Building an Email Program That Survives the Privacy Decade

Privacy, personalization and AI-readiness used to be three separate issues. Now, they're interlinked.

About 20 U.S. states now enforce comprehensive privacy laws, cookie-based retargeting keeps getting harder to rely on and AI personalization is only as good as the data feeding it. The brands pulling ahead aren’t winning because they have the biggest legal teams or ad budgets. They’re successful because they ask customers directly what they want, then use that consented data to fuel their AI. 

Here’s how to build that exact type of system for your brand.

Why Email Became the Privacy-Ready Channel

Privacy law is now the baseline. About 20 states have comprehensive consumer privacy laws in effect this year, with Indiana, Kentucky and Rhode Island joining the list on January 1. 

Across all 20, the direction is the same. There are more overlapping rules and real enforcement behind them. For example, California's privacy agency has landed multi-million-dollar settlements this year, and coordinated enforcement sweeps in California, Colorado and Connecticut show regulators now see privacy as a necessary operational function.

In plain English, this means companies can’t treat privacy like a box-checking exercise anymore, because it now affects how they collect data, run marketing and avoid costly penalties.

Retargeting is quietly weaker, but not for the reason you think. Contrary to the "cookies are dead" narrative, Google reversed its plan to deprecate third-party cookies in Chrome and retired most of its Privacy Sandbox replacements. While Chrome still allows them, the real erosion is coming from elsewhere: 

• Safari and Firefox block third-party cookies by default. 

• Apple’s App Tracking Transparency has made tracking much harder for marketers, with most users choosing not to opt in.

• And roughly 33% of internet users run ad blockers. 

Together, most advertisers are working with an addressable retargeting pool that's shrunk somewhere between 30–60% since 2022. And that’s a hole email and SMS never had, because your list was never dependent on a cookie in the first place.

AI needs real signals. 75% of marketers now use generative AI in at least one recurring workflow, per Salesforce's 2026 State of Marketing report. Add in the early moves toward agentic commerce—where AI shopping agents discover and compare products on a customer's behalf—and both AI personalization engines and shopping agents run on structured, accurate, consented data.

All of these forces point to the same fix: the data your customers actually gave you permission to use.

Consented Data Is an Appreciating Asset. And It’s Fuel for Your AI.

Third-party data depreciates. It decays, restricted by new laws and browsers are steadily cutting off access to it. 

Zero-party data, which customers voluntarily share directly with a brand, does the opposite. It compounds and it’s consented, so it survives the next law change. It's also more accurate, because the customer declared it instead of you inferring it. Plus, no competitor can buy their way into it.

Progressive profiling is how you build that asset without friction. Instead of one long signup form, ask for the minimum up front—usually just an email address—then layer in preferences over time, triggered by real behaviors, such as a second visit, a cart abandonment or a first purchase. 

For example, a fashion brand might ask about style and size after the first order. A home goods brand can ask what room or project someone's shopping for. Brands can keep score on what drives individuals to engage and purchase, by tracking customer ID across POS, email, SMS and loyalty. Brands can also use email to ask questions that visibly improve the customer's experience, later using it for content personalization.

Make the most of your preference center. This is where customers tell you how often they want to hear from you, what channels they prefer and what they care about. That information should immediately shape your segmentation and email content. Brands that actually use their zero-party data often see stronger conversion performance, because they can personalize more precisely.

Here's where it gets interesting for your AI strategy. A lot of email and CRM platforms are now using reinforcement learning, which means they don’t rely on fixed rules anymore. They test different options for each customer—like an offer, channel, send time or creative variant—then learn from what works. If a click, purchase or re-engagement happens, the system treats that as a positive signal and gets smarter over time. The goal is to figure out the best next action for each person, not just for each segment.

The catch is that a reinforcement learning system is only as good as the signal it starts with. Feed it noisy, inferred data and it has to guess its way toward a decision, sometimes optimizing for the wrong thing entirely. 

Feed it zero-party data and it’s no longer guessing. Braze itself describes AI decisioning as dependent on solid data foundations and a clear reward signal to optimize toward. That signal starts with the preference-center and progressive-profiling work above. 

You can see this compound in email, specifically. Perry Ellis went AI-first on campaign creative with Backstroke and saw a 17.5X return, with predictive creative beating one-size-fits-all sends by up to 45% in revenue per recipient. Thirdlove tested predictive templates across 1.6 million subscribers and saw revenue per recipient rise 25%. Cozy Earth's agentic AI workflow personalized campaigns in under five minutes and lifted click rates 48%. None of that happens without clean signal feeding the model first.

Every preference toggle, newsletter click and quiz answer you collect is training data for your personalization engine.

Consent → Personalization → Agent-Readiness

Consented data is what makes real personalization possible. McKinsey found that effective personalization can cut acquisition costs by up to 50%, lift revenue by up to 15% and improve marketing ROI up to 30%, with fast-growing companies generating about 40% more revenue from personalization.

Done poorly though, personalization backfires. Gartner found 53% of customers who experienced "creepy" personalization reported a negative reaction, making them over 3X more likely to regret a purchase and 44% less likely to buy again. Declared preferences lower your risk of coming off as creepy, because you're acting on exactly what a customer told you.

This also sets you up for what’s next with AI shopping assistants. If your product data is clear, complete and accurate, those assistants are more likely to find your brand and recommend it to shoppers. And since most of these tools still send people back to the brand shop to finish the purchase, you keep the customer relationship, including their login, loyalty info and consented profile. In other words, the same zero-party data that helps you personalize today also helps keep you visible when AI agents start doing more of the shopping work.

A Simple 90-Day Plan to Tap Into Your Customer Data

Here's a quick checklist to help you make customer data more useful right away:

1. Start by launching or updating your preference center this quarter so customers can set frequency, channel and topical preferences. 

2. Then add two or three low-friction profile questions across your welcome series, post-purchase flow and birthday emails. 

3. Make sure all customer data rolls up to one ID, with clear consent tracked by channel. 

4. Be sure to use the preferences people share right away in segmentation and dynamic content, instead of letting them sit in a database. 

5. And if you’re testing an AI decisioning tool, decide upfront what success looks like. Track the metrics that matter most to you, like revenue or retention, so your AI tools can learn from clean zero-party data from the start.

The privacy decade rewards brands that truly earn their audiences’ data and interests. Consent is the asset. Your data is the compounding interest. And AI brings everything together, orchestrating a predictive email program that delivers winning creative before you hit send.