Customer Contact 3.0

Back-Office Automation

Shrink 

Insight Reports

 we spent three days at Web Summit in Lisbon — speaking with founders, operators and investors, and seeing hundreds of demos, workshops and main-stage sessions. Across it all, five themes kept resurfacing. Each theme points to a deeper shift in how retailers, brands and enterprise organisations will build, operate and compete over the next 18–36 months. 

Below are the ones we believe matter most: 

1. Agentic Commerce Moves from Theory to Deployment

Agentic commerce was everywhere this year — and not as a thought experiment. AI-driven traffic to ecommerce sites reportedly grew more than 4,700% year-on-year in July 2025, and retailers like Walmart and Etsy are already enabling transactions directly inside AI applications like ChatGPT and Perplexity. It’s a clear signal: a growing share of shopping journeys will soon begin with an agent, not a human. 

That shift has consequences. If the primary “customer” becomes an AI agent optimising for structured data — price, stock accuracy, delivery terms, returns clarity — then the traditional discovery journey compresses. The battleground becomes data quality, not merchandising; clarity, not persuasion. 

Dana Lawson’s session on Zero-UI engineering captured this well: retailers will need to design for AX (agent experience), not UX. That means cleaner schemas, machine-readable policies, real-time availability and well-structured APIs that let agents interpret and compare products confidently. 

A wave of early infrastructure is emerging. 

 lets brands create a dedicated AI-readable web layer. 

 provides a real-time MCP/API bridge so agents can fetch accurate availability, pricing and policy data — reducing the risk of hallucinated recommendations, out-of-date stock, or unsanctioned purchases. 

But major questions remain. Agents still recommend out-of-stock or unsuitable products, largely because most aren’t yet integrated into live commerce systems. And as agents transact independently, intent, liability and dispute resolution become non-trivial. Retailers will need strong recourse models before this scales. 

The Missing Layer in Enterprise AI: Context, Not Capability

Enterprise AI surfaced repeatedly, but the mood was clear: adoption is rising; ROI is not. Speakers cited MIT’s finding that roughly 95% of AI projects fail to scale, alongside McKinsey’s estimate that nearly 80% of companies using AI see no material bottom-line impact. 

The message was the same across the board: the bottleneck isn’t model capability — it’s context. 

Many enterprise deployments still sit on top of weak foundations: fragmented knowledge stores, shallow retrieval layers, inconsistent metadata and limited orchestration across systems. Without a shared, structured source of truth, even highly capable agents behave unpredictably, generating confident answers that collapse under scrutiny. In other words: enterprises don’t have a model problem, they have a memory and meaning problem. 

A growing set of startups is positioning against this gap. 

 each focus on strengthening the knowledge layer itself — building the contextual scaffolding that lets agents reason, retrieve and respond with reliability rather than surface-level fluency. 

Until that layer matures, experimentation will continue to outpace impact.  

Cybersecurity Becomes a Front-Line Commercial Risk

Cybersecurity felt unusually front-and-centre this year — especially for retailers, where the financial and reputational stakes are rising sharply. Recent breaches at Louis Vuitton, Dior and M&S illustrate how quickly incidents now escalate, with losses reaching tens or hundreds of millions. Cyber risk has now moved well beyond IT. It’s now a commercial issue. 

The newest wave of startups reflects that reframing. 

 takes a human-centred approach, using AI-powered simulation and personalised training across email, SMS, voice, chat and deepfakes — acknowledging that most breaches originate with people, not perimeter failures. Whereas 

 covers the technical heavy lifting — automated pen-testing, round-the-clock monitoring and dark-web intelligence — aimed at organisations without a full internal security function. 

A recurring theme: tools alone aren’t enough. Resilience — clear roles, practiced response plans and decision-making discipline — now matters as much as detection. In sectors built on trust, a breach doesn’t just compromise systems; it compromises brand credibility.  

4. The Creator Economy Becomes a Performance Channel

The creator economy had a visible presence at Web Summit — not just in startup pitches, but across the headline speakers. Talks from Grace Beverley and Josh Richards underscored how far the space has moved: influence is no longer the product; it is the distribution engine for IP, brands and communities. Creators aren’t just producing content anymore — they’re building brands. 

In parallel, we saw a new layer of tooling emerge to help brands run creator partnerships with more structure and less guesswork. 

 is targeting Shopify-powered brands with automated creator matching, briefing and outreach. 

 is pushing into enterprise-level campaign governance with workflow control, audience authenticity and analytics. 

 is taking a different angle, offering performance-based pricing where brands pay only for verified commercial outcomes. 

What we’re seeing is a shift toward creators becoming a structured, performance-driven commercial channel. Brands that treat this as an ongoing, measurable capability — supported by the right platforms — are the ones unlocking repeatable value rather than sporadic campaign wins. 

5. Robotics and Embodied AI Move Toward Practicality

Robotics and embodied AI were highly visible this year, from roaming floor demos to full humanoid models capable of shaking hands and holding conversations — including a striking blue-haired humanoid showcased by the Artificial Intelligence Association. But after the novelty wore off, the real question felt more practical than futuristic: what design, environment and ROI case will make robots useful, not just impressive? 

Two of the strongest examples came from companies focused on operational efficiency rather than anthropomorphic design. 

 demonstrated how hospital-grade cleaning and workflow automation can unlock real clinical value, while 

 showcased multi-purpose cleaning, delivery and building automation robotics for hospitality, logistics and smart infrastructure. Both prioritise safety, uptime, and integration over human-like interaction — and that felt far closer to deploy-ready reality. 

At this stage, the robots gaining real traction are the ones built for a specific job and a clear business case — not humanoids designed to feel like co-workers. The bigger open question remains: as hardware capability accelerates, where will commercial value justify the investment, and what form factor will end-users truly trust — tool-like, human-like, or something in between? 

Web Summit gave a clear view of where the industry sits today. Curiosity has met capability, and the focus is shifting toward what creates repeatable operational value. Clean data, strong architecture, governance and clear ROI are becoming the real differentiators. 

The takeaway wasn’t that everything is ready. It’s that the centre of gravity has moved. The next wave belongs to operators who can turn promising technology into measurable outcomes. 

In our latest blog, we break down five themes we saw at Web Summit 2025, from agentic commerce to enterprise AI, cybersecurity, creators and robotics, and what they signal about how retailers and brands will build and compete over the next 18–36 months. 

Web Summit 2025: Five Themes Shaping the Next Wave of Tech  

The GenAI Divide: Why AI Works for People but Not for Companies at Scale (Yet)

MIT made headlines everywhere with their claim that only 

“5% of generative AI pilots make it to production and deliver measurable enterprise value”

Now, even if you doubt the validity of that statistic as some have, one thing is undeniable the ‘informal’ use of personal AI tools in organisations, also known as shadow AI, is booming versus enterprise-led AI initiatives. 

Across every industry, employees are using tools like ChatGPT, Claude and Gemini to write emails, craft campaigns, summarise research or build slides. They’ve quietly reinvented how they work. 

If AI works so well for individuals, why is it still struggling to transform organisations? 

 the widening gap between personal productivity gains and enterprise-level impact. 

In other words, everyone’s experimenting with AI, but almost no one is scaling it. 

It’s easy to assume the difference lies in model quality or investment levels. But the truth is simpler, and more structural.

Generative AI works brilliantly for individuals because the cost of forgetting is low. 

When you use ChatGPT, it doesn’t need to remember much. You provide the context it gives you the output; you tweak and move on. The feedback loop is quick and cheap. If it forgets your tone or last week’s preferences, you just re-prompt it. You, the human, supply the memory – either directly in the prompt context window, or through third party tools like 

 (in our VC portfolio) which allows consumers to take their personal context with them across the internet. 

Inside the enterprise, forgetting is a much bigger problem. 

When an AI system forgets a compliance rule, it risks fines. When it forgets a customer’s history, it undermines trust. When it ignores internal feedback, it repeats the same mistake across thousands of users. 

Fixing that isn’t a quick re-prompt, it’s a full process clean-up. 

That’s why consumer AI thrives despite its amnesia, while enterprise AI stalls because of it. 

Shadow AI succeeds because humans absorb the learning. 

Enterprise AI fails because systems don’t. 

The learning gap that holds business back

Most enterprise AI systems today don’t 

 in any meaningful sense. They generate, but they don’t improve. 

Each interaction begins with a blank slate. No memory, no feedback history, no adaptation. 

MIT calls this the learning gap: the absence of feedback loops, context retention, and workflow integration that allow AI systems to evolve. 

 study echoes the finding. Nearly 80% of firms now use AI in some capacity, yet fewer than 20% can show bottom-line impact. The strongest driver of success? Workflow redesign

AI isn’t failing because it’s weak, it’s failing because we’re forcing it into old processes that were never designed for learning. 

Traditional enterprise workflows are deterministic: a sequence of fixed inputs, rules, and approvals. 

Generative AI thrives in the opposite environment; ambiguity, iteration, and feedback. When we drop AI into a rigid structure, one of two things happens: 

Either way, the learning loop never closes. 

Meanwhile, employees using personal AI tools have unconsciously redesigned their micro-workflows. They iterate, adjust, and feed their own feedback back in. Every prompt is a tiny experiment. That fluidity is exactly what enterprise processes need, but rarely allow. 

The companies crossing the GenAI Divide start with process redesign, not model selection. 

Rather than asking, “Where can we add AI?”, they ask, “Which parts of our workflow need to be rebuilt so AI can learn?” 

Together, these mechanisms turn a static tool into a self-improving system. 

The next wave of AI transformation won’t be decided by who has the biggest model or the deepest pockets. It will be decided by who can make their systems – and processes – learn. 

To cross the GenAI Divide, organisations need to do two things at once:  

This is how enterprises can unlock the same adaptability and impact that’s driven the success of ‘shadow AI’ and, ultimately, realise the trillion-dollar AI opportunity.

“The GenAI Divide: State of AI in Business 2025”

“The State of AI: How Organisations Are Rewiring to Capture Value”

Why does AI empower individuals but stumble when scaling across organisations? In our third blog on Large Language Models, Advisory Principal George Brain unpacks what MIT calls the "GenAI divide."

Advisory, reimagined

Trusted by the world’s best companies

We are outcome-focused

How we help

Technology Scouting

How we help

Strategy & Execution

How we help

Executive Support

Insight Reports

Technology Partnerships

Innovation Challenges

innovation stories

News

Web Summit 2025: Five Themes Shaping the Next Wave of Tech

Insight

The GenAI Divide: Why AI Works for People but Not for Companies at Scale (Yet)