Building trust in digital-first brands: What we learned at All Access: Digital CX 2026

Research conducted for the CX Network report CX Horizons: The Global State of CX 2026, found that AI-powered technologies for operations is the top CX trend among practitioners for the second year in a row. Signaling where the market is heading, agentic AI and AI agents made the top 10 trends for the first time in 2026, while AI-first customer journeys also ranked highly as CX trends and customer behaviors.

AI governance, however, was found to be maturing but incomplete. The percentage of organizations with an organization-wide approach to generative AI best practice and governance grew six percent between 2025 and 2026, but 20 percent reported still having no policy at all. 

It was in this context that we ran All Access: Digital CX 2026. We delved into what it means to build trust as a digital-first brand, and how to meet customers where they are in a post-funnel world. We also dug into the key topics around successfully deploying AI for smoother customer journeys and increased personalization, while reflecting on what the rise of agentic commerce will mean for digital CX moving forward, and taking an honest look at how to get started with AI implementation. 

If you missed the series live, we've got you covered. You can watch all sessions on-demand here and read on for our roundup. 

Don't miss any news, updates or insider tips from CX Network by getting them delivered to your inbox. Sign up to our newsletter and join our community of experts. 

Learn More

How Fun.com overhauled search and recommendations with vector AI

The proliferation of LLMs is having an unanticipated influence on ecommerce, with customers increasingly using long-tail terms for search and discovery as they become more used to describing what they want using natural language. 

Fun.com had accumulated more than 5,000 manual redirects trying to compensate for a legacy search engine that wasn't equipped for this change. Vector AI, says Dianna Lyngholm, director of website and creative services at Fun.com, is the way forward. It learns over time, based on users' clicks, hovers, and browsing behavior. 

"Think of a chat when you go to GPT or Gemini. If you start a new chat, it doesn't remember what you were talking about before. Same thing with LLM search. But with vector, you're understanding the user and the user behaviour, and long-term it continues to get better and better," Lyngholm explained. 

After rolling out vector AI, the costume and novelty retailer was able to eliminate the 'no results found' page and drive a 25 percent increase in search revenue. Revenue Per Visitor (RPV) is the north star metric, and vector AI-driven personalized recommendations has lifted this by 31 percent. 

Fun.com spent two years exploring and testing the vendor market before settling on Malachyte, a relatively new player in the game. Lyngholm recommended partnering as closely as possible with vendors to see the best possible results. She said: "We have a separate Teams and Slack channel with them, we have weekly calls. You feel like they're part of your family, they're not just this vendor where you set it and forget it."

Watch the full session on-demand here.

The three-point framework for AI readiness

Shailesh Nalawadi, chief transformation architect at delight.ai from Sendbird, joined the series to lay out a framework for AI readiness. He noted that many organizations treat AI deployment the same as any other software purchase. But AI agents, he explained, touch legal, compliance, engineering, content, and CX, so treating it as a single-team decision can cause pilots to fail. 

"You often discover, after the contract is signed and you're partway through the implementation process, that nobody really has authority to approve the decisions that are actually required to deploy," he said. 

To mitigate this, he recommended setting up an "AI council", before even beginning to evaluate vendors. It must be cross-functional, including an executive sponsor, CX, operations, product, engineering, legal, and compliance. He advised: "Assemble the council before you start, not as a formality, but really as a forcing function. In that first meeting, write down the decision rights. Who makes the call when engineering and legal disagree? Who owns a blocking issue at 2AM during implementation? All of this sounds procedural, but it has a direct bearing on how fast you move."

He emphasized three key pillars of AI readiness: organizational, infrastructural and strategic, warning that "you can't compensate for weakness in one with strength in another." 

When asked which of these three pillars is most challenging, Nalawadi said infrastructural, specifically relating to knowledge base quality. Knowledge bases designed for humans do not translate to AI, because humans bring a level of instinct and judgement when looking at imperfections that AI simply does not "AI does a great job of automating at scale, but if you give it a messy system, it will automate the mess at scale," he explained. 

When asked about the practicality of implementation, Nalawadi recommended using a three-tiered framework, with level one covering FAQ and knowledge retrieval, in which you can achieve "close to 100 percent automation." Level two can pull live customer data and level three can make changes to systems. The higher the level, the more governance and testing is needed.

Watch the full discussion on-demand here.

Redefining customer journeys in the age of AI and machine customers

Customer journeys are no longer linear, and the "funnel approach" is no longer fit for purpose. 

This was the premise on which we began the panel discussion between Anne-Kathrine Nissen, experience design lead at H&M, and Liliana Caimacan, professor of marketing, innovation and strategy at Hult International Business School. 

"The big shift we see is moving from the funnel, and moving customers to different stages of the funnel, to really maintaining a relationship with the customer. It is about a flywheel... we need be part of their life," said Caimacan. 

The agentic commerce space is developing quickly and brands are losing control of their own channels – and how customer interact with them. With AI agents shopping on behalf of customers, brands need to look inwardly and define how they want to show up when customers are not directly in front of them. 

Nissen said: "We are increasingly having to let go of how our brand is seen in all these other channels. We have TikTok shops, we have iOS 16, we have agentic commerce. It's important that there is cross-functional alignment. We set the vision for how we would like it to be and how we would like to come across as a brand."

Developments in AI are also paving the way for exciting new personalization opportunities, but these will fall flat when simple selling is prioritized over relationship building. Transparency around data use is essential for maintaining trust, especially post-GDPR: 
"If personalization is about what they should buy next, then it's not a relationship - it's a monologue, not a dialogue with the customer. Less is more, unless the customer shows a very clear intent," warned Nissen. 

Both speakers also pointed to the challenge of translating data into meaningful action across organizations that are often very siloed. 

Building internal buy-in for CX transformation requires internal selling. Nissen pointed out that people generally have a bias towards loss – they know what they have and fear what change may cost. 

She advised: "You have to have your people on the other side perceive that their winnings will be bigger than the loss they will experience when you change their daily life and their daily tools. You have to create clarity about what has to change, create an understanding of what it takes to get there, and establish that strong dissatisfaction with the status quo, while you paint the picture of how much better the future will be."

Watch the full discussion on-demand here.

Designing for CX flow with a 'Journey Atlas' approach

Customers do not experience touchpoints, they experience continuous journeys irrespective of internal siloes, departments, and product teams. They "have an end-to-end experience, and this requires end-to-end collaboration and synergy", said Eelko Lommers, director of CX design and innovation at zooplus. 

Flow is what will buy customer loyalty. Borrowing from Kahneman's System 2 Thinking for complex reasoning, Lommers explained why friction is so costly.

"Transactional experiences cultivate transactional customer behavior, which is amazing if you're going for a singular transaction, but not so great if you want customers to come back because they loved your experience or your company so much […] The journey is only how we move. The experience is what we remember. No flow, no go," he said.

Zooplus built the 'Journey Atlas' to map customer lifecycles as "jobs to be done", rather than business processes: "We map journeys from the perspective of the needs of customers, and not the needs of the business. We track jobs to be done, and if a customer has a need, we aim to understand how they get to that need, and what steps they would normally take."

When looking ahead, Lommers said that the future of CX lies in content feeds, not in fixed pages. Zooplus is looking at the acceleration of agentic commerce and is building a modular, token-and-component design to enable dynamic and personalized content streams.

Lommers said: "The trust we build through consistent page, touchpoint and platform experience will soon come from consistent content streams... Our company and brand become a story, a product, and a service multiverse – all part of the same zooplus universe."

Watch the full presentation on-demand here.

How Back Market builds trust as a digital-first brand in an unfamiliar category

Refurbished tech is a nascent and low-frequency category. This brings a challenge to Back Market UK, the refurbished tech brand that actively discourages overconsumption with a goal to extend device lifecycles from 18-24 months to five years. 

In this context, traditional promotional marketing doesn't fit. During his session, Luke Forshaw, head of brand and marketing at Back Market UK, emphasized the importance – and challenges – of staying relevant between transactions in such a marketplace. 

"The way that we go about promoting ourselves as a company is hugely important, because promoting an oversell doesn't fit with what we want to be as a brand, and I think people are getting quite tired of constantly being sold to," Forshaw said. 

In an unfamiliar category, distrust can show up anywhere in the funnel, and Back Market's in-house team responds proactively. "Whenever there's a negative experience that we see – whether that's someone on their own social feed, someone in our social feed, someone popping up on Trustpilot - we're in that space trying to understand what their problem is, solve the issue, and move it swiftly to DMs." 

Part of building trust is rooted in design and online experience, both of which have had an overhaul under Forshaw's watch. Bringing design in-house has enabled brand consistency across 17 markets. 

Forshaw explained: "At a time where we're trying to grow trust not just the brand, but also the category, we felt that being more uniform was simply the best thing to do. By bringing the work in-house, we mitigated third parties that, through no fault of their own, would have done things slightly differently." 

Simplifying the online experience while empowering customers with greater choice has driven impactful commercial results. "Since we've made the changes, dwell time is higher, AOV is higher, conversion rate is higher. Despite bringing more people into the platform, we're not seeing an increased amount of wastage, we're seeing far more engagement. We feel that's done by making the process simpler, by putting more content out there as and when people want it, but not forcing it down their throats."

When asked for parting advice for digital-first brands entering unfamiliar categories, Forshaw said: "Know what your consumer wants. Speak to people, do research sessions, talk to people on the street about what they want, what they need, what their concerns are, and use that to shape how you offer the brand. The users are what is going to make or break your business."

Watch the full session on-demand here.

Quick links

• What happened at All Access: The AI Revolution in CX
• The Future of Customer Listening: dynamic, active, and AI-powered
• How the big brands are revolutionizing CX design and journey management