Today on The Merchant Desk, we're tracking the architectural choices being made for the next wave of commerce. Two major M&A deals land — Nuvei buying Payoneer and Salesforce acquiring AI agent platform Fin — while the Visa and ChatGPT integration we've been monitoring officially goes live, making agentic commerce a reality. At the same time, a new strategic framework for pricing in the 'agent economy' provides a playbook for how to actually build a business on these new rails.
Following Visa's rollout of its agentic commerce toolstack last week, its high-profile integration with OpenAI is now live in production. The partnership allows ChatGPT to search for products, compare options, and autonomously complete purchases on behalf of users within predefined limits, secured by Visa's existing tokenization rails.
Why it matters
This moves agentic commerce from a conceptual PowerPoint slide to a live production environment at massive scale. While we've tracked the build-up, this confirms the payment rails are officially connected to the world's most popular LLM. It forces every e-commerce merchant and payment provider to confront the reality of a software-driven buyer. The strategic questions now shift from 'if' to 'how': how to optimize product data for agent discovery and how to handle attribution when the 'customer' is an AI.
According to Visa's latest Stay Secure study, 77% of South African consumers use AI for shopping assistance, such as for product recommendations and comparisons. However, a significant trust gap remains at the point of transaction, with only 23% stating they would trust an AI agent to complete a purchase on their behalf. The study also found that 37% of consumers experienced a financial scam in the last year, many originating on social media platforms.
Why it matters
This data puts a fine point on the central challenge for agentic commerce in South Africa: the 'last mile' of trust. While consumers are comfortable using AI as a research tool, they are hesitant to cede financial control. For payment and merchant tech operators, this highlights the critical need to build robust security, transparent user controls, and clear liability frameworks into any AI-driven checkout experience. Winning in this space will be as much about building confidence as it is about building technology.
Woolworths is overhauling its customer loyalty strategy, phasing out its long-standing WRewards and MySchool programs in favor of a new, integrated platform called MyDifference. According to reports on Monday, the new program is designed to offer more personalized rewards, leverage customer data for deeper insights, and provide greater transparency in the company's charitable contributions, marking a significant strategic shift for the South African retailer.
Why it matters
This is a major move in the South African retail landscape. Woolworths is betting that a more modern, data-driven, and personalized loyalty engine can drive engagement more effectively than its legacy programs. For competitors and the broader retail tech ecosystem, this signals that the bar for customer loyalty is rising. Success will depend on the ability to translate vast amounts of shopping data into relevant, timely offers that genuinely influence consumer behavior, moving beyond simple points and discounts.
Google’s AI-powered Shopping Graph, enhanced by its Gemini 2.5 model, is beginning to erode Amazon's long-held dominance in product search, especially for high-value purchases among 25-44 year olds. A Monday report, citing early data, shows a significant increase in Google Shopping-referred sessions and higher conversion rates. This is forcing brands and e-commerce aggregators to reallocate advertising budgets toward optimizing for Google's AI Overviews and rich product results.
Why it matters
This marks a structural shift in the e-commerce landscape. For years, Amazon was the default starting point for product discovery; Google is now leveraging its AI prowess to intercept that journey earlier. For merchants, this means the 'front door' to their store is moving. It's no longer just about optimizing for Amazon's A9 algorithm but also ensuring product data is structured for Google's AI to find and recommend, fundamentally changing go-to-market and ad spend strategies.
Building on the Zalando 90% AI-generated Concept Store content milestone we noted recently, Co-CEO David Schröder detailed further operational impacts on Monday: a 10% reduction in returns driven by improved AI sizing, and 10 million active users on its shopping assistant. Schröder also warned that Europe risks ceding its commerce infrastructure if it fails to support scaling tech companies.
Why it matters
Zalando provides one of the clearest case studies to date of AI moving beyond hype to deliver measurable, bottom-line results in e-commerce. The 10% reduction in returns is a particularly powerful signal, as returns are a massive cost center for fashion retail. This demonstrates how practical AI applications in fit, discovery, and content are becoming table stakes for competitive online retail, offering a concrete model for other merchants on where to focus AI investment.
Online furniture retailer Wayfair has implemented an AI system, in partnership with OpenAI, to automate the correction of flawed product listings and handle routine supplier support requests. According to a report on Monday, the system has already corrected 2.5 million product attribute tags and now processes 41,000 supplier support tickets monthly, significantly reducing manual work for Wayfair's team and improving data quality across its vast catalog.
Why it matters
This is a prime example of AI being deployed to solve a messy, unglamorous, but operationally critical problem in e-commerce: data hygiene at scale. Poor product data leads directly to bad customer experiences and lost sales. By automating this cleanup, Wayfair not only cuts costs but also improves product discoverability and reduces 'not as described' returns. It's a practical, high-ROI application of AI that any large-scale merchant can learn from.
Nuvei announced on Monday a definitive agreement to acquire Payoneer for approximately $2.75 billion in cash. The deal combines Nuvei's acquiring and technology platform with Payoneer's B2B payments network, creating an entity with a projected $3 billion in annual revenue and processing over $500 billion in annual volume. The explicit goal is to build a unified financial infrastructure supporting local and cross-border commerce for merchants of all sizes.
Why it matters
This is a significant consolidation move in the global payments space, directly challenging players like Stripe and Adyen. The combination creates a formidable competitor with deep reach into both enterprise acquiring and the long tail of SMB/freelancer cross-border payments. For operators in markets like South Africa, this signals that the M&A landscape for payment infrastructure is heating up, with scale and comprehensive global-to-local capabilities becoming the key competitive battleground.
In a sweeping regulatory overhaul announced on Monday, the Central Bank of Nigeria (CBN) has directed all banks, fintechs, and payment service providers to store transaction data generated within Nigeria on local servers by January 1, 2027. The new rules also introduce market share limits to curb concentration, capping any single institution at 25% of card issuing or merchant acquiring volume. The regulations also mandate the disclosure of Ultimate Beneficial Ownership (UBO) to increase transparency.
Why it matters
This is a significant move by Nigeria's regulator to assert data sovereignty and reshape the competitive dynamics of its massive payments market. For operators like Paystack, Flutterwave, and major banks, this requires substantial investment in local data infrastructure and a potential re-evaluation of growth strategies to avoid hitting the new market share caps. It signals a broader trend across emerging markets toward greater regulatory control and a desire to foster a more resilient, less concentrated local fintech ecosystem.
An analysis on Monday of Chime's business model reveals a masterclass in building a sustainable consumer fintech. The US neobank has built a multi-billion dollar business by optimizing interchange revenue, driving high-margin volume from its credit builder product, and creating stickiness with its 'SpotMe' micro-advance feature. Despite its valuation normalizing from a 2021 peak of $25 billion to an estimated $10-12 billion, the company is strengthening its balance sheet in preparation for a public listing.
Why it matters
Chime's story provides a crucial playbook for profitability in a sector often criticized for burning cash. For fintech operators, particularly those in competitive neobanking spaces like South Africa, it demonstrates how a disciplined focus on interchange, a sticky credit product, and durable, low-cost deposits can create a viable, self-sustaining financial model. It's a case study in execution and adapting to market realities away from hype-driven valuations.
In a significant policy reversal, the Central Bank of Nigeria (CBN) has prominently featured stablecoins in its new Payments System Vision 2028 (PSV 2028) document. Mentioning stablecoins 68 times, the framework, released June 1, signals a shift from its previous outright ban towards developing a regulated environment for their use. The goal is to leverage stablecoins to address foreign exchange challenges, lower remittance costs, and improve cross-border payments. The proposed rules require 100% fiat reserves and real-time regulatory visibility.
Why it matters
This is a watershed moment for digital assets in Africa. Nigeria, as the continent's largest economy, is setting a potential precedent for other regulators to move from prohibition to supervised integration. For fintech operators, this opens a massive, previously closed-off market for stablecoin-based treasury management, remittances, and trade finance. The detailed framework, including the 'regtech node,' suggests the CBN is serious about building a robust, transparent ecosystem rather than just paying lip service.
A new analysis outlines a strategic framework for pricing and positioning in the emerging 'agent economy,' arguing for a shift from flat-rate subscription software to outcome-producing, credit-priced agent systems. Published on Monday, the playbook details choices across five domains: stack position (infrastructure vs. application), value architecture (what you sell), pricing architecture (how you charge), competitive positioning, and trust/governance, advising companies to build for the long-term economics of autonomous systems from day one.
Why it matters
This is a critical strategic document for any operator building or selling software. It provides a clear-eyed framework for navigating the most significant business model shift in decades: from selling tools (SaaS) to selling outcomes (AI agents). For a fintech operator focused on merchant services, understanding how to price for value delivered by an AI—rather than per seat or per API call—is fundamental to designing profitable, defensible products for the next five years. The emphasis on 'Phase 3 economics' is a direct challenge to the incrementalism common in enterprise SaaS.
Salesforce announced on Monday its plan to acquire Fin, an autonomous AI agent platform specializing in customer query resolution, for approximately $3.6 billion. The acquisition is a strategic move to bolster Salesforce's 'Agentforce' platform and accelerate the deployment of sophisticated AI agents within enterprise customer service and sales operations. Fin's platform is known for resolving customer queries across multiple channels using its proprietary AI model, Apex.
Why it matters
This is a major bet on the future of enterprise operations moving from human-assisted to AI-led. By acquiring a mature AI agent company instead of building from scratch, Salesforce signals the urgency to own this layer of the stack. For operators in the B2B SaaS and merchant tech space, this validates the market for vertical AI agents that can replace or augment entire workflows, not just assist with single tasks. It also raises the competitive stakes for smaller players aiming to provide AI-powered customer service.
Consolidation in Payments and AI Nuvei's $2.75B acquisition of Payoneer and Salesforce's $3.6B purchase of Fin signal a major trend of consolidation. In payments, the goal is to create unified global platforms for cross-border commerce, while in enterprise software, the race is on to acquire mature AI agent capabilities to accelerate the shift to autonomous operations.
Agentic Commerce Moves to Production Visa's integration with ChatGPT is now live, marking a critical shift from demos to real-world agentic transactions. This move, combined with a new strategic framework on pricing for the 'agent economy,' shows the infrastructure and business models for AI-driven commerce are rapidly maturing.
AI Adoption Accelerates in Retail Operations Major retailers like Zalando, Wayfair, and Amazon India are reporting significant, measurable ROI from AI deployed in core operations. Use cases have moved beyond chatbots to include automated product listing correction, AI-generated merchandising content, and dramatic reductions in returns, proving the technology's value at scale.
Regulatory Frameworks Adapt to New Realities Central banks in emerging markets are actively shaping new payment landscapes. Nigeria is a key example, with new rules mandating local data storage, imposing market share caps to curb concentration, and, in a major reversal, formally exploring stablecoins as part of its national payment vision for 2028.
The Trust Gap in AI Commerce New studies from Visa in Nigeria and South Africa reveal a consistent pattern: while a large majority of consumers (77-88%) use AI for shopping assistance and research, only a small fraction (23-34%) trust AI agents to complete the actual purchase. This highlights a critical trust gap that must be closed for agentic commerce to reach mass adoption.
What to Expect
2026-06-18—MTN MoMo Ghana's interim dividend payable to shareholders.
2027-01-01—Deadline for Nigerian financial institutions to store all domestic payment data on local servers, per new CBN directive.
How We Built This Briefing
Every story, researched.
Every story verified across multiple sources before publication.
🔍
Scanned
Across multiple search engines and news databases
452
📖
Read in full
Every article opened, read, and evaluated
196
⭐
Published today
Ranked by importance and verified across sources
12
— The Merchant Desk
🎙 Listen as a podcast
Subscribe in your favorite podcast app to get each new briefing delivered automatically as audio.
Apple Podcasts
Library tab → ••• menu → Follow a Show by URL → paste