Episode 09 — Hyper-Personalization without exclusion

HOST A:  I want to open with a sentence from the report that I think is the sharpest line in the entire piece. Personalization that adapts to what a shopper sees without adapting to how they access it isn’t truly personal — it’s exclusion.

HOST B:  That’s a genuinely uncomfortable reframe, because personalization is usually talked about as this purely positive, customer-obsessed trend. Nobody pitches it as a potential exclusion mechanism.

HOST A:  And to be clear, the underlying demand for personalization is completely real and well-documented. Seventy-one percent of shoppers expect personalized experiences. Ninety percent of brands report higher ROI from personalization efforts. And sixty-one percent of consumers say they’ll spend more with brands that personalize well.

HOST B:  So this isn’t a trend anyone’s debating the value of. The business case for personalization is about as settled as anything in eCommerce right now.

HOST A:  Completely settled. Which is exactly why the exclusion risk matters so much — because companies are pouring real investment into this, at scale, often without thinking at all about whether the personalized experience they’re building actually works for every customer it’s being personalized for.

HOST B:  Let’s make this concrete, because “personalization can exclude people” is a bit abstract until you see the actual mechanism.

HOST A:  Sure. Think about how personalization typically gets built. A lot of it relies on dynamically generated layouts — different users see different arrangements of content based on their behavior or preferences. It relies on visual recommendation modules — carousels of “you might also like” items, often image-heavy with minimal text. It relies on tailored promotional banners that shift based on browsing history.

HOST B:  And all three of those have an accessibility dimension that’s easy to miss if you’re only thinking about the personalization logic itself.

HOST A:  Exactly. If your personalization engine generates a layout dynamically, but that dynamic content isn’t built with proper semantic markup, a screen reader user encountering that page may get an incoherent, poorly structured experience — possibly worse than if the page had just been static, because at least static pages tend to get more thorough one-time accessibility review. If your recommendation carousel is purely visual, with no accessible alt text or keyboard navigation, a user with a visual impairment or motor impairment simply cannot engage with the personalized recommendations you built specifically to engage them.

HOST B:  So the irony is that the more sophisticated and dynamic your personalization gets, potentially the worse the accessibility experience gets too, if accessibility wasn’t part of how that system was designed.

HOST A:  That’s precisely the risk, and I think it’s an underappreciated one, because personalization and accessibility teams very rarely sit in the same room when these systems get built. Personalization tends to live with marketing or growth. Accessibility, when it’s owned at all, tends to live with legal or compliance. They’re not talking to each other, even though their work directly affects the same customer experience.

HOST B:  Let’s talk about what accessible personalization actually looks like when it’s done right, because I don’t want this episode to just be a list of failure modes.

HOST A:  The report gives some good concrete examples. Screen-reader friendly menus, so that as personalized navigation options shift, a screen reader user still gets a coherent, well-announced experience. Simplified flows for neurodiverse shoppers — personalization can actually work in their favor here, by reducing visual clutter and cognitive load rather than adding to it, if it’s designed thoughtfully. And user-controlled display adjustments, so that every customer can actually engage with their tailored experience on their own terms — things like adjustable contrast, adjustable motion, adjustable text size, layered on top of the personalized content itself.

HOST B:  That third one is interesting, because it reframes personalization slightly — it’s not just “the system decides what’s best for you based on your data,” it’s also “you get to decide how that personalized content is actually presented to you.”

HOST A:  I think that’s actually the more sophisticated and more genuinely customer-centric version of personalization. True personalization shouldn’t just mean algorithmic content selection. It should also mean giving the user control over their own experience parameters — which is, not coincidentally, exactly what good accessibility design has always been about.

HOST B:  Let’s get into the gap that currently exists, because I imagine most companies are not there yet.

HOST A:  They’re really not, and the report is direct about it. Twenty percent of eCommerce leaders admit they are not sure how to personalize effectively for shoppers with disabilities.

HOST B:  One in five leaders, openly admitting uncertainty, not even pretending to have an answer.

HOST A:  And I’d actually read that statistic somewhat charitably — it suggests an honest awareness of a gap, rather than a population that’s confidently building exclusionary systems without realizing it. The challenge seems to be less “we don’t think this matters” and more “we genuinely don’t know the playbook for doing this well yet.”

HOST B:  Which is actually a useful starting point, because it suggests the fix here is largely educational and procedural, not adversarial.

HOST A:  Exactly, and I think there’s a clear starting point for any team in that position. The same accessibility principles we’ve covered all season — proper semantic markup, keyboard operability, screen reader compatibility, user-controllable display settings — those don’t get suspended just because the content on the page is now dynamically personalized rather than static. The technical requirements are the same; they just need to be applied to a more dynamic content layer, which does require some additional engineering discipline.

HOST B:  Let’s talk about testing specifically, because I imagine testing a personalized experience is more complex than testing a single static page.

HOST A:  It is, meaningfully more complex, because by definition different users are seeing different versions of the experience based on their behavior and data. You can’t just audit “the homepage” once and be done — you potentially need to audit multiple personalization states, or at minimum the underlying components and logic that generate those states, to be confident the system holds up across the range of experiences it’s capable of producing.

HOST B:  Which probably argues for accessibility being built into the personalization engine’s components from the start, rather than tested after the fact on every possible output.

HOST A:  That’s exactly right, and it echoes something from earlier episodes — building accessibility in from the start is dramatically more efficient than retrofitting it after the fact, especially in a system where the number of possible output states is, practically speaking, enormous.

HOST B:  Let’s bring in the connection back to loyalty and trust, because I think that’s actually the more compelling business case here than pure risk avoidance.

HOST A:  The report makes this connection directly: accessible personalization unlocks loyalty. And I think the logic is intuitive once you say it out loud — if a brand has clearly invested in making sure its personalized experience genuinely works for you, specifically, including any accessibility needs you might have, that’s an extremely strong signal of how much that brand actually values you as a customer, compared to a brand that’s personalizing aggressively for everyone except you.

HOST B:  Which connects back to that sixty-one percent figure from earlier — consumers spending more with brands that personalize well. If personalizing “well” implicitly excludes a meaningful share of your customer base, you’re not actually capturing that full loyalty upside.

HOST A:  That’s the synthesis I’d want listeners to take from this episode. Personalization and accessibility aren’t competing priorities, even though they’re often built by separate teams who rarely talk to each other. They’re actually the same project, viewed from two different angles — both are fundamentally about making sure every individual customer gets an experience that genuinely works for them.

HOST B:  Let’s set up next episode, because I think this is a natural bridge into something bigger. We’ve now covered five major trends across nine episodes — litigation, cart abandonment, voice, AI search, and personalization — and in every single one, accessibility has turned out to be the underlying mechanism that determines the outcome.

HOST A:  Which is exactly what we’re naming directly next time: accessibility as a growth engine. Not a side effect of good ethics, not just risk mitigation — an actual driver of revenue and competitive advantage in its own right.

HOST B:  That feels like the right pivot point for the back half of this season. See you there.

HOST A:  Thanks for listening.