UK retail powerhouse The Very Group is reimagining the customer journey in the age of AI. Muktar Mahama, Head of Digital Product, shares how the digital-first retailer is building an agentic commerce strategy — powered by composable tech and grounded in customer obsession.
At The Very Group, we’ve always stood at the intersection of great products, flexible payments and digital innovation. Our much-loved retail brands, Very and Littlewoods, offer customers everything from fashion and beauty to home and electricals — all under one digital roof.
What really sets us apart is how we make those experiences flexible and accessible. The combination of amazing brands, rich product ranges and FCA-regulated flexible payment options, through Very Pay, helps us serve millions of families across the UK and Ireland.
As a business without physical stores, our digital experience is our shopfront. That means every click, every scroll and every interaction has to feel effortless and personal. And while we’ve made great strides in this direction, we, like every other retailer, have to contend with a new player in the game: AI.
Whether we like it or not, AI is here to stay — and it’s already changing our lives. It’s also fundamentally changing how customers search, browse and buy products. Even customers themselves are changing as AI agents enter the fold. This means the familiar conversion funnel the retail industry has worked around and towards (awareness → discovery → consideration → purchase) is collapsing.
Recent studies show that click-through rates from search results are dropping rapidly, and many more queries now end without a click, which means fewer opportunities for customers to visit eCommerce sites. Online click-through rates have dropped considerably this year across the industry, resulting in fewer visitors to eCommerce sites, which in turn impacts conversion rates.
Social commerce started taking the customer away from online brands: You can see a jacket on TikTok, buy it in-app and never visit a retailer’s website. And now, AI platforms like ChatGPT, Perplexity and Google’s new search experiences are accelerating that shift. People can ask, “What’s the best winter coat under £100?” and the AI channel can surface product results directly, sometimes even facilitating the purchase.
Shoppers don’t have to land on the brand’s eCommerce site. Ever.
So, as retailers, we have to ask:
What happens when customers no longer start or end their journey on our website?
How do we remain visible AND relevant when discovery happens in someone else’s ecosystem?
For me, AI isn’t something to fear. I genuinely believe AI can be a force for good if we use it responsibly. It can drive enormous value for both our customers and our colleagues. However, to achieve this, we need to rethink how discovery, personalization and data integration come together.
That’s the journey we’ve been on at The Very Group.
How it started and how it’s going: Our AI journey so far
We began our AI journey with two perspectives in mind: Our customers and our employees.
As a proudly customer-centric company, we asked ourselves: How can we make shopping easier, more personal, and more inspiring for our customers?
And just as importantly: How can we make our work more efficient — and even more inspiring — across our organisation?
This mindset ensures our AI initiatives stay grounded in real business value, rather than adopting technology simply because it’s new or “cool.”
As the use of generative AI for content creation started to gain momentum, we immediately recognised the opportunity to enhance the content on our product detail pages (PDPs).
Being the place where customers make key decisions on whether to buy the product, providing accurate, complete and rich descriptions is absolutely essential. Before AI, our copywriters had to manually update each and every detail. As a multi-brand retailer with millions of SKUs, this process was slow and prone to errors.
Using generative AI has been a game-changer. By generating richer and more accurate descriptions, as well as feeding models with product attributes and images, AI has significantly accelerated the process. Of course, our human copywriters — now more like editors — continue to play a fundamental role in reviewing and refining content to ensure each description fits our brand standards, as well as avoiding AI hallucinations.
The result is measurable: We've improved our product descriptions, reaching 90% of all PDPs for our Fashion and Sports categories, which has driven clear improvements in conversion and even reduced returns.
Our experience proves that AI doesn’t replace creativity; it amplifies it. Our copywriters now focus on quality and storytelling, rather than repetitive admin tasks.
Our focus on leveraging AI hasn’t stopped on the Product Pages; we’re also focusing on product discovery. Our goal is clear: To help customers find the right product quickly across our extensive product catalog.
We partnered with Constructor, an AI-powered discovery platform, as part of our broader migration to a composable architecture (more on that later).
This system learns from each customer’s clickstream data — their on-site searches, navigation and browsing behavior — and then tailors the results we display. So, if you and I both search “black shirt,” the products we see may differ based on what we’ve each been exploring. It’s subtle, but powerful.
We compared this AI-driven discovery experience with our previous solution and observed clear improvements in both customer satisfaction and commercial metrics. It’s proof that AI-powered personalization — done well and transparently — benefits everyone.
We’ve also been exploring how AI can help our colleagues behind the scenes. With our new Amplience content management system — another composable partner — we’re running an innovation programme focused on AI use cases for content.
Things like automatically generating alt text for accessibility, creating buying guides to improve SEO or drafting promotional copy. These are small efficiencies that add up to a significant cultural shift, allowing our teams to spend less time on manual tasks and more time on creativity and strategy.
Looking ahead, I’m excited about the immense possibilities of all things AI — agentic commerce included.
AI-driven shopping assistants are another pillar of our strategy, as they can leverage personal preferences to recommend products across brands in a conversational manner. Think of it as your own virtual stylist or personal assistant.
To support discovery in an evolving digital landscape, we’re focused on improving the quality and structure of our product content. This helps ensure it’s represented accurately across AI platforms like ChatGPT and Perplexity, while also enhancing the customer experience on our own websites and apps.
The foundation for innovation: Why composable architecture matters
None of the above would have been possible if we hadn’t fundamentally changed our technology stack.
For many years, we were operating on a legacy on-prem commerce platform, Oracle ATG. It had served us well, but over time, it became complex, rigid and difficult to manage. Our ambitions had outgrown its capabilities.
We needed a flexible, scalable and enterprise-level commerce platform — a foundation to stay ahead with both customer expectations and emerging technologies like AI. That’s why we decided to migrate to a composable architecture, which we’re calling Skyscape.
I like to think of composable architecture as Lego for eCommerce. Each block — commerce, content management system, search, IDAM and more can be swapped, upgraded or improved without disrupting the entire system. We’re free to choose best-of-breed partners, each one focused on building a brilliant product. When they innovate, we get that innovation instantly, and our customers and colleagues feel the benefit.
Our composable ecosystem now includes:
commercetools for commerce and checkout.
Constructor for product discovery and personalization.
Amplience for content management.
AWS Cognito for Identity & Access Management (IDAM).
Finally, we’ve built our own presentation layer on top because that’s where our differentiation lies, and we want uniqueness where it matters most. For us, that’s the way we bring retail and flexible ways to pay together to feel uniquely “Very.”
Our decision to move to a composable architecture gives us the kind of agility we need to innovate faster, test ideas quickly and respond instantly to what our customers want.
The same rationale applies to AI initiatives. We can experiment without worrying about disrupting the core infrastructure and roll out AI-based solutions faster. Say we want to try out an AI-enabled skin analysis tool for our beauty experience: We don’t have to wait months. We can do it far quicker than we’ve ever been able to. It becomes a business decision instead of a technology hindrance.
And that’s the whole point: Technology should never distract from customer obsession. Having a composable foundation enables us to spend less time maintaining systems and more time designing great experiences — for humans and AI agents alike.
5 lessons learned and recommendations for retailers
If I could share a few lessons from our journey so far, it would be these:
AI and platform migrations are as much about people and culture as they are about technology. From day one, we involved our colleagues in identifying problems, evaluating solutions and reviewing outputs.
When we introduced AI-generated product descriptions, for example, our copywriters weren’t sidelined — they were empowered. They became editors, curators and tone-of-voice guardians. That buy-in has been crucial. If people feel a sense of ownership, they become your biggest champions.
Transformation is disruptive — and that’s okay. But you need to prepare your teams for it. Make sure everyone understands the “why,” not just the “what.”
For us, that meant showing how composable architecture and AI aren’t abstract tech decisions — they’re enablers of better customer experiences and simpler workflows for our colleagues.
Get to know what your colleagues do today and how their jobs may change with the introduction of new technologies. Then, ensure you have the right training to support their transition to a new way of working, if necessary.
We made a conscious decision not to build everything ourselves. There are brilliant companies — like commercetools, Constructor, Amplience, AWS and others — whose sole focus is to make best-in-class solutions for their domains.
Partnering with them allows us to focus on our differentiator: The customer experience.
Conducting a thorough RFP with clear problem statements and use cases should help you select the right partner.
You don’t need to have a five-year AI roadmap on day one. Start with a clear problem. Test something small. Measure the results.
At The Very Group, we’re building our AI adoption on a test-and-learn mindset — fail fast, adapt and move forward. That agility has been key to our progress so far.
It’s easy for technology to become the headline. But ultimately, for us, this is all about helping customers to get more out of life — how we make their lives easier, their journeys simpler and their experiences more personal.
In an AI-driven world where customers might discover, compare and buy products without ever visiting your site, relevance becomes everything. That relevance will come from great data, flexible tech behind the scenes and an unwavering focus on what your customers actually need.
Looking ahead
We’re only at the beginning of what AI can do for the retail industry. From personalized discovery to agentic shopping assistants, virtual try-ons and hyper-personalized offers, the potential is huge.
But success won’t come from chasing every shiny new capability. It will come from building the right foundation, empowering your teams and staying relentlessly focused on the customer.
At The Very Group, that’s exactly what we’re doing:
Our composable foundation provides us with the flexibility to integrate new tools quickly and adapt rapidly to new market dynamics or emerging customer needs.
Our AI strategy helps us stay relevant as the funnel evolves.
And our people — across product, design and UX — are the heartbeat that keeps it all customer-first.
I genuinely believe AI will be a force for good for customers, colleagues and commerce as a whole. The retailers who thrive will be the ones who see it not as a replacement, but as an enabler of better experiences.
Interested in learning more about customers who have achieved success by migrating to commercetools? Visit our customer stories page.
Muktar Mahama leads the vision and strategy for the Digital Customer Experience team across web and mobile platforms for brands like Very and Littlewoods. With a background in software engineering, he focuses on driving digital transformation and utilizing data-informed decision-making to enhance the customer journey.