The last era of commerce and public services was built around people interacting directly with brands, platforms and institutions. AI has now inserted itself between them.
AI is increasingly acting as a mediator for how consumers and citizens discover, evaluate and access products, services and support — reshaping decision-making before organisations even know a journey has begun. This represents more than the arrival of another digital channel. It fundamentally changes how organisations are discovered, evaluated and chosen.
“The assumption that customers will engage directly with your brand is already outdated,” says Clare Allum, global head of consumer and manufacturing at PA Consulting, a global innovation consultancy. “AI intermediaries now decide what gets seen, compared and chosen. If you’re not designing for that layer, you’re not competing and you’re being filtered out. The winners will be those that design for machine judgement while keeping human trust.”
This shift presents a challenge for both public and private sector organisations — but also an opportunity to respond more intelligently. Becky Noble, a public services expert at PA, says: “We’re seeing the opportunity for AI to equip organisations with early insight to anticipate needs before they are explicitly expressed.”
From the attention economy to the recommendation economy
“For decades, organisations competed for attention. Increasingly, they’ll compete for recommendation,” says Allum. “It’s no longer enough to be visible; you need to be understandable and credible to the systems helping customers make decisions.”
Marketing, search and customer experience investments were designed to put brands in front of customers at the right moment. Success depended on memorability and persuasion.
AI is changing that equation. Recommendation engines may compare products using pricing, performance or other criteria — such as reviews or consumer sentiment — which surface strengths and weaknesses beyond the scope and control of traditional marketing.
It’s no longer enough to be visible; you need to be understandable and credible to the systems helping customers make decisions
“We’re shifting from a world where information was designed primarily for human consumption to one where it must also be interpreted by AI systems,” says Donald Cameron, a financial services expert at PA. “Being found and recommended increasingly depends on structured, detailed information that helps AI evaluate products and services.
”The same principle applies to public bodies,” says Noble. “Trust is not a brand attribute; it is legitimacy. Citizens need to know when AI is being used, what it has done, what it has not done, and how a human can challenge the outcome. This last principle is crucial, and we’ve seen it applied to good effect already in AI implementations; such as our work with HM Courts & Tribunals Service.”
Trust in a world of recommendation and delegation
If recommendation is becoming the new battleground, then the advantage will go to those able to deploy trust as a differentiator. When customers see recommendations provided by a large language model, trust is not limited to the accuracy of information, but also the source and brand behind it.
PA’s annual Brand Impact Index – a nationally representative survey of US consumers – shows that consumers feel more positive about, and are more likely to make purchases from, brands that deliver reliably and make life easier.
Allum adds: “They’re looking to organisations that reduce friction, solve problems, and earn trust. Brand leaders have a responsibility not just to deliver that but to ensure that AI engines know and surface this information.”
Brand leaders have a responsibility not just to deliver that but to ensure that AI engines know and surface this information
Cameron adds that some consumers are becoming more comfortable discussing sensitive financial topics with AI, highlighting both the opportunity to improve access and the need to maintain trust, empathy and oversight.
Yet he also warns that a focus on AI optimisation cannot come at the expense of the substance behind the experience.
“The quality of products and services, and the people behind them, need to withstand scrutiny,” says Cameron. “Especially as AI evaluates offerings based on evidence, outcomes and customer value, not just messaging.”
As intelligent systems compare products on evidence, outcomes and customer value, weak propositions will be harder to disguise behind marketing.
From reaction to anticipation
As AI moves from recommendation into delegation, it also creates the conditions for a more fundamental shift in how services are delivered — enabling organisations to anticipate needs, not just respond to them.
By combining AI, analytics and connected data, organisations can identify signals earlier and respond before customers or citizens actively seek support.
The opportunity is particularly pointed in the public sector, says Noble, where technology can bridge data gaps between departments.
Noble gives the example of an elderly person living alone. “There are lots of small breadcrumbs that might suggest something isn’t quite right,” she explains, from missed appointments to unusual financial patterns. Individually, these signals may appear insignificant. Together, they can indicate that intervention is needed and trigger support mechanisms earlier.
“People expect the services they use to work together around their needs, not organisational boundaries, because from their perspective it’s all part of the same life event,” says Noble. “Technology can play a key role in improving anticipatory services, ensuring citizens get the support they need in the moments that matter.
Singapore’s LifeSG platform demonstrates how this can work in practice. Designed around life events such as childbirth, retirement and job loss, it brings together government services and surfaces relevant support before citizens need to search for it.
Focus on the moments that matter
Yet becoming anticipatory does not mean trying to predict everything. The greater opportunity lies in identifying those touchpoints that deliver the most value; where timely intervention can prevent harm, build trust and improve lives.
“Not every touchpoint carries the same weight,” says Allum. “A small number of moments shape how customers judge you, determining whether they convert in that interaction and whether they choose to come back again, building loyalty over time.”
The focus on these moments should never come at the expense of the customer experience investments of the past decade, warns Allum. “Frictionless journeys, intuitive interfaces and strong service design all remain essential, but they are now table stakes because we’re all getting less patient and more likely to switch,” she adds.
Those that succeed in an AI-mediated economy will build on these foundations rather than replace them. They will become machine-legible, trusted and anticipatory: ensuring AI can evaluate what they offer, maintaining transparency and identifying needs before problems emerge.
As AI becomes the new front door to commerce and public services, competitive advantage will belong to organisations that are easiest to understand, easiest to trust and best placed to anticipate and act when it matters most.
Discover how to create intelligent experiences with PA Consulting
The last era of commerce and public services was built around people interacting directly with brands, platforms and institutions. AI has now inserted itself between them.
AI is increasingly acting as a mediator for how consumers and citizens discover, evaluate and access products, services and support — reshaping decision-making before organisations even know a journey has begun. This represents more than the arrival of another digital channel. It fundamentally changes how organisations are discovered, evaluated and chosen.
"The assumption that customers will engage directly with your brand is already outdated," says Clare Allum, global head of consumer and manufacturing at PA Consulting, a global innovation consultancy. "AI intermediaries now decide what gets seen, compared and chosen. If you’re not designing for that layer, you’re not competing and you’re being filtered out. The winners will be those that design for machine judgement while keeping human trust.”
