Everyone’s talking about AI. But for most retailers, the biggest challenge isn’t whether to adopt AI — it’s where to begin.
Do you focus on internal efficiencies, better customer experiences, or fixing pricing and promotions? And what kind of AI do you even need? LLMs? GenAI? Agentic AI?
Here’s a simple way to frame your starting point:
Ask: Where is the biggest opportunity for impact in your business today?
- Is the biggest immediate opportunity in optimizing your cost base and supply chain?
- Do you need to improve pricing, promotions, and your supplier collaboration?
- Are you struggling with loyalty, engagement, or growing basket size?
McKinsey estimates that GenAI could unlock $240–$390 billion in value for retail annually — adding 1.2 to 1.9 percentage points to margins.
Yet most companies are still stuck in pilot mode. Adoption is high, but scale is rare. The winners are those who choose a focused path and rewire workflows accordingly.
Understanding the AI Landscape in Retail
To set the foundation, here’s a quick breakdown of the types of AI we’ll refer to in this post:
- Specialised AI: Domain-specific models, often based on machine learning, used in areas like forecasting & optimisation engines. Think of this as the automatic breaking capability in your car.
- Generative AI (GenAI): AI that can generate new content — such as product descriptions, marketing copy or chatbot responses, — based on patterns in data. To continue the car reference, think of this as the part of your car that reads all the millions of data points your car is transmitting at any given time.
- Agentic AI: Systems that not only generate or analyse data, but autonomously plan and take actions toward a goal — combining GenAI, automation, and reasoning. Think of it as an AI-powered co-worker or planner. Think of this as the autonomous car, able to guide all the specialised capabilities to do the right thing at the right time.
Let’s discuss a few concrete examples of how to use AI today and where to start.
Optimising the Retail Value Chain with AI
Retailers win or lose on execution — and that starts with how you manage your supply chain.
AI can already drive tangible value across the end-to-end value chain. Here are five high-impact areas where AI is helping retailers today:
1. Accurate Forecasting
Domain-specific AI models can predict demand by SKU, location, and channel with far greater precision, helping reduce overstock and stockouts.
2. Replenishment Optimisation
Go beyond forecasts. AI can factor in constraints like truck capacity, store space, delivery frequency, spoilage risk, and inventory accuracy to optimise ordering at scale.
3. AI Agents for Root-Cause Analysis
Agentic AI can automatically detect and explain issues in availability, spoilage, or inventory accuracy — and even recommend actions. These systems operate continuously and learn over time.
4. Inventory Counting Automation
LLMs and computer vision are streamlining store inventory counts, reducing labour and improving accuracy.
5. In-Store and DC Process Optimisation
Specialised AI can guide store employees or warehouse teams on optimal picking and shelving sequences to reduce time and errors.
Many Retailers have already deployed either some, or many, of the above AI capabilities to help make smarter, and more informed operational decisions.
The takeaway: AI in the value chain is not futuristic — it’s operational. And it’s where many retailers are starting to see rapid ROI.
Smarter Pricing & Promotions with AI
Promotions can make or break your margins. Yet many retailers don’t know which ones are actually working.
1. Analyse Existing Promotions with AI
Use AI analytics to determine which promotions add value and which destroy margin. Retailers are seeing:
- 10%+ uplift in promo profitability
- 15%+ increase in promo revenue
Most retailers run extremely high promotional shares of revenue, and this is the area where huge quick impacts can be made. The number of data and the fact that suppliers keep spending most of the promotion suggestions makes it almost impossible for any human to manually review, analyse and understand the true impact of all these promotions.
AI can do this in seconds.
2. Plan Better Promotions with AI Agents
Once you know what works, AI Agents can help generate promotion plans aligned to your goals — fast, consistent, and data-backed. Think about this in a way of guiding the machine to do the investigation and clicking on your behalf. “I need x amount of revenue for next month for this category, please plan a campaign that is compliant with my strategy, is based on theme y and will be active for z period.
3. Dynamic Pricing Strategy
AI can help you:
- Monitor competitors
- Model elasticity and simulate outcomes
- Optimise for profit or volume within your pricing strategy
Another area where Agentic AI can add concrete value, is to continuously analyse internal and external data, recommend changes, and even automate execution — freeing your team to focus on strategy.
While GenAI creates the content, Agentic AI handles the action — analysing, adapting, and executing price strategies in real time.
The result: A shift from firefighting price issues to strategically guiding them — with AI as your partner.
Using AI to Drive Loyalty, Basket Size & Engagement
Loyalty is no longer just about points. It’s about creating intelligent, personalised experiences.
AI is powering new kinds of customer interaction:
1. Personalised Chat Assistants
LLMs and GenAI are behind intelligent chatbots that:
- Help customers discover new products
- Guide them through shopping journeys
- Answer service queries 24/7
According to McKinsey, 82% of retailers have experimented with GenAI in customer service — but only 36% have managed to scale.
Once again highlighting that the biggest challenge with adopting AI currently is not the lack of ideas and use cases, it’s focus on the right things and ensuring value is generated for specific challenges.
2. Tailored Shopping Lists & Recommendations
Agentic AI can understand customer intent and behaviour to generate:
- Personalised meal or product suggestions
- Auto-generated weekly shopping lists
3. In-Store Assistance
From voice-enabled product finders to GenAI-powered kiosks — the goal is to reduce friction and increase convenience.
Why it matters: These interactions not only increase basket size — they also deepen customer relationships.
Final Thoughts:
Retailers don’t need to embrace all of AI at once. Instead, they need a clear, focused path — aligned to their biggest opportunities.
AI is not hype. It’s already reshaping how the best retailers operate — and the window to catch up is now.
Retailers that focus on scaling just a few high-impact use cases — while rewiring workflows and investing in AI literacy — are pulling ahead.
References: https://www.mckinsey.com/industries/retail/our-insights/llm-to-roi-how-to-scale-gen-ai-in-retail
From AI to ROI 
Leave a comment