambiguity, filling gaps in your information with judgment and estimation. In agentic commerce, the agent enforces context from your structured data and tolerates almost no ambiguity, so vague copy that a human forgives becomes a hard disqualifier.
The second difference is the buyer you optimize for. Traditional CRO targets human psychology through design, persuasion, and trust signals, while agentic commerce rewards complete, unambiguous, machine-readable data and a checkout an agent can actually complete. A store can have excellent human conversion and near-zero agent conversion at the same time, which is why we track the two as separate KPIs. The economics also shift from capturing attention to being the best computable fit in the agent’s candidate set.
What are the key characteristics of agentic commerce?
The key characteristics we emphasize are autonomy, delegation, precision, and verifiability. Autonomy means the agent carries the full funnel without a human in the loop for most steps. Delegation means the agent acts on authority granted by a human, which pushes trust and identity to the center of the transaction. Precision means ambiguity is intolerable: delivery dates, sizing, and policies must be computable, not merely described.
Verifiability ties it together, because two software actors transacting need machine-readable trust rather than the ambient signals humans rely on. Add to those the fact that outcomes are often silent, with wins and losses leaving no trace in human-oriented analytics, and you get a commerce environment that behaves very differently from the one most merchants have optimized for over the past two decades.
Do I need UCP to participate in agentic commerce?
You need some machine-readable, standards-based way for agents to discover, trust, and transact with your store, and the Universal Commerce Protocol is the open standard we build on because it addresses the full pipeline rather than just the catalog. You can attempt bespoke per-agent integrations, but that approach does not scale across a web of many agents and millions of stores, and it tends to drift out of sync with your real inventory and pricing.
Our honest position is that a protocol-based foundation is largely portable across the leading standards, so adopting one now is low risk relative to the silent cost of being invisible to agents. We track the standards competition closely so our clients are not locked into a premature bet, but we do not advise waiting, because the orders you lose while standing still leave no record and cannot be recovered.
Does passing UCP validation mean agents can buy from my store?
No, and this is one of the most important nuances we teach. Passing validation means your manifest is well formed and conformant, which is necessary but not sufficient. A conformant UCP manifest is not the same as an agent being able to complete a real checkout. For context, UCP Checker, which independently monitors more than 20,120 storefronts, reports that roughly 81% pass full UCP validation with 16,376 verified, but that sample skews heavily toward Shopify and says nothing about whether those stores can actually complete an agent-driven purchase.
In our implementation work, the gap between a passing badge and a completable checkout is where most of the real effort lives. Fraud rules tuned for human behavior, human-only CAPTCHAs, custom address-validation steps, and payment flows that assume a human all block legitimate agents despite a valid manifest. The only reliable test is to run a real agent through a real purchase end to end and fix every point where it stalls.
How do I measure whether my agentic commerce efforts are working?
Track agent conversion rate as a first-class KPI, separate from human conversion, and instrument your agent funnel by pipeline stage so you can see discovery-to-transaction drop-off. In the first 30 days, confirm a live, validated manifest with 100% of active SKUs exposed and at least three documented end-to-end agent test purchases. By 60 days, aim for 100% agent checkout completion on your standard test suite and zero legitimate agents blocked by fraud or bot rules.
By 90 days, you want agent conversion trending up week over week with stage-level visibility, zero inventory and pricing drift between your human and agent channels, and your first attributed, reconciled agent-mediated orders. That last milestone is the one that proves the funnel produces real revenue rather than just a passing validation badge, which is the distinction that separates merchants who are genuinely agent-ready from those who only look it on paper.
Sources
- Universal Commerce Protocol: the discovery layer for agentic commerce
- The third wave: from predictive to agentic AI in ecommerce
- Agentic commerce conversion rate with UCP
- UCP versus manual implementation: the strategic guide to agentic commerce in 2026
- Agentic commerce protocol: the official 2026 UCP specification deep dive
- UCP versus ACP: which standard will rule the agentic web in 2026
- Shopify UCP guide 2026: enabling agentic commerce for your store
- UCP security: building the trust layer for agentic commerce in 2026
- The future of UCP and agentic commerce in 2026 and beyond
- Agentic commerce roadmap 2026
- UCP versus ACP: the battle for the agentic commerce standard
- Universal Commerce Protocol 2026: the strategic roadmap for agentic commerce
- Agentic conversion rate with UCP
- Agentic commerce 2026: the strategic guide to AI-mediated trade
- Shopify AI automation: unlocking the agentic plan in 2026
- Talk to the UCPhub team


