AI is transforming the way customers shop, how retailers work and where profit flows. We explore how retailers can accelerate their response in 2026 and catch up with consumers, manufacturers and eTailers who are ahead on AI adoption. The next five years will reward retailers that rebuild their customer value propositions, economics, capabilities and tech stacks around AI, versus those that treat it as just another tool to plug into today’s model.
Stay ahead with BCG insights on the consumer products industry
AI’s Impact on the Retail Business Model
AI is reshaping four core dimensions of retail, from how customers discover products to where and how value is created.
Customer Journeys: Missions, Not Products
Instead of browsing for individual products, customers are increasingly looking to solve a ‘mission’ when they shop like ‘refresh my winter wardrobe’ or ‘host a toddler’s birthday party’. Rather than navigating store aisles or clicking through categories on a webpage or via a search bar, customers are more likely to discover products that meet their needs on interfaces that best grasp their context, constraints and preferences.
Channels: Digital Informs, Stores Confirm
Digital channels, particularly AI assistants, will become the default space for research on considered purchases, with most customers deciding on a purchase or shortlist before they reach a store. Within a store, customers will seek out confidence, service and fulfilment. The role of the store associate will evolve and AI-driven labour scheduling, task automation and coaching will be critical to free up their capacity to consult with customers, make sales, solve customer problems and build loyalty.
Profit Pools: Fragmented and Asymmetric
Destination retailers – those that customers seek out directly – will retain healthier margins and drive incremental revenue by monetising owned data for loyalty, personalisation and media. On the other hand, evaluation retailers – those that rely on traffic from AI platforms – will face margin pressure as they compete on cost, fulfilment speed and agent visibility, with margin increasingly skewed toward a small set of scale and specialist winners.
Differentiation: Advantage Moves Up the Stack
As traditional sources of advantage like promotions, replenishment and forecasting become standardised by algorithms, they will no longer be meaningful sources of edge. Competitive retailers will instead have distinctive customer value propositions (CVPs), smart rules that shape algorithms and human critical thinking to spot friction, identify whitespace and build new strategic narratives.
Responses to These Shifts
In the face of these changing customer behaviours, widespread adoption of algorithms and new channels, retailers will need to make dramatic changes over the next few years to evolve their operating models, capabilities and investment priorities. To keep up with customers, manufacturers and eTailers, they will need to fundamentally reshape how they work, with AI at the heart – not just as a tool to plug in.
Operating Model: Human-AI Teaming as the Default
To embed AI at scale, retailers need to transform all components of the operating model (Exhibit 2).
Strategy and Governance (A and B)
First, retailers must define their endgame as AI platforms upend the customer journey: do they want to be a destination player (where customers shop directly), an evaluation player (that wins through recommendations from AI platforms) or somewhere in between? This will guide their AI roadmap, priorities and where they invest for advantage. Destination players will need to invest in assisted discovery, data-driven personalisation and loyalty to own missions, while evaluation players focus on cost leadership, fulfilment speed and agent visibility.
This choice must be owned at the top. Executive teams need to set clear value priorities, commit to AI governance and oversee enterprise-scale adoption. That means establishing responsible AI guardrails and human-AI decision rights, and building investment frameworks that track value, not just activity.
Operating Model Building Blocks (C, D, E, F and G)
Moving down the pyramid, the next row of building blocks (capabilities, workflows, AI applications, organisational structure and KPIs) need to be re-shaped for blended human-AI teams, rather than AI tools layered into existing roles. Workflows must be reimagined from first principles with every step challenged: maximise agentic efficiency, question where humans need to be involved and redraw functional boundaries to reduce friction and increase speed.
These workflows will become the blueprint for custom AI applications, agentic processes, new structures with optimised managerial spans and layers and refreshed KPIs that unlock real productivity gains. Given the pace of AI’s evolution, retailers should look to off-the-shelf applications that offer scale and continuous upgrades and reserve in-house development for applications that differentiate their CVP, economics or channel experience (e.g. pricing).
In parallel, retailers must rapidly lift the AI fluency of their organisation. Teams must be trained and encouraged to use tools like ChatGPT Enterprise and CoPilot responsibly, avoiding the 20%+ productivity drag seen in early AI rollouts where adoption was unmanaged.
Enablers (H and I)
AI outputs are only as strong as the data they ingest. The success of AI-powered workflows depends on real-time, reliable data flows, robust platform foundations and resilient technical infrastructure. Yet most retailers are held back by fragmented, low-quality data due to years of underinvestment. Lifting data availability, quality and connectivity should be prioritised based on the ROI, with pragmatic workarounds where needed.
As AI increasingly touches customer data and decisions, cyber security becomes non-negotiable. Retailers must protect the integrity of their AI models, training data, interaction interfaces and agentic tools against threats like prompt injection, data poisoning and others. This is critical for data safety and to earn customer and employee trust.
Sustained Change (J)
Advanced models and algorithms are no longer the bottlenecks to scaling AI. Lasting impact and a truly AI-enabled operating model require the largest workforce shift since the introduction of the personal computer, multiyear investment in re-skilling, real-time data, human-AI workflows, platform modernisation and application development.
The biggest value comes from widespread adoption which starts with people. Returns extend far beyond cost and productivity with impacts on accelerated decision-making, innovation and growth. Leaders must therefore treat AI adoption as an ongoing journey of listening, co-creation and adaptation: communicating a shared purpose, involving employees in shaping that purpose and helping them embrace
Organisational Capabilities: Leaner, Smarter and More Strategic Organisations
In AI-mature organisations,
Three strategic capability stacks – merchandising, customer growth and digital activation, and technology – will account for around 70% of above-store employee costs for destination retailers (Exhibit 3). We outline their roles and capabilities below.
- Merchandising (27%) teams will curate mission-ready assortments, plan inventory flows and design pricing rules that win AI agent trust and convert customers. Category managers will become ‘mini CEOs’, focused on driving greater category performance, with the time-consuming parts of their job – including vendor negotiation fact-pack generation and negotiation, competitive monitoring and price recommendations, and demand forecasting – heavily supported by AI.
- Customer growth and digital activation (13%) teams will defend the direct relationship between customer and retailer, preventing disintermediation by orchestrating loyalty, personalisation, CX, digital channels and retail media to monetise traffic, create repeat missions and build proprietary data signals. This will require upskilling teams in customer journey engineering (designing end-to-end interactions, feedback loops, human escalation protocols), mission design (designing loyalty nudges and habitual journeys) and predictive personalisation (generating upsell/cross-sell signals during a mission).
- Technology (27%) will shift from being a back-office cost centre to a role of enterprise-wide enablement, embedding intelligence across the organisation with real-time data engineering, agent orchestration, algorithm governance and AI literacy uplift. Retailers will need to develop capabilities in machine-learning engineering, agent orchestration, machine learning ops, responsible AI oversight and workforce enablement. And existing roles such as product managers, data engineers, software developers, architects and cyber security specialists will need upskilling in AI platforms, patterns and controls to safely embed AI across the operating model.
Investment Plans: Sustained and Strategic
Given the scale of operating-model change required, we expect total annual investment levels to rise by around one third over the coming years, driven by a combination of capex uplift and digital and IT operating spend.
Within the capex envelope, the underlying composition will also shift. More capex will flow to supply chain automation, AI-enabled platforms, data infrastructure and digitally-enabled customer experience. And less will flow to store fit outs and hardware as retailers look to lighter maintenance and selective upgrades focused on safety, loss prevention and targeted task-level automation (Exhibit 4).
In parallel, digital, IT and AI operating expenditure is set to increase by at least one third as retailers fund subscription AI platforms, cloud computing, model operations, cybersecurity and enterprise-wide reskilling.
Financial returns will play out over a 3–5-year horizon as workflows and systems modernise and adoption grows. Retailers will need to manage shareholder expectations for longer paybacks, while staying disciplined on value tracking, use case prioritisation and delivery.
The cost of delay is real. Retailers that hesitate on this investment risk falling further behind eTailers and brands already scaling AI adoption at pace.
Actions for 2026
We see five imperatives for retailers to accelerate their responses to the changing retail business model in 2026:
- Define your strategic AI endgame: Destination or evaluation? Articulate your AI endgame and CVP to win customer share-of-wallet as AI-assisted channels reshape discovery, engagement and conversion.
- Set an enterprise AI roadmap: Identify the use cases that justify end-to-end workflow redesign and can materially lift productivity, speed or customer outcomes. Define the investment envelope you require in year one (across data, platforms, training and tools), your strategy to build or buy solutions and your 3–5-year ROI ambition and align shareholder expectations early.
- Lift your workforce’s AI fluency and adoption: Equip your teams to use existing enterprise AI tools responsibly and effectively. Drive adoption with hands-on training, change leadership and clear role‑based use cases that make AI integral to daily work.
- Build the foundation for enterprise-wide AI: Set up AI guardrails and shared AI infrastructure to support the safe deployment of early AI use cases across your entire organisation.
- Roll out select big bets: Scale 2–3 flagship AI use cases (e.g. AI-driven pricing, markdowns, HR support or talent acquisition) across your organisation to demonstrate early value and fully transform 1–2 core functions with AI-enabled workflows, new roles, tools and performance metrics to sustain momentum.
AI is fundamentally changing retail as we know it. To respond takes more than applying AI tools to legacy retail models; the real winners are redesigning their CVPs, economics, capabilities and operating models end to end.
This is the year to define your AI endgame, commit to an investment plan and kick off end-to-end operating model redesign. According to BCG’s AI Radar (a global survey of 2,400 business executives), four out of five CEOs are more optimistic about the ROI of their AI investments than they were a year ago and nearly all CEOs believe that AI agents will produce measurable returns in 2026.
Get in touch to make it happen and learn from AI pioneers pulling ahead.
