Marketers Have The Data For Better Personalization. They Don't Have The Time, The Team, Or The Trust To Act On It.

Marketing teams love to talk about their 360-degree customer views. They've consolidated data, built out CDPs, and invested heavily in first-party data infrastructure. And yet, in the same breath, many of these organizations are still blasting five-year loyalists with introductory trial texts. When personalization breaks down at that level, the instinct is to blame the data. But the data usually isn't the problem.

The real constraint is far more human: marketing teams lack the capacity to act on what they already know, and their organizations remain deeply hesitant to let autonomous systems operate on their behalf. "If you don't trust it to work autonomously, that's a trust problem, not a data problem," says Eric Miao, Chief Strategy Officer at Attentive. Miao has spent nearly a decade at the company, rising from VP to CSO, and previously held roles in mobile advertising at Twitter. In a previous conversation, he outlined how Attentive is building technology designed to act on a marketer's behalf. Here, he turns to the organizational and operational barriers that keep most brands from using the tools they already have.

Lift the floor before chasing the ceiling

The industry's idealized version of personalization involves crafting unique, perfectly timed messages for every individual customer. That's a worthy long-term goal, but Miao argues that chasing it prematurely misses the opportunity sitting right in front of most teams. Instead of pursuing perfect one-to-one messaging for cold prospects, a growing number of brands are investing in continuous, autonomous testing loops that raise the quality of every journey across the board.

The math isn't complicated. Perfect personalization requires deep customer knowledge, and for the majority of any brand's subscriber base, that depth simply doesn't exist. "Is the most optimized baseline journey as good as literal one-to-one messages made up for each person from scratch? Maybe not for certain people, but for most people they're actually the same," Miao says. "Because for most people you don't know that much information about them. You can't actually come up with something sufficiently amazing for a person who's a stranger to you."

Rather than building elaborate segments from scratch, Miao advocates for a simpler path: start with a single journey, split it on one attribute, test the split, then split again. Purchaser vs. non-purchaser. One-time buyer vs. recurring subscriber. Recent signup vs. dormant. Each iteration lifts the floor. "You can A/B test your journey continuously, and an agent can do that right now. All of the tools exist, the MCP servers exist, the computer use exists. It's all right there." The approach mirrors what well-staffed agencies have always done manually, except that agentic systems can run those loops indefinitely without burning out or billing by the hour.

The constraint on better personalization today is marketer time, not a lack of data. Most brands don't have a team large enough to build and test the journeys the data would support, and that gap is what the technology is meant to fill. "It's not a data problem. You have the data right now to have an insanely better baseline," Miao says. "You don't have the agent to do it for you. You don't have the marketing team. If you had the marketing team, you wouldn't need the agent."

The compounding problem

Most marketing budgets fund production. Briefs get written, emails get designed, campaigns get scheduled, and the cycle repeats. The work creates output but generates no compounding value. Miao points to one major retailer that employs a 20-person in-house email design team whose sole function is to follow briefs and build emails. The cost runs into several million dollars annually. The team doesn't make the emails better over time. It just makes more of them.

"People have been spending their money and their time on things that are required but do nothing to make the marketing actually get better over time," Miao says. "There is nothing compounding about a 20-person email design team. It doesn't get better. It just keeps churning out emails." As content production costs collapse, Miao argues that the savings should be redirected toward experimentation, testing, and strategic reinvestment rather than returned to shareholders.

Lower production costs and lower testing costs are linked. "The easier it is to do the task, the more people will be able to test stuff. The more you can test stuff, the better it's going to get. The more personalized it's going to get." The path to genuine personalization runs through operational efficiency first, data science second.

The trust gap and the identity shift

Automating tedious, low-leverage work is the easy part. The harder transition is psychological. Many marketers built their careers on hands-on execution, and asking them to step back from that work and into a strategic editorial role requires a genuine identity shift.

The lowest-risk entry point for autonomous workflows is the work marketers already resent. Miao sees teams connecting AI to their CRMs and automating weekly reporting: downloading data, reformatting it, copying it across platforms, and surfacing summaries. "In that job, zero leverage came from running reporting," says Miao. "Marketers are creative people. It's not the thing they're the best at, and they don't get energized from it." Reporting is where trust gets built, because the stakes are low and the upside is immediate.

The uncomfortable next step is acknowledging that the tasks marketers enjoy doing aren't necessarily the ones they do best. "If you thought it was fun to code together emails, that part changes," Miao says. "If you thought it was fun to come up with a test plan yourself by looking at all of the profiles and trying to learn about them, that changes too." He frames this as a universal experience, noting that he's had to go through the same reckoning himself. The marketer who resists that evolution becomes the bottleneck.

The organizational challenge also extends beyond individual contributors. Any system involving many people is limited by the number of handoffs between them, and marketing teams tend to be deeply interconnected with inventory, e-commerce, acquisition, and SEO. Miao argues that leaders need to audit every approval chain and feedback loop that slows decisions. "If the CEO looks at all of the emails and texts and has to approve them, that's the kind of thing you'll want to change," he says. "Otherwise it's hard to get more productive."

Product, brand, speed: what matters when everyone can personalize

As AI commoditizes execution, the question shifts from who can personalize fastest to who has something worth personalizing for. Miao sees competitive advantage returning to timeless fundamentals: product quality, brand conviction, and organizational speed.

When categories converge on similar products, the winners are the organizations that move into new opportunities faster and learn from the market sooner. Miao points to athletic apparel as an example: Adidas, Nike, Puma, and Lululemon all produce similar products, but customers choose based on how effectively a brand captures the cultural moment. "Why do people buy one thing and not the other? It actually is better somehow, or it captures the zeitgeist, which is a marketing thing, or it's the brand," he says. "The question is, how did the brand decide to get into that category, and were they able to get into it quickly? It's a speed thing."

AI won't replace brand storytelling. Miao notes that major brands rely on human intuition and rigorous testing to protect their identity rather than trusting automated creative. "Coca-Cola isn't running AI-generated commercials that no one's ever watched. They test this stuff carefully." In a world where execution gets cheaper, brand becomes the scarce asset that determines whether a customer chooses you over a functionally identical competitor.

After spending an entire conversation advocating for AI-driven execution, Miao acknowledges a natural tension at the top of the market. Elite brands thrive on cohesion and aspiration, and aggressive personalization can undermine both. "You could make an argument that personalization and brand are achieving opposite objectives" he says. "I don't think Hermès wants to have one-to-one personalization for everything. When they are talking to the average person, they want everyone to be imagining the same Hermès. It's only once you're in the store that they cater to you one to one. So there's real tension." The brands that navigate this tradeoff most intentionally, knowing where to let agents run and where to preserve a singular, human-driven voice, will define the next era of marketing.