In 1961, a Catholic priest named Father Schmitt sat down at an IBM 650 mainframe and did something that almost nobody in marketing was doing: he computerized his parish's donor records.

The IBM 650 processed dozens of records per hour. It was roughly the size of a refrigerator and cost more than a house. Father Schmitt used it to do something deceptively simple: personalize his outreach. Instead of sending the same appeal letter to every parishioner, he matched the message to the donor — their giving history, their participation patterns, their capacity. The technology was primitive. The principle was timeless.

Personalize the outreach. Match the message to the recipient. Remember what worked.

Someone at a conference once asked me why I keep reaching back to ancient history when talking about the agentic future. It's a fair question. Here's my answer: Father Schmitt's mainframe processed dozens of records per hour. An AI agent evaluates dozens of brands per second. That's six orders of magnitude faster. But the logic — the underlying principle that knowing your audience and personalizing your approach is the key to every transaction — didn't change at all.

This is the thread that runs through every story in this series. The technology changes. The principles compound.

The Sixty-Five-Year Arc

Trace the line from Father Schmitt to today and you see the same idea expressed through successively more powerful technology.

1961: Mainframe processes donor records. Personalize the letter. Father Schmitt's IBM 650.

1995: Database marketing scales direct mail. Build segments. Target households. RFM models predict who'll respond.

2005: Digital advertising creates real-time behavioral signals. Cookies track browsing. DMPs organize audiences. Brands can find individuals, not just segments.

2015: Customer data platforms unify identity. Cross-device stitching. Journey orchestration. The promise of understanding the whole customer.

2025: AI agents act on the customer's behalf. The customer delegates. The intermediary between brand and person is no longer a browser or an inbox — it's an intelligent agent with preferences, evaluation criteria, and negotiating capability.

Each era brought a new technology. Each technology was declared revolutionary — and it was, in the narrow sense that it enabled things that were previously impossible. But the underlying logic never changed. Know your audience. Match the offer. Remember what works. Be relevant.

Father Schmitt would recognize what we're doing. He'd just be amazed at the speed.

Data advantage is temporary. Someone will always have more data, fresher data, better data. But the principles that govern how you use it — those compound across every generation of technology.

What the Agents Don't Change

I've spent nine stories in this series arguing that agents change everything — frameworks, funnels, measurement, discoverability, activation, the fundamental relationship between brands and customers. And I believe all of that.

But I want to end the series by arguing the opposite: what the agents don't change.

They don't change the fact that humans set preferences. Every agent in the world operates on instructions from a person. Those instructions are shaped by experience, narrative, emotion, memory, trust, and all the other irreducibly human factors that stories are built to influence. The agent is a tool. The human is the principal. And the principal's preferences are formed through exactly the same mechanisms that Father Schmitt was leveraging in 1961 — just at a different scale.

They don't change the fact that relevance wins. Father Schmitt's personalized letter outperformed the generic one. Casey's personalized pizza email outperformed the stock photo. Campbell's weather-triggered ad outperformed the always-on campaign. An agent-facing product API that precisely matches the customer's preferences outperforms a generic product listing. The technology is different. The principle is identical: relevance beats volume.

They don't change the fact that trust compounds. The brands people trust are the brands people tell their agents to prefer. Trust isn't built in the agent's evaluation architecture — it's built in the human's experience of the brand over time. Every positive interaction, every kept promise, every moment like the Porsche on the tarmac — they accumulate into a preference that the human encodes into their agent. Brands that build trust over decades have a structural advantage that no amount of agent-optimization can replicate.

They don't change the fact that stories move people. Not agents — people. The CMO who needs to invest eight figures in agent-ready infrastructure won't be moved by a technical spec. They'll be moved by the story of Apex Home Goods doing everything right and still getting blindsided. The PMM who needs to rethink their entire approach won't be motivated by a framework diagram. They'll be motivated by the image of Joe's agent buying a car in six days while their multi-touch attribution model measured a journey that no longer exists.

The Story Before the Agent

Here's the insight I keep coming back to, the one that ties every piece in this series together: the story happens before the agent gets involved.

The customer who tells their agent "I only buy from sustainable brands" got that value somewhere. A documentary they watched. A conversation with a friend. A brand campaign that landed at the right moment and shifted their worldview. That preference — the one the agent now enforces with algorithmic precision — was formed through storytelling.

The customer who tells their agent "always book Delta" formed that loyalty through an experience — maybe a Porsche on the tarmac, maybe a decade of reliable service, maybe just a feeling that Delta gets it. The agent doesn't have that feeling. The agent has an instruction. But the instruction originated in a story.

The customer who tells their agent "I like Heineken" didn't arrive at that preference through a structured comparison of beer attributes. They arrived at it through a hundred small moments — a cold bottle at a summer party, the distinctive green color catching the light, a sense of worldliness that the brand cultivated over decades. The agent might eventually optimize away from that preference if a competitor offers better value on the attributes that matter. But the starting position — the default the agent works from — was set by story.

Brand marketing in the agentic era isn't obsolete. It's elevated. It moves upstream, into the preference-formation layer where humans decide what they value, what they trust, and what they want their agents to do on their behalf. The agent executes. The story initiates.

The Principles Compound

Father Schmitt would be bewildered by the technology. But he'd understand the logic immediately.

You have an audience. Learn about them. Match your outreach to what matters to them. Remember what works and do more of it. Build trust over time. Be relevant. Be human.

The IBM 650 did that at dozens of records per hour. The CDP did it at millions of profiles per second. The AI agent does it at a scale and speed that makes both of those look like cave paintings. But the principles are the same. They've always been the same.

I've told ten stories across this series — a $5 chip, a pizza, toilet paper, weather, playlists, a Porsche, a bullseye, a plateau, a car, and a mainframe. They span sixty-five years and six orders of magnitude of computing power. And the thread that connects all of them is the same thread that Father Schmitt tugged on in 1961: the people you're trying to reach are human, and humans respond to relevance, recognition, and relationship.

The agents are new. The principles are ancient. And the companies that remember that — the ones that build for agents without forgetting they're ultimately serving humans — are the ones that will win the next sixty-five years.

The technology changes. The principles compound. That's the whole story.