Teams still design for an imaginary buyer. In reality, AI and algorithms filter choices long before people decide. Here’s how to adapt.
Open Amazon, search for almost anything, and most people will choose from whatever shows up first. That choice feels personal. But it’s often the output of a system: ranking, badges, availability, reviews, shipping speed, and paid placement all shaping the shortlist before a human really compares anything. This is the gap personas can’t fully explain anymore.
For decades, marketing teams have relied on personas as a core planning tool. A persona is a fictional, research-informed “stand-in” for a group of customers—usually described with basics like age and lifestyle plus motivations, habits, and what might persuade them. Personas help teams align on who they’re building for, what messages might resonate, and which channels to prioritize. They worked best when brands had a relatively direct path to people. That path is now heavily mediated.
Personas assume attention. AI systems assume efficiency. Today, many purchase decisions are shaped before someone ever actively chooses. Algorithms decide what appears. Ranking systems narrow the list. People still decide—but they increasingly decide inside systems that have already structured the choice set.
Personas assume visibility. Platforms control it
Traditional personas assume a person encounters a meaningful set of options and then evaluates them. In practice, people rarely see the full category. Platforms decide what surfaces, in what order, and with what framing. That changes what “good marketing” even means. It’s not only the story you tell. It’s whether the system lets you show up at all—and whether your story survives the way the system packages it.
Deep dive: Amazon and the illusion of consumer choice
Amazon is one of the clearest examples of how decision-making has shifted from persuasion to system design. Amazon’s ad products explicitly sell premium visibility, including placements at the top of search results, which means some of what looks like “the best options” is also “the best-positioned options.” Amazon also layers shortcut signals into the experience. Badges like “Amazon’s Choice” and “Best Seller,” plus star ratings, delivery promises, and review volume, act as permission slips that let shoppers stop searching earlier than they otherwise would.
From the shopper’s perspective, it feels like comparison shopping. From the system’s perspective, it’s a ranking and filtering exercise that happens before the shopper has meaningfully compared alternatives. By the time someone clicks, the decision space has already been narrowed. So the decisive question isn’t “What would this persona want to hear?” It’s “What does the system reward—and are we building for that reality?”
From personas to decision pathways
Here’s an alternative that holds up better in a mediated environment: replace static personas with decision pathways. Instead of trying to perfect a profile of “who” the customer is, map how a choice actually gets constructed and delivered. You can think of it as three linked moments: what gets you surfaced, what keeps you on the shortlist, and what makes the final click feel like the right call.
Start with the algorithmic trigger. What causes the product to show up at all? On marketplaces and feeds, this can include relevance signals, engagement signals, pricing and availability, and paid placement. Then map the filter criteria. What gets you removed from consideration? Weak reviews, slow shipping, confusing pages, unclear differentiation, or claims that don’t survive being compressed into a tile, snippet, or quick summary. Finally, define the human decision point. What does the person need to feel in the last mile—relief, confidence, identity, delight—so choosing you feels easy rather than risky?
A simple template teams can use is: “Our product needs to rank for [X signals], survive compression as [Y hook], and deliver emotional reassurance through [Z proof points].” The system decides whether you get a chance. Emotion decides whether you get chosen.
The new emotional job isn’t persuasion. It’s the right feeling at the right moment
People still buy with emotion. But emotion shows up differently when choice is filtered. In mediated environments, the dominant emotion for many everyday purchases is risk management: “Will I regret this?” “Will this be a hassle?” “Is this legit?” That’s why ratings, shipping promises, return policies, and familiar cues often outperform beautiful storytelling at the point of selection. They create comfort when the person is deciding quickly.
This doesn’t kill emotional marketing. It relocates it. A lot of emotional work now happens at the moment of confirmation, not discovery. Brands still need a point of view and a voice; they just need those things to translate into proof and clarity at the moment a system has already narrowed the list.
Liquid Death shows what still works: emotional clarity that survives compression
Liquid Death is a useful counterweight to the idea that brands must feel “safe” to connect. It’s intentionally loud—tallboy cans, heavy-metal aesthetics, and a provocation as brand posture. The more rigorous lesson is that Liquid Death’s “risk” is mostly aesthetic. The underlying choice is low-stakes. It’s still water. The brand takes cultural risk so the buyer doesn’t have to.
It’s also built for mediated environments. The idea reads instantly in a thumbnail, a shelf glance, or a two-line mention. You don’t need a manifesto to understand it. That portability is the point: differentiation works when it can survive compression and still feel like itself. To be precise, this isn’t proof that “compression-friendly hooks always win.” Plenty of bold brands fail because they don’t back the hook with distribution, repeat exposure, or product experience. Liquid Death’s tighter lesson is that a sharp human truth, repeated consistently, can travel through modern interfaces without losing its emotional charge.
Where this framework matters most, and where it matters less
This decision pathway model matters most when platforms and algorithms heavily shape visibility and shortlists. Think everyday goods, marketplace purchases, search-led decisions, and any category where ranking and recommendations do the gating. It matters less when the primary driver is extended deliberation, relationships, or lived experience—high-consideration purchases, deep loyalty categories, or situations where a trusted human advisor is central. In those contexts, personas can still capture the primary dynamic because the system isn’t doing as much of the filtering.
The point isn’t to throw personas away. It’s to stop pretending personas are the whole map.
Final Thoughts
Personas aren’t useless. They’re incomplete. The best teams do both: they understand the human truth—what people want to feel—and they engineer for the systems that decide whether that feeling ever gets a chance to land.
The tension is about to sharpen. As AI summaries and compressed interfaces get better, will differentiated brands survive translation—or get flattened into generic labels? Platforms will keep evolving their filters. The brands that thrive won’t just optimize for today
Leave a comment