CUSTOMER EXPERIENCE

What Netflix Knows That Your Bank Doesn’t

Streaming platforms don’t have more data than your favorite brand. They just never stop paying attention.

Photo by BoliviaInteligente on Unsplash

Netflix has never asked my age. It has never asked where I live, what I do for work, or how much I earn. It doesn’t have my name on a birthday list. And yet it has recommended, with unnerving accuracy, three shows in a row that I finished in a single weekend.

Meanwhile, my bank, which knows my exact income, my mortgage balance, my spending categories down to the dollar, and has had my name and date of birth on file for eleven years, emailed me last month to wish me a happy birthday and suggest I “explore our travel rewards card.” I don’t travel. There is a decade of transaction history proving I don’t travel. The bank has more raw data about me than Netflix will ever have. And it is, by a wide margin, worse at knowing me.

That gap is the whole story. It isn’t a data problem. It’s an attention problem.

The company that has less information and gets it more right

Photo by Sushil Nash on Unsplash

Here’s the part that should bother every marketer who’s ever pitched a “360-degree customer view”: Netflix’s advantage was never the size of its dataset. Roughly 80 percent of what people watch on the platform originates from its recommendation engine, and that engine runs almost entirely on one type of signal, what you actually did. What you finished. What you abandoned nine minutes in. What you rewatched twice in a month. What time of night you reach for a documentary instead of a thriller.

It has none of the identity information a bank, an airline, or a retail loyalty program collects as a matter of routine. No income. No address. In most cases, not even a real name. It has behavior, and it has treated that behavior as a live signal instead of a one-time form.

That’s the actual difference. Most brands collect identity once, at signup, and then treat it as permanent truth. Netflix, Spotify, and Amazon collect behavior continuously, and treat every new action as a small update to a belief that’s always slightly incomplete. One approach freezes a customer in time. The other keeps watching.

Two operating models, not two datasets

Photo by Brett Jordan on Unsplash

It helps to think about this as two different systems, not two different amounts of data.

The first is what most companies actually run: an address book. You fill out a form once, maybe update it occasionally, and the company files that information away as the definition of who you are. Every message that follows gets built on top of that static file. Your name goes in the greeting. Your age determines which segment you fall into. Your zip code decides which regional promotion you receive. Nothing about this system updates itself based on what you do next, because it was never designed to watch, only to file.

The second is what Netflix, Spotify, and Amazon actually run: an attention system. Every click, skip, pause, and repeat view feeds back into a profile that’s never considered finished. There’s no moment where the system declares “we now know this customer” and stops updating. It assumes it’s always slightly wrong and keeps correcting.

An address book updates once. A behavior model updates constantly.

This is why a customer who watched three seasons of a crime drama last month and then switched entirely to cooking shows this month gets different recommendations almost immediately on Netflix, but keeps receiving the same category of email from a retailer for the next six months. The retailer isn’t lacking data. It’s lacking a system that treats yesterday’s behavior as more relevant than last year’s signup form.

The cold start problem most brands never solve

Photo by Aerps.com on Unsplash

Engineers who build recommendation systems talk about something called the cold start problem: what do you show a brand-new user before you have any behavioral data on them? Streaming platforms treat this as a temporary, uncomfortable phase to get through as fast as possible, usually by asking a few quick preference questions and then correcting rapidly once real behavior starts flowing in.

Most brands never leave the cold start phase. The demographic profile collected at signup becomes the permanent substitute for behavior, forever. A customer who joined an email list two years ago as a 24-year-old renter might now be a 26-year-old homeowner with entirely different needs, and the brand has no idea, because nothing after the signup form ever got recorded as meaningful.

The cold start problem isn’t a technical constraint most brands have failed to solve. For most brands, it is the entire personalization strategy. They never move past the first guess.

Why this shows up as a trust problem, not just a relevance problem

Photo by Towfiqu barbhuiya on Unsplash

It would be easy to write this off as a minor inefficiency, a slightly-off email here and there. The research suggests it’s more damaging than that.

Gartner’s 2025 survey of nearly 1,500 B2B and consumer buyers found that personalized marketing produced a negative experience for 53 percent of customers, who were 3.2 times more likely to regret their purchase and 44 percent less likely to return. That statistic gets read as evidence that personalization itself is the problem. It isn’t. It’s evidence of what happens when a company acts confident about something it actually knows very little about. A message built on a two-year-old signup form, delivered with the tone of “we know you,” reads as presumptuous the moment it’s wrong. And a static profile is wrong constantly, because people change and forms don’t.

Attentive’s 2025 research on consumer attention backs this up from the other direction: 8 in 10 shoppers now actively ignore messages that feel irrelevant, and a full quarter say a generic message makes them less likely to purchase at all. Customers have learned to read the difference between a company that’s watching and a company that’s guessing, and they’ve stopped extending the benefit of the doubt.

The reward for closing that gap is not small. McKinsey’s research on personalization leaders found revenue lifts of 5 to 15 percent and acquisition cost reductions of up to 50 percent, concentrated specifically among companies that built real behavioral capability rather than fancier segmentation. The money was never in knowing more names. It was in reacting faster to what people actually did.

The honest limits of the comparison

Photo by SK Sherazi on Unsplash

It’s worth being fair to the marketers reading this and rolling their eyes at the Netflix comparison, because the comparison has real limits.

Streaming platforms have a structural advantage almost no retail or service brand shares: a single, continuous interface where every interaction happens in one place, generating clean behavioral data by default. A bank, a clothing brand, or a B2B software company collects behavior across a dozen disconnected systems, a website, an app, an email platform, a support line, a physical store, and most of those systems don’t talk to each other. Building one unified attention model across all of that is a genuinely harder problem than anything Netflix has to solve, and it’s fair to say the comparison undersells the difficulty.

There’s also a ceiling on how far behavioral tracking should go before it becomes its own kind of failure. Gartner’s same research found that even accurate personalization increased feelings of being rushed or overwhelmed when it wasn’t paired with some way for the customer to stay in control. Watching someone closely and acting on it without their sense of consent is exactly what produces the “how did they know that” discomfort brands are rightly afraid of. Netflix gets away with constant behavioral tracking partly because the exchange is transparent and expected. A retailer trying to replicate that level of tracking without the same transparency risks crossing from attentive into invasive.

The lesson isn’t “become Netflix.” It’s narrower than that: stop treating the signup form as a finished picture, and start treating behavior as the thing that’s actually still being written.

What actually changes

None of this requires a streaming-platform budget. It requires a shift in what a company decides counts as real information about a customer.

The practical version looks like this: before sending a message, ask whether it’s responding to something the customer told you once, a long time ago, on a form, or something the customer did recently, on their own, without being asked. The first kind of message is filing. The second kind is listening. Most personalization budgets are spent making the first kind of message look more polished, better subject lines, sharper segmentation, nicer templates, when the actual gap is that it’s answering a question nobody asked anymore.

A customer who searched the same product category three times this week has told a company more about what they want right now than a form they filled out two years ago ever could. The company just has to be willing to treat that as the more important piece of information, which requires admitting that everything collected before it might already be out of date.

The thing worth sitting with

Netflix doesn’t know you. It has just been paying attention longer, and more consistently, than your bank, your favorite retailer, or the software company that emails you every Tuesday.

That’s not a technology gap most companies can’t close. It’s a decision most companies haven’t made: whether the version of you they act on is the one from the signup form, or the one from this week.

One of those versions is still accurate. The other one is just easier to file.

Hey, hey! I’m Manya. Marketer by trade, writer by obsession. I share stories, swipe insights, and rant (lovingly) about freelancing, tech, and the business of words. Just hit 22k on LinkedIn , let’s connect!

Want to write with us? Simple! Follow The Hub, submit your draft, and let’s tell your story

[embed]Write For The Hub Publication: Submission Requirements We amplify your bold voice and deliver your inspiring stories to our curious and hungry readers.thehubpublication.com