Atul Sabharwal is a founder and CEO of Snipp, a leading AI-powered value-added SaaS company in the global loyalty and promotions sector.
getty
Loyalty programs spent the last few decades moving beyond points and discounts toward experiences, personalization, status and emotional affinity. The strongest brands understood a basic truth: Customers don't stay loyal just because the math works.
But when a shopper is being represented by an AI agent, what exactly will loyalty be measured against? And, more importantly, how do we know who we are actually rewarding?
Agentic AI isn’t a hypothetical: It’s already present and active, browsing product pages in growing volume. HUMAN’s 2026 State of AI Traffic & Cyberthreat Benchmark Report found that traffic from AI agents and agentic browsers grew 7,851% year over year in 2025, with retail and e-commerce, streaming and media, and travel and hospitality accounting for more than 95% of that traffic—sectors where loyalty programs are often central to customer value.
For loyalty leaders, the more important development is that these agents are beginning to operate within logged-in sessions on behalf of users, navigating accounts, authentication flows and checkout.
A human may choose a brand because of a remembered preference, a status benefit, a personalized offer or simply the feeling of being known. An agent may reduce the decision to price, availability, convenience, redemption value and friction. In that sense, loyalty may be entering a strange full-circle moment: After years of moving from transactional rewards to emotional engagement, AI may reintroduce a more logical, ROI-driven filter at the point of decision.
This creates two problems for loyalty. The first is legitimate agentic commerce: trusted AI assistants acting on behalf of real customers, helping them discover products, compare offers, apply coupons or redeem loyalty benefits. For brands, this is not inherently a bad thing. It can become a valuable new interface for high-intent customers.
But many brands aren’t ready for this: PayPal found that most merchants can see at least some AI-agent activity on their sites, but only about one in five have most of their product catalog structured in a way those agents can easily read and use. And product data is only the first step. Loyalty data will be next: Tier rules, points balances, member benefits, eligibility and redemption options also need to be clear enough for trusted AI systems to understand. Without that structure, brands risk becoming invisible when agent-assisted shoppers are ready to engage.
The second problem is AI-enabled abuse. Historically, more users, more enrollments, more clicks and more redemptions were often treated as positive signals. In an agentic environment, those metrics become less reliable.
Bad actors can (and already do) use automation to game the system. Points, rewards and discounts all have value, but many programs are still secured like marketing databases, not financial systems.
And while the financial cost is a huge part of this problem, the data damage may be just as significant. If brands can’t tell whether engagement is human, agentic or fraudulent, they may optimize campaigns around distorted signals.
So what should leaders do now?
"More activity" is no longer automatically good activity. The better question is: Which activity reflects real customer value? If you can’t answer that, your growth numbers, conversion data and loyalty metrics may already be muddied.
Look beyond payment and checkout to eligibility, enrollment, login, referral, redemption, account recovery and points-transfer flows. In agentic commerce, brands must verify both the customer and the agent acting on their behalf. These flows also need to withstand automated attacks at scale.
Even if the overall journey is secure, individual promotions can still create loopholes. Every offer should be evaluated not only for customer appeal, but for automation risk. Can it be claimed repeatedly? Can accounts be created cheaply to exploit it? Can rewards be stacked, transferred or redeemed faster than abuse can be detected? If so, the offer should be redesigned before the campaign goes live.
Which of your loyalty benefits are actually legible to a machine? Benefits that show up only in marketing copy may be invisible at the moment of decision. Tier pricing, points balances, member benefits and stacked promos need to be exposed in ways a trusted agent can parse.
None of this means emotional loyalty stops mattering. It simply moves earlier in the journey, to the moment when customers decide which brands they trust enough to let their AI agent choose from. The transaction layer may become more dispassionately logical, but the relationship layer still belongs to people.
That means loyalty leaders need to preserve the emotional benefits that give people a reason to stay—experiential rewards, status recognition and community—while making rational benefits clear enough for trusted agents to understand and act on.
The next loyalty challenge will be navigating a hybrid market where nonhuman users are not all the same: Some will be legitimate, some opportunistic and some outright fraudulent. The companies that adapt most effectively will stop asking, "How do we get more users?" and start asking, "Which users can we trust, and can they understand the value we are offering?"
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。