What 800 signals actually buys over a traditional segment model
Most e-commerce personalization still runs on customer segments built from historical purchase data, a shopper gets bucketed into a category based on what they bought before, and offers get tailored to that bucket. Made With Intent's platform works differently: it analyzes more than 800 behavioral signals in real time, meaning the system is reading how a shopper is behaving on the page right now, hesitation patterns, scroll behavior, comparison browsing, rather than relying solely on what that shopper or similar shoppers bought in the past.
That real-time behavioral layer is what allows the system to determine both what a shopper sees and precisely when they see it, surfacing a discount offer or a payment option at the specific moment a behavioral signal suggests the shopper is wavering, rather than displaying the same static offer regardless of where the shopper actually is in their decision process. This moment-level targeting is the core technical differentiation from a segment-based personalization system.
The revenue number, and what it actually measures
Sarah Hanna, Digital Director at The Very Group, reported revenue per customer increasing by a fifth in one of the company's agentic strategies within the first month of the partnership, alongside an uplift in overall conversion. Revenue per customer rising specifically, rather than just total revenue, is a meaningful distinction: it suggests the platform is extracting more value from existing shoppers through better-timed offers and relevant product surfacing, not simply driving more traffic to the site.
The conversion uplift reported alongside that revenue figure is the more operationally significant number for other retailers to weigh, since it indicates the personalization is helping shoppers who were already engaged actually complete a purchase, rather than just increasing the average basket size of shoppers who were going to buy regardless. A platform that lifts revenue per customer without lifting conversion would be a weaker signal than this combined result.
Why payment-moment personalization is the less obvious feature
Beyond product and offer targeting, the platform also personalizes payment options presented to individual shoppers, a feature that gets less attention than product recommendations but addresses a genuinely underexplored friction point in e-commerce checkout. Different shoppers have different payment preferences and different sensitivity to financing or installment options, and presenting the wrong payment method prominently, or failing to surface a buy-now-pay-later option to a shopper who would have used it, is a quiet but real source of abandoned carts.
Retailers evaluating personalization platforms often focus evaluation criteria entirely on product recommendation quality and overlook payment-moment customization as a separate, addressable lever. The Very Group's results suggest that lever is contributing to the broader conversion uplift, which argues for retailers to treat payment personalization as a distinct evaluation criterion rather than an afterthought bundled into a broader recommendation engine purchase decision.
Reading an early, vendor-adjacent result with appropriate care
These results come from one month of operation and are self-reported by both the retailer and the vendor's founder, which is standard for an early case study but means the numbers have not been independently verified or tested against a longer operating period that would reveal whether the initial lift holds, decays, or compounds over time. First-month results in any new personalization deployment often benefit from a novelty effect as the system is still calibrating against fresh behavioral data it has not seen before.
That caveat leaves the result's relevance as a signal fully intact, while still arguing for treating it as an early proof point rather than a guaranteed outcome any retailer could replicate by simply adopting the same platform. The specific fifth-of-revenue lift is tied to The Very Group's particular customer base, product catalog, and prior personalization maturity level, all of which shape how much headroom existed for a new system to capture in its first month.
What this means for retailers still running segment-based personalization
The broader signal for any retailer still relying primarily on historical segment-based personalization is that real-time behavioral signal analysis is producing measurably different results, at least in this early case, than the segmentation approach most e-commerce personalization stacks were built around over the past decade. Segment-based approaches still have a place in a broader personalization stack, and the gap this case highlights is wide enough to be worth testing directly against current performance regardless.
Retail technology leaders evaluating their own personalization roadmap should treat the 800-signal, real-time behavioral approach as a distinct category worth piloting separately from incremental improvements to an existing segment-based system, rather than assuming the next version of their current platform will eventually close this gap through gradual feature additions. The architectural difference between the two approaches is deep enough that incremental upgrades to a segment-based system are unlikely to fully replicate what a purpose-built real-time behavioral engine delivers.
The industry recognition context worth noting
Made With Intent's inclusion on the Retail Technology Hot 100 List, an industry recognition list compiled independently of this specific partnership, provides some external context for The Very Group's choice of vendor beyond the self-reported results. Direct evaluation still carries more weight than any recognition list, though this one does indicate the platform had already drawn attention and credibility within the retail technology community prior to this specific case study being published.
For retail technology leaders building a vendor shortlist for personalization platforms, that kind of independent industry signal is a reasonable input for narrowing a shortlist, though it should sit alongside direct technical evaluation and reference calls with existing customers, including ideally a customer running a comparable catalog size and complexity to your own, rather than resting on the recognition list or the vendor's own case study alone.
