McKinsey: GenAI could add $150-275bn to fashion operating profits in 3-5 years

McKinsey estimates that generative AI could add between $150 billion and $275 billion to the apparel, fashion and luxury sectors' operating profits within the next three to five years, according to the consultancy's 2023 analysis. The range reflects a conservative to optimistic scenario, and the prize is large enough to justify the sector's growing investment appetite. More than 35 per cent of fashion executives already use generative AI in daily operations, according to the State of Fashion 2026 report published by McKinsey and the Business of Fashion in December 2025.

Yet adoption maturity lags far behind intent. While 92 per cent of fashion organisations say they will increase their investments in generative AI, only 1 per cent describe their deployment as mature, according to the same survey of executives conducted for the State of Fashion 2026. Many remain stuck in pilot mode, testing siloed AI solutions with limited impact, with scaling often seen as too complex or costly.

The contrast between ambition and execution matters because the operating environment for fashion is deteriorating. Forty-six per cent of fashion executives expect industry conditions to worsen in 2026, up from 39 per cent in the previous year's survey. Tariffs, declining consumer confidence and inventory pressure are forcing brands to seek new efficiency levers. In that context, AI is shifting from a competitive edge to a business necessity, McKinsey argues.

Where will the $150 billion to $275 billion in value come from?

McKinsey's estimate spans the full value chain, from design and merchandising to marketing, supply chain optimisation and store operations. The 2023 McKinsey report, Generative AI: Unlocking the Future of Fashion, outlines use cases across six major areas: merchandising and product, marketing, digital commerce and customer experience, supply chain and logistics, store operations, and organisation and support functions.

In merchandising and product, generative AI can convert sketches, mood boards and descriptions into high-fidelity designs, enrich product ideation by collaborating with AI agents that generate creative options from past product lines and inspirational imagery, and customise products for individual consumers at scale. A Hong Kong-based design group from the Laboratory for Artificial Intelligence in Design held a fashion show in December 2022 featuring generative AI-supported designs, an early signal of what the technology could enable.

In marketing, around 22 per cent of tasks can be automated by 2030 with generative AI, particularly in strategy development, campaign ideation and content creation, according to the State of Fashion 2026 report. Communications and marketing functions could see the greatest increases in time savings from automation with generative AI in the United States fashion industry. McKinsey estimates that AI can increase marketing productivity by 10 to 20 per cent through better targeting and optimisation, and that personalisation can increase revenue by 5 to 15 per cent.

In supply chain and logistics, AI-driven demand forecasting can reduce forecast errors by 20 to 50 per cent and cut lost sales by up to 65 per cent, while inventory optimisation can lower inventories by 20 to 50 per cent, according to McKinsey research. Digital ordering systems and AI tools can improve stock visibility and optimise in-store labour allocation.

The McKinsey estimate is sector-wide. At the individual company level, returns depend on deployment quality, data maturity and the ability to integrate AI into workflows rather than leaving it in experimental silos.

What adoption scenarios determine who captures the value?

Morgan Stanley tracked AI adoption among consumer durables and apparel companies and found that the share classified as AI adopters more than doubled from 20 per cent to 44 per cent during the first half of 2025. The bank's framework suggested that fashion brands and retailers could see $6 billion in total cost savings opportunity and potential at the midpoint of estimates, translating into a 20 per cent increase for earnings before interest and taxes in 2026.

Companies that have scaled AI across at least two functions report 8 to 12 per cent higher earnings before interest and tax margins compared with peers that have not adopted AI, according to McKinsey's State of Fashion 2025 report. The gap between AI adopters and non-adopters is widening, driven by superior pricing power and efficiency gains among the former, Morgan Stanley noted.

Adoption is accelerating fastest in customer-facing applications. More than 35 per cent of executives report already using generative AI in areas such as online customer service, image creation, copywriting, consumer search or product discovery. Customer product discovery and search was the top-ranked use case for generative AI in 2025, according to fashion executives surveyed for the State of Fashion 2025.

The State of Fashion 2026 report notes that companies across industries are expected to unlock productivity gains of more than 30 per cent over the next five years due to automation and generative AI. By 2030, around one third of employee time across industries could be automated by generative AI and other technologies in Europe and the United States.

How much must fashion companies invest to capture GenAI value?

Investment requirements vary by company size and ambition. For small and independent brands, annual AI tool costs typically range from $350 to $2,400, yielding savings of $5,000 to $50,000 or more in the first season through reduced sampling, photography and iteration costs. Mid-sized operations report average savings of $2 million to $5 million per season from generative AI that reduces physical sampling costs by up to 90 per cent, according to BCG analysis cited in fashion industry research.

Large enterprises face higher upfront costs. Typical investment for enterprise-scale AI deployment ranges from $2 million to $10 million or more, covering platform licenses, data infrastructure, integration work and talent acquisition. However, first-year returns can be substantial when AI is deployed across multiple functions. One analysis estimated that a large fashion retailer could achieve annual savings of $2.16 million in return processing costs alone by reducing returns through AI-powered virtual try-on and fit tools.

The State of Fashion 2026 report emphasises that 92 per cent of organisations say they will increase investments in generative AI, but many are stuck in pilot mode, testing siloed AI solutions with limited impact. Scaling is the bottleneck. A 2025 MIT study found that only 5 per cent of AI pilots extract millions in value, with data and operating models, not the AI models themselves, identified as the primary constraint.

Which companies have deployed GenAI at scale, and with what results?

Zalando is among the most advanced adopters. The European fashion platform reported that generative AI reduced image production costs by 90 per cent, according to the State of Fashion 2026 report. In the fourth quarter of 2024, 70 per cent of its editorial content was AI-generated. The share of AI-generated product content increased from almost zero to 90 per cent within one year, while campaign creation timelines dropped from six weeks to a few days, according to company statements in May 2026.

Zalando expanded its AI assistant to include beauty recommendations alongside fashion advice in early 2026. Close to 10 million customers used the assistant in the first quarter of 2026 alone, compared with 6 million across the whole of 2025. The company reported group gross merchandise volume growth of 21.7 per cent year on year to €4.29 billion in the first quarter of 2026, while revenue rose 23.8 per cent to €3 billion. Adjusted earnings before interest and tax increased by 38.7 per cent year on year to €65 million, supported by €10 million in synergies from the ABOUT YOU acquisition before integration costs.

Ralph Lauren launched Ask Ralph, an AI-powered conversational shopping assistant, in September 2025. Built in partnership with Microsoft and powered by Azure OpenAI, the tool uses natural language processing to serve shoppable visual layouts of complete outfits based on conversational prompts. The company reported better than expected second quarter fiscal 2026 results in November 2025, with net revenues of $2.01 billion, up 16.5 per cent on a reported basis and 14 per cent in constant currency. Net income for the second quarter was $207.5 million, compared with $147.9 million in the prior year period.

H&M piloted digital twins in 2024 and 2025, creating AI models from 30 human models with explicit consent and rights management. The company uses explicit disclaimers on images of digital twins. Mango launched a teen campaign built entirely with AI in 2025 and now uses templated AI ad variants to scale creative.

Etro uses AI to enhance product descriptions, support multilingual translations and generate creative assets for product pages, as presented at Business of Fashion VOICES in December 2025. Shopify's Winter 2026 Edition, released in December 2025, introduced over 150 AI-powered features including SEO-optimised product description generation, built-in image editing and an AI catalogue system that automatically infers product categories across billions of listings.

What are the risks and limitations of GenAI in fashion?

Algorithmic bias represents one of the most widely discussed ethical concerns. AI models trained on image data can unintentionally reinforce biases concerning body type, skin colour and cultural norms. Research demonstrates that AI-driven fashion systems frequently reflect mainstream consumer preferences because training data is typically sourced from major brands, social media and leading retail markets. If datasets are incomplete, skewed or unauthorised, the result could be misleading trend predictions, incorrect sizing or culturally insensitive content.

Data privacy and security concerns are rising. AI fashion tools require personal data such as photos, body measurements, style preferences and behavioural patterns. In 2026, privacy-conscious consumers are increasingly selective about which tools they use. Tools that explain clearly what data they collect and how it is used see higher adoption rates. Some tools process wardrobe photos locally rather than sending them to cloud servers, a growing differentiator.

The provenance of training data behind the AI imagery models powering fashion brands' outputs remains an active question. Several major lawsuits in 2024 and 2026 addressed whether AI image generators trained on uncredited photographs constitute fair use or infringement. Brands using AI imagery vendors should ask about training data sources and prefer vendors with clean, documented training data origins.

Hallucinations and accuracy issues remain a technical risk. AI-generated content can produce factually incorrect information or nonsensical outputs. Disclosure and transparency are emerging regulatory requirements. By the end of 2027, fashion brands using AI-generated imagery in advertising may face mandatory labelling in multiple major markets. The European Union AI Act demands obligatory labelling of synthetic media starting from 2026. New York's Fashion Workers Act, effective June 2025, requires explicit consent for digital replicas of models.

Employment concerns also arise. Human models and creative workers may lose opportunities to automated systems, creating tension between efficiency and fair labour practices. McKinsey frames generative AI as a productivity multiplier across design, marketing and merchandising, with roughly 30 per cent of employee time automatable by 2030, while creative direction and taste remain human. In practice, designers use AI for ideation and rapid iteration, not for running collections on autopilot.

Returns on investment remain uncertain for many. Gartner predicted that at least 30 per cent of generative AI projects would be abandoned after proof of concept by the end of 2025. The MIT NANDA study found that only 5 per cent of AI pilots extract millions in value, with data and operating models, not the AI models themselves, identified as the primary bottleneck.

How does GenAI impact different functions across the fashion value chain?

In design and product development, McKinsey research highlights that a quarter of generative AI's value in fashion could stem from product styling and development. Twenty-eight per cent of industry players were already integrating AI into creative design processes as of 2023. AI helps brands visualise concepts faster while designers remain responsible for editing and finalising garments, according to a Wall Street Journal report from January 2026.

In merchandising and planning, AI is increasingly used to support demand forecasting, pricing decisions and assortment planning. Merchandising teams can respond more quickly to performance trends. AI-supported insights help guide allocation decisions and reduce excess or shortage risk. By 2030, around one third of employee time in fashion could be automated by generative AI and other technologies.

In marketing, organisations that are implementing agentic workflows can expect to see 10 to 30 per cent revenue growth from hyperpersonalised marketing, according to McKinsey research. Agentic systems are estimated to accelerate the creation and execution of marketing campaigns by 10 to 15 times by speeding up both the brainstorming and vetting of ideas. Marketing is moving from content production to curation, with around 22 per cent of marketing tasks automatable by 2030.

In operations and reverse logistics, returns are a major target. An estimated 19.3 per cent of United States online sales were returned in 2025, according to the National Retail Federation. Online apparel return rates run in the 24 to 30 per cent range. Retailers implementing virtual try-on averaged a 23 per cent reduction in apparel return rates, with some categories hitting 40 per cent, according to a 2025 Forrester analysis.

What counter-arguments challenge the $150 billion to $275 billion estimate?

The McKinsey estimate is a sector-wide projection over three to five years, not a guaranteed outcome for individual companies. It assumes that fashion companies will successfully scale AI beyond pilots, integrate it into core workflows and capture efficiency gains. The State of Fashion 2026 data shows that 92 per cent plan to increase investment but only 1 per cent have mature deployments, suggesting that execution risk is substantial.

The estimate also assumes that productivity gains translate into operating profit rather than being competed away through lower prices or invested into growth initiatives that do not immediately lift margins. In a deflationary environment where consumers are price-sensitive and value segments are expanding, brands may choose to pass AI-driven cost savings to customers rather than retain them as profit.

Some industry observers argue that the value may be front-loaded in areas like image generation and customer service, where early wins are visible, but harder to capture in areas like design and merchandising, where AI augments rather than replaces human judgement. The Wall Street Journal reported in January 2026 that AI helps brands visualise concepts faster while designers remain responsible for editing and finalising garments, suggesting that productivity gains in creative functions may be more modest than in operational functions.

There is also debate about the size of the addressable market. McKinsey's estimate applies to apparel, fashion and luxury operating profits. The global fashion market was valued at approximately $997 billion in 2026, according to McKinsey and Business of Fashion. If the industry posts low single-digit growth as projected, and if operating margins remain under pressure from tariffs and input costs, the absolute pool of operating profit available to lift may be smaller than the $150 billion to $275 billion range suggests.

Finally, the estimate does not account for the costs of AI deployment, including platform licenses, data infrastructure, integration work, talent acquisition and the write-off of failed pilots. If 30 per cent of generative AI projects are abandoned after proof of concept, as Gartner predicted, the net value captured by the sector will be lower than the gross opportunity.

Frequently asked questions

What should fashion companies do now?

1. Audit current AI pilots and kill the ones that do not scale. If a pilot has been running for more than six months without a clear path to production deployment, either resource it properly or shut it down. The 1 per cent maturity figure signals that most companies are accumulating pilots, not scaling solutions.
2. Prioritise use cases with measurable ROI and short payback periods. Start with image generation, product descriptions and customer service, where vendors offer proven tools and results are visible within weeks. Avoid complex custom builds in areas like trend forecasting until the operational foundations are in place.
3. Invest in data quality and integration before models. The MIT NANDA study found that data and operating models, not AI models, are the bottleneck. Clean product data, integrated inventory systems and unified customer profiles are prerequisites for scaling AI across the value chain.
4. Set clear governance for bias, transparency and consent. Establish guidelines for training data provenance, require vendors to document data sources and implement bias detection and diversity measures. Prepare for mandatory disclosure requirements by labelling AI-generated content now.
5. Reskill teams and recruit for new roles. Key hiring criteria include expertise in data quality, system integration, digital twins and marketing automation, according to the State of Fashion 2026. Upskill existing employees in AI-assisted workflows rather than assuming automation will simply displace headcount.
6. Measure productivity gains, not just cost cuts. Track time savings, campaign velocity, iteration speed and error reduction alongside cost metrics. The largest returns come from doing more with the same resources, not just doing the same with fewer resources.
7. Plan for agentic AI in 2027 and beyond. While current deployments focus on generative AI for content and recommendations, agentic AI that can make decisions and automate systematic processes is emerging. Brands should monitor developments in agentic commerce, where AI agents assist with search, product selection and purchases.