Everything Old Is New Again… What Direct Marketers Always Knew

Much of today’s conversation about first-party data feels urgent, even frantic. Brands and agencies, as Digiday recently reported (https://digiday.com/marketing/inside-the-brand-and-agency-scramble-for-first-party-data-in-the-ai-era/), are scrambling to pull customer data out of platforms and into environments they control, often motivated by AI, privacy concerns, and a growing unease with over-reliance on walled gardens.

But for anyone with a direct marketing background — especially those of us who cut our teeth in catalogue, CRM, or direct response — this moment feels less like a revelation and more like a rediscovery.

Because the core truth was always obvious:

Walled gardens optimize for acquisition. First-party data exists to drive retention and growth first, acquisition second.

And direct marketers have known that for decades.

Long before programmatic media, identity graphs, and AI copilots, direct marketers worked from a simple but powerful asset: the customer file. Names, addresses, purchase history, recency, frequency, monetary value. Yes, there were ongoing acquisition efforts, but long-term growth and profitability revolved around understanding who had already raised their hand and how to increase their long-term value. (Bob Kestnbaum would be very unhappy if I said "lifetime value".)

Catalogue marketers didn’t ask, “How do I reach the most people?” because they were budget constrained. They asked, “Who is most likely to buy again — and what should I send them next?” and “Where can I find more like them?”

That mindset naturally prioritized:

• Relevance over reach

• Lifetime value over initial conversion

• Profitability over media efficiency

In other words, retention and growth were as important as acquisition.

So why did it take the rest of the marketing world so long to catch up?

The answer lies in what digital platforms made possible — and what they quietly deprioritized.

Walled gardens like Google and Meta were built to solve an enormous problem: customer acquisition at global scale. They abstracted away identity, simplified targeting, and made it easy to buy demand quickly. For a generation of marketers, success became synonymous with clicks, conversions, and CPM efficiency.

But that convenience came at a cost. The customer relationship was mediated by platforms. Signals were aggregated. Feedback loops were short. And most importantly, long-term customer value became someone else’s problem.

As long as acquisition was cheap and measurable, few questioned the tradeoff.  And spending to “acquire” the same customer, over and over, was lost in lower CACs.

Digiday’s reporting highlights how AI has accelerated today’s first-party data scramble. Brands want their data ingested into cloud environments they control, where it can train AI models securely and meaningfully. Agencies are being pulled closer to data strategy, governance, and interpretation — not just media execution.

This isn’t because AI invented the need for first-party data. It’s because AI exposes the limits of acquisition-only thinking.

AI systems trained on platform signals get better at buying ads. AI systems trained on first-party data get better at understanding customers.

That distinction matters. Retention, churn prediction, cross-sell, and lifetime value modeling all depend on longitudinal data — the kind that only exists inside a brand’s own ecosystem. These are direct marketing problems, even if we’ve stopped calling them that.

What’s striking about the current moment is how familiar it feels to anyone with a direct response background. The industry is rediscovering ideas that once felt foundational:

• Not all customers are equal

• Past behavior is (generally) the best predictor of future behavior

• Growth comes from deepening relationships, not endlessly replacing them

The difference now is scale, tooling, and computational power. AI allows brands to apply direct marketing principles across millions of customers in real time — but only if the underlying data strategy supports it.

The lesson isn’t that walled gardens are bad. Acquisition will always matter. But it can’t be the only strategy.

As brands rethink data ownership in the AI era, a few principles stand out:

1. Treat first-party data as a growth engine, not a fallback plan.

2. Measure success in customer value, not just media efficiency.

3. Build systems — and partnerships — that prioritize learning over targeting.

The irony is that none of this is new. The tools have changed. The scale has changed. The stakes have changed. But the strategy hasn’t.

The future of marketing may be powered by AI — but its foundation looks a lot like direct marketing, rediscovered.