Agentic commerce is going to reshape consumer journeys – and that has some unexpected advantages for messaging providers and MNOs
Agentic commerce is going to radically reshape how consumers find and buy things, but it is also going to reshape the messaging market. Think about it: agentic commerce basically means AI agents that can find, compare, and buy things on behalf of the user – often via chat. That’s directly in the slipstream of RCS and WhatsApp, which are already the main “commerce chat” channels.
Morgan Stanley thinks that by 2030 nearly half of US online shoppers will use AI shopping agents, adding $115bn to US e-commerce alone. Visa finds 47% of consumers are interested in using agents for commerce, while 92% of shoppers who already use AI for shopping say it improves their experience, with a 1,200%+ spike in traffic from GenAI sites to merchants in 2024.
If we look at how consumers view messaging, we find that, according to Salesforce 66% of online adults prefer messaging apps like WhatsApp to talk to businesses. WhatsApp itself is the monster channel: about 3 billion unique active users and around 140 billion messages per day as of 2024, according to DoubleTick.
But RCS Business Messaging is catching up: in retail and ecommerce, with global communications platform Infobip forecasting more than 3.9 billion messages exchanged between brands and customers this 2025 shopping season, marking a 15% increase year on year. In fact, its data shows that RCS will be the fastest-growing channel in 2025, with volumes rising 161% on Black Friday and 269% on Cyber Monday worldwide. SMS remains the leading retail channel with nearly 2.7 billion interactions.
So, combined with agentic AI and you have big money, lots of automation and buyers that prefer messaging.
What it means for messaging
In the agentic AI world, AI agents need a chat surface where they can ‘live’ and talk to users, ideally with buttons and carousels and forms, as well as built in access to payments. RCS and WhatsApp – in fact all of today’s rich business messaging – tick all these boxes.
For that reason, RCS is becoming something of an agent-native channel. RCS now has more than 1 billion monthly active users and retail/e-commerce RCS usage is exploding by some 684% growth, according to data from Master of Code Global.
Other Ecommerce-focused research estimates RCS adoption jumped five times in 2024 and RCS is forecast to reach 7% of global business messaging revenue by 2029 and 18% of operators’ business messaging revenue (up from 3% in 2024).
It also drives consumer engagement, with people 35 times more likely to read a rich RCS message than an email, 74% of buyers say they’re more inclined to engage brands via RCS and some campaigns show 80% conversion rates and 2–3× higher CTR vs other channels.
What does this mean in an agentic marketing world? Agents will send proactive RCS messages: “I’ve found you a cheaper flight; tap here to switch”, “It’s time to reorder coffee; pick one of these three bundles” when shoppers have been looking, for example.
Combining this with payments will add even more power. The Agentic Commerce Protocol (ACP) from OpenAI and Stripe is an open standard letting AI agents execute purchases across merchants and payment processors. An ACP-enabled agent could use RCS as the front-end surface: show ACP-driven offers in an RCS carousel, collect user confirmation and shipping preferences, then complete checkout over card rails or wallets.
What about WhatsApp?
WhatsApp, meanwhile, is already the default conversational commerce rail in many markets. Looking at the hard numbers, WhatsApp has some 3 billion unique users, sending around 140 billion messaged per day. Meanwhile some 50 million busiensses use WhatsApp Business reaching nearly 800 million active users per month. In fact, according to Mobilesquared, CPaaS providers expect WhatsApp to become the most trusted and effective omnichannel CPaaS channel by 2025, overtaking SMS, as brands flood to richer, less fraud-prone messaging.
And there are already examples of WhatsApp agent experiments that show where this could be heading. JioMart in India sells groceries over WhatsApp, with full end-to-end shopping inside WhatsApp chat. Reliance reported a seven-fold increase in monthly orders and six-fold month-on-month growth in new customer conversions after launching a WhatsApp AI shopping experience.
Also in India, boAt, a D2C electronics seller, saw conversational commerce via WhatsApp generated $236,000 revenue in six months and recovered 185,000 abandoned carts via automated WhatsApp flows.
This is already agentic commerce, just with less autonomy: the agent provides options, but the user taps to confirm. As Visa notes, we’re moving from “AI recommends” toward “AI initiates” and eventually “AI transacts” under tight controls.
What does it mean for MNOs?
For MNOs this shift to rich messaging as part of the broader Agentic Commerce market is a mixed blessing. On the one hand it could generate significant RCS A2P revenues for carriers, with research by Juniper and Omdia forecasting RCS to rise from around 1 to 3% of business messaging revenue in 2024 to some 7 to 18% by 2029, with operators capturing that share.
There is also money to be made from verified, SIM-anchored identity and fraud controls, which are crucial when AI agents start transacting autonomously. Visa, Stripe and others are explicitly emphasising the need for stronger identity/guardrails in agentic commerce, according to McKinsey.
The opportunites for MNOs are there. They can offer enterprises an ‘agent-ready RCS’ product: pre-built templates, conversational flows and easy integration with ACP/agent platforms.
They can also bundle RCS with network identity APIs, location, and fraud tools so agents can do smarter things, such as delivery options based on cell-tower location, SIM-based authentication.
There are also opportunities for billing and carrier wallet integration too. Agents need ways to pay. Visa, AWS, Amazon, Stripe, et al are racing to provide card/wallet rails for agentic commerce and, as we have seen on page 18 mobile money and carrier billing could be exposed to agents – within spends caps and SCA – giving MNOs a slice of automated micro-transactions.
However, there are risks for MNOs. The ever present threat of becoming ‘dumb pipes’ looms large when OTT messaging gets any sort of boost. WhatsApp, Instagram, Messenger and future super-agents – Meta AI, Amazon agents, – all sit above the network. If MNOs don’t aggressively own RCS and network APIs, agentic commerce will route around them via card schemes, wallets and OTT messaging.
Apple’s RCS support ensures rich messaging is available by default on both major OSs; if operators don’t move, most of the value will be captured by cloud CPaaS and card networks.
The big upside for MNOs lies in leveraging RCS, payments and identity. Win this and MNOs are suddenly in a very strong position in the Agentic commerce world.
*******************************************************
Channel realities: WhatsApp versus RCS for telemedia
So, what are telemedia companies to make of all this? We have already seen in previous issues how AI and agentic commerce shift how content and services are going to be marketed, but when it comes to messaging, what should they choose? Here we offer a simple view.
- WhatsApp
- Huge reach, great UX, and already proven for commerce. But WhatsApp has strict policies on gambling, adult content, and subscription-style monetisation, and Meta is reducing scope for third-party “AI as the product” bots.
- For telemedia, WhatsApp will be strongest in utility, support, and lightweight upsell (“add this tip service to your existing sports subscription?”), less so for the classic “premium SMS” style billing.
- RCS
- Verified business senders, rich content and carousels, and most importantly operator-controlled policies and billing models.
- Telemedia players can rebuild classic VAS models using RCS + carrier billing, but now with app-like UX and AI-powered personalisation (e.g. an RCS agent recommending content packs, reminding about expiring passes, etc.).
In short, telemedia that leans into RCS + operator billing + agent APIs can grow. Telemedia that stays stuck in one-way PSMS and static WAP pages will get bypassed by smarter agents using card/wallet rails. You have been warned.
*******************************************************
