What is the Stripe and Thoughtworks report about?

The joint report by Stripe and Thoughtworks explores how agentic commerce is advancing in Australia and New Zealand. It reveals that while consumer adoption and AI-referred traffic are rising, a major execution gap remains – with only a small fraction of organizations operating live integrations. The report offers business leaders a practical pathway to bridge this gap by establishing AI-ready data architectures and a dual strategy that maintains visibility to AI engines without losing direct customer relationships.

What is the methodology behind the Stripe and Thoughtworks report?

The report is based on research conducted by YouGov in July 2026, surveying 1,505 consumers (1,002 Australia and 503 New Zealand) and 206 senior decision-makers (155 Australia and 51 New Zealand) in e-commerce, digital experience or online sales strategy. The data evaluates consumer readiness for AI-mediated buying decisions alongside business capabilities required to support agentic commerce.

How common is AI usage among online shoppers?

81% of ANZ consumers have used AI tools to support their online shopping, primarily to offload cognitive tasks like complex product comparisons.

Are consumers comfortable letting AI recommend or select products?

41% of ANZ consumers are comfortable allowing AI to select products for them. As recommendation engines continue to improve, shoppers are increasingly delegating product discovery and curation to AI tools.

Will consumers allow AI agents to make purchases on their behalf?

Only 22% of ANZ consumers are comfortable letting an AI agent complete a purchase on their behalf. Consumer hesitation is primarily driven by security concerns, including fraud risk (58%), data privacy (46%) and loss of spending control (44%).

How much e-commerce traffic is generated by AI tools?

98% of ANZ businesses report receiving AI-referred traffic to their websites.

Does AI-referred traffic convert better than traditional web traffic?

76% of ANZ businesses report that AI-referred traffic converts at a higher rate than their site average. Because AI tools pre-qualify user intent before driving a click, these shoppers enter the sales funnel with higher purchase intent.

How much revenue will agentic commerce generate for businesses?

62% of ANZ businesses expect agentic commerce to account for more than 10% of their revenue within two years.

How many companies currently have live agentic commerce integrations in place?

Only 11% of ANZ businesses are currently operating live integrations, despite 63% of leaders saying they believe their organization is ready for agentic commerce. This reveals a massive competitive gap between early adopters already capturing AI-driven revenue and the majority of companies still in the planning stage.

Joint research by Stripe and Thoughtworks

Get the report

August 2026

Agentic commerce report: From confidence to execution

Consumer attitudes and business readiness across Australia and New Zealand

Mockup image of the Stripe and Thoughtworks agentic commerce report on an tablet

Ready or not, the conditions for agentic commerce are forming faster than most realize.

More consumers are using AI to find and buy products, and businesses are seeing a growth in AI-referred traffic that’s converting at premium rates. Despite this momentum, few have deployed agentic commerce integrations – most are stuck, believing they’re ready but yet to prove it in practice.

This joint study by Stripe and Thoughtworks brings these dynamics into focus – revealing how consumers are embracing AI-mediated shopping, where businesses are falling behind and what it takes to close the gap.

Get the report

Executive summary

The way consumers shop is beginning to change. AI platforms like ChatGPT, Gemini and Copilot are becoming primary commerce channels for product discovery, evaluation and selection. This joint study by Stripe and Thoughtworks across Australia and New Zealand reveals a clear pattern: consumer adoption is rising, AI-referred traffic is growing and it’s converting at premium rates. Yet as agentic commerce takes shape, a crucial gap has emerged between where consumers are and where business readiness really stands.

Key highlights

Consumer adoption becomes mainstream: 81% of ANZ consumers already use AI tools to support their online shopping journeys, mostly to outsource the heavy thinking when comparing products.
Trust falls at payment but single-use tokens fix it: Consumer trust plummets from 52% at research and 41% at selection to just 22% at checkout due to privacy and security concerns. However, issuing single-use tokens that mask card details restores confidence, increasing willingness to allow AI-mediated transactions by 30 percentage points overall.
Business confidence is high but execution remains low. While 63% of leaders believe their organizations are ready for agentic commerce, only 11% are currently scaling live integrations. Unprepared businesses don’t just risk being invisible to AI engines; they risk misrepresentation, competing blind and being locked out of the recommendation flywheel.
The barriers are plumbing: The primary hurdles to agentic readiness are foundational infrastructure issues, including fragmented data (22%), legacy systems (17%) and missing API layers (10%). Only a minority of businesses currently maintain structured, machine-readable catalogs (35%) or API-first product feeds (37%).
Sustainable advantage requires trusted, agentic-ready foundations. To bridge the execution gap, organizations need to move beyond basic product feeds toward real-time, highly legible and tightly governed data architectures that are visible to AI agents and instill trust in consumers.

“Consumer adoption is further advanced than most businesses recognize, and the commercial signal is already visible. Yet only 11% of businesses are actively scaling live integrations. The challenge is no longer simply making data available. It is turning that data into connected, governed capabilities that work reliably across live customer journeys.”

— Karl Durrance (Country Lead, ANZ, Stripe) and Kathryn Heffernan (Market Director, SEA & ANZ, Thoughtworks)

Key findings

Based on YouGov research surveying over 1,500 consumers and 200 businesses across Australia and New Zealand.

Consumer findings

Consumers are increasingly using AI when purchasing, but they want to remain in control.

81%

of consumers have used AI tools to support online shopping.

41%

of consumers are comfortable allowing AI to select products.

22%

of shoppers would let an AI agent buy on their behalf.

58%

of consumers cite security and fraud risks as their leading adoption concern.

Business findings

Revenue expectations for agentic commerce are high, but most businesses lack the foundations to support it.

98%

of businesses report seeing AI-referred traffic.

76%

of businesses say AI-referred traffic converts at a higher rate than their overall average.

63%

of businesses expect agentic commerce to account for >10% of revenue within two years.

11%

of businesses are scaling live integrations.

Trust collapses at payment but tokens restore it

Consumer comfort drops sharply as AI moves down the shopping funnel – from 52% for research, 41% for product selection to just 22% at checkout. Consumer hesitation is primarily driven by security concerns, specifically fraud risk (58%), data privacy (46%) and loss of spending control (44%).

However, this trust cliff is solvable. When offered a single-use payment token that masks real card details, consumer comfort increases by 30 percentage points overall. The good news is that the technology is already here – shared payment tokens (SPTs) initiate payments using a buyer’s saved payment method without exposing payment credentials. Stripe’s payment infrastructure already supports this technology. For businesses, integration requires no new checkout architecture – only a token-compatible payment flow.

Is your business ready to be found, chosen and transacted with?

Agentic commerce is already changing how customers discover, evaluate and buy. The opportunity now is to identify where it can create real commercial value, understand what stands in the way and move deliberately from readiness to execution.

1. Start with one high-value commercial flow

Identify a customer journey where agentic commerce could create measurable value, such as product discovery, complex comparison, dynamic bundling or assisted purchase. Focus on one revenue-critical flow where you can prove impact before expanding further.

2. Assess whether your business is agent-ready

Having data available is not the same as making it machine-readable. Ensure product feeds, catalogs and APIs are structured, attribute-rich and accessible to AI agents, so your products can be accurately discovered and represented.

3. Enable trusted, secure transactions

As agentic commerce moves from research to action, trust becomes critical. Establish explicit authorization checkpoints and security guardrails that keep agentic commerce interactions safe and ensure customers remain in control.

4. Measure what works, then scale

Evaluate the evidence to determine which customer journeys warrant deeper investment. Ultimately, organizations that learn fast and prioritize infrastructure investment in the journeys that prove value will build compounding competitive advantage.

FAQs

The joint report by Stripe and Thoughtworks explores how agentic commerce is advancing in Australia and New Zealand. It reveals that while consumer adoption and AI-referred traffic are rising, a major execution gap remains – with only a small fraction of organizations operating live integrations. The report offers business leaders a practical pathway to bridge this gap by establishing AI-ready data architectures and a dual strategy that maintains visibility to AI engines without losing direct customer relationships.
The report is based on research conducted by YouGov in July 2026, surveying 1,505 consumers (1,002 Australia and 503 New Zealand) and 206 senior decision-makers (155 Australia and 51 New Zealand) in e-commerce, digital experience or online sales strategy. The data evaluates consumer readiness for AI-mediated buying decisions alongside business capabilities required to support agentic commerce.
81% of ANZ consumers have used AI tools to support their online shopping, primarily to offload cognitive tasks like complex product comparisons.
41% of ANZ consumers are comfortable allowing AI to select products for them. As recommendation engines continue to improve, shoppers are increasingly delegating product discovery and curation to AI tools.
Only 22% of ANZ consumers are comfortable letting an AI agent complete a purchase on their behalf. Consumer hesitation is primarily driven by security concerns, including fraud risk (58%), data privacy (46%) and loss of spending control (44%).
98% of ANZ businesses report receiving AI-referred traffic to their websites.
76% of ANZ businesses report that AI-referred traffic converts at a higher rate than their site average. Because AI tools pre-qualify user intent before driving a click, these shoppers enter the sales funnel with higher purchase intent.
62% of ANZ businesses expect agentic commerce to account for more than 10% of their revenue within two years.
Only 11% of ANZ businesses are currently operating live integrations, despite 63% of leaders saying they believe their organization is ready for agentic commerce. This reveals a massive competitive gap between early adopters already capturing AI-driven revenue and the majority of companies still in the planning stage.

Financial infrastructure platform for businesses. Stripe’s Agentic Commerce Suite connects businesses to AI shopping agents across ChatGPT, Gemini, Copilot and more.

Thoughtworks is a global technology consultancy that integrates design, engineering and AI to help organizations respond to shifts such as agentic commerce with practical, scalable innovation. For more than 30 years, we’ve delivered extraordinary impact together with our clients by helping them solve complex business problems with technology and culture as the differentiator.