The Boundaries Products Are Quietly Redrawing

The moment that made it click

Last week, I sat in on a pricing review that was supposed to be routine. Early-stage SaaS, thoughtful team, solid traction. We were debating whether to simplify their plans—three tiers instead of five, fewer add-ons, cleaner page. The usual arguments surfaced quickly.

“Users are confused.”

“Competitors are simpler.”

“We need fewer decisions in the funnel.”

Then someone said something quieter: “But if we remove this option, we’re deciding for them.”

The room went still for a second. Not because the statement was dramatic, but because it landed uncomfortably true.

That same tension keeps showing up across the product conversations I’ve been following this week. Pricing guides promising simplicity. Platforms like eBay explicitly banning AI agents that act on users’ behalf. Articles questioning whether every user should really get the same interface. Ongoing debates about whether personas still matter.

On the surface, these are different topics. Underneath, they’re wrestling with the same question:

Where should the product draw the line between helping and deciding?

As a product manager, I’ve learned that most hard decisions aren’t about features or flows. They’re about boundaries. And right now, we’re collectively renegotiating where those boundaries belong.

Simplicity isn’t neutral—it’s a choice

Let’s start with pricing, because it’s the most honest interface most products have.

There’s a reason “simple SaaS pricing” content keeps circulating. According to a 2023 ProfitWell analysis of over 400 SaaS companies, reducing the number of pricing plans can increase conversion rates by up to 20%—in the short term. That stat gets cited a lot. Less often discussed is what happens six months later.

In my experience advising teams, simplified pricing often does three things at once:

It reduces cognitive load for first-time buyers
It forces internal clarity about who the product is really for
It removes edge-case flexibility for users who don’t fit the median

That last point is where the discomfort lives.

I once worked with a B2B product serving operations teams across wildly different company sizes. When they collapsed their pricing into two clean tiers, self-serve signups jumped. Support tickets dropped. The dashboard looked healthier almost immediately.

But three months later, sales started flagging a different signal: their most sophisticated customers—often the most profitable long-term—were negotiating harder, asking for exceptions, or quietly churning at renewal.

The product hadn’t become worse. It had become more opinionated.

And here’s the thing we don’t say enough: opinionated products are not automatically respectful products.

Simplicity decides what matters. And when we simplify, we’re not just removing friction—we’re redistributing agency.

Practical insight I’ve learned the hard way:

If you simplify pricing, explicitly name who you’re no longer optimizing for
Track not just conversion, but who converts
Watch renewal and expansion behavior, not just signup velocity

Simplicity without acknowledgment of its trade-offs isn’t clarity. It’s avoidance.

When automation crosses into authorship

The eBay update banning AI “buy for me” agents landed quietly, but it says something loud about where platforms are drawing boundaries.

At first glance, it looks defensive—protecting marketplace integrity, preventing abuse. All true. But zoom out, and it’s part of a broader discomfort with tools that don’t just assist users, but act as them.

This matters for product design because agency is not binary. It’s graduated.

There’s a meaningful difference between:

Tools that help me decide
Tools that suggest what I might do
Tools that take action in my name

Most UX patterns collapse these distinctions too quickly.

I’ve watched teams justify aggressive automation by pointing to efficiency metrics. Fewer steps. Faster completion. Higher task success rates. And yes—those numbers often look great.

But here’s the data point that keeps nagging at me: a 2024 survey by Pew Research found that 62% of users feel uneasy when software takes actions they don’t fully understand, even if the outcome is positive.

That unease doesn’t always show up in usability tests. It shows up later, as hesitation, mistrust, or over-cautious behavior.

eBay’s move is interesting not because it limits AI, but because it clarifies responsibility. You are the buyer. You click purchase. The platform refuses to blur that line.

As product leaders, we should be asking:

Where does our product stop assisting and start authoring?
Do users understand when that handoff happens?
If something goes wrong, do they know who’s accountable?

Automation that obscures authorship may feel helpful today—and erode trust tomorrow.

The myth of the “average” user interface

Another thread that keeps resurfacing: the hidden cost of giving every user the same interface.

On paper, consistency is kind. It’s fair. It’s easier to maintain. But fairness and sameness are not identical.

I’ve seen this tension most clearly in enterprise products. A designer will argue for progressive disclosure or role-based views. Engineering will worry about complexity. Product will compromise with a “mostly universal” interface.

And then research sessions reveal something uncomfortable: users bending themselves to fit the product.

One study from Nielsen Norman Group showed that personalized interfaces can improve task efficiency by 15–25% for experienced users—but can hurt novices if done poorly. That nuance often gets lost in the personalization debate.

The deeper issue isn’t whether personalization is good or bad. It’s whether we’re willing to admit that different users are doing fundamentally different jobs, even inside the same product.

Personas get dragged into this argument a lot. “Do they still matter?” people ask. I think that’s the wrong framing.

What matters isn’t the artifact. It’s the act of recognizing difference.

When teams abandon personas without replacing the underlying observational work, sameness creeps in by default. The interface becomes optimized for whoever speaks loudest in roadmap discussions—or whoever the product lead unconsciously identifies with.

A practical test I use with teams:

Ask, “Who does this screen frustrate the most?”
If no one can answer concretely, you’re probably designing for an imaginary average

Designing one interface for everyone often means designing deeply for no one.

Boundaries are where trust is built

What ties all of these conversations together—pricing, automation, interfaces—is a shared discomfort with boundaries.

Boundaries decide:

What the product takes responsibility for
What the user remains responsible for
Where judgment lives

When those boundaries are clear, users feel oriented—even if the product is complex. When they’re blurry, even simple products feel unsettling.

I’ve learned to look for trust signals that don’t show up in dashboards:

Users double-checking automated actions
Customers asking for “manual modes” or overrides
Support tickets that say, “I just want to understand what happened”

These are not signs of failure. They’re signs that people are trying to locate themselves inside your system.

As product leaders, we often chase ease. But ease without orientation is fragile.

Some questions worth sitting with:

Where does our product intentionally slow people down?
What decisions do we refuse to make on the user’s behalf?
How do we signal those boundaries—not just document them?

The teams I respect most aren’t the ones shipping the fastest or automating the most. They’re the ones willing to say, “This is as far as we go—and here’s why.”

Choosing where to stand

Back in that pricing meeting, we didn’t resolve the debate cleanly. We ended up keeping an option that complicated the page—but preserved user choice for a specific segment. It wasn’t elegant. It was intentional.

That’s what these product conversations are really circling: intentionality.

Not whether to simplify, automate, personalize, or standardize—but where, for whom, and at what cost.

Products don’t just solve problems. They teach people how much control they have.

And in a moment where software is increasingly confident, decisive, and fast, the most human thing a product can sometimes do is draw a clear line—and invite the user to step over it themselves.

That’s not indecision. That’s care.

And care, in the long run, is what people feel—even when the metrics look the same.