The Renaissance of Marketing

Every schoolbook gets the Renaissance backwards. It was not a flowering of genius. It was the first copy crisis, and nearly every lasting invention it produced was a marketing invention.

Before 1450, a book was a scribe’s year. By 1500, Europe had printed roughly 20 million volumes (the standard estimate from Febvre and Martin). Making had become cheap. Within two generations the complaint we treat as modern was already on the record. In 1545 Conrad Gessner opened his catalogue of all known books by protesting the “confusing and harmful abundance of books” and asked kings and princes to do something about it. The first documented information overload is 481 years old.

What followed the flood is the part worth studying. Manuscripts had no title pages. The title page was invented for printed books as an advertisement. The printer’s mark (Aldus Manutius’s dolphin and anchor in Venice) was the first publisher brand. Ann Blair’s work on the period shows that the sense of too many books produced new genres built to sort them: the universal bibliography, the book review, the index. The abundance of making created the profession of being found.

Then the lawsuit. Around 1506 a Bolognese engraver named Marcantonio Raimondi bought a set of Dürer’s Life of the Virgin woodcuts in Venice and engraved copies so faithful they were sold as Dürer’s, complete with his AD monogram. Dürer went to Venice and complained to the Senate. The ruling is the whole essay in one line. Raimondi could keep copying the images; the compositions belonged to everyone. He could not use the monogram. The court would not protect the work. It protected the mark.

That was the verdict in 1506. It is the verdict in 2026.

A good product sells itself. You have heard this from a founder or a CEO, and if you work in marketing you have been budgeted according to it. It was never true. Not now, and not in the scarcity era it was supposedly true in.

The United States Patent Office has granted more than 4,400 mousetrap patents, the most of any device in American history. Fewer than two dozen ever earned their inventor a cent. The trap that wins is a snap trap from 1899, sold under the Victor name, and it outsells every other American mousetrap combined. Thousands of people built a better mousetrap. The world did not beat a path to anyone’s door.

Peter Thiel made the same point in Zero to One without the mice: superior distribution can build a monopoly with no product advantage at all, and the reverse is not true. Engineers underrate this because distribution looks like noise to people who build.

So why did the myth survive? Because for 30 years, building was the bottleneck. A proper product took capital, engineers and time, so capital flowed to builders and marketing was the department you cut first. I watched it happen in telecom, gaming and fintech. To a tiny extent the myth held: when a category had three products, the best one could win on word of mouth. Count the three-product categories you can find today.

The numbers moved faster than most leadership teams have noticed. Google said 25 percent of its new code was AI-generated in October 2024, over 30 percent by early 2025, and by April 2026 its CEO put it at three quarters. The conservative industry measurement, across 4.2 million developers, puts AI-authored production code at about 27 percent in early 2026. Y Combinator reported in 2025 that a quarter of its startup batch had codebases that were about 95 percent AI-written.

Read those as one sentence: if you know what you want built, you can have it built. The skill moved from writing code to specifying it. Anyone can copy a product now, and anyone who can describe one can ship one.

And because everyone is specifying into the same three or four models, the outputs converge. I have ranted a lot about the coral-highlighted dark landing page that every founder shipped this year. The point here is bigger than design. Software, copy, decks, pages, product screens: the floor of quality jumped, the ceiling did not, and a crowded market became a market of near-identical credible options.

In November 2024, for the first time, more new articles on the open web were generated by AI than written by humans (50.3 percent, according to Graphite’s analysis of Common Crawl). Graphite estimates more than 10 billion AI-generated pages have been published since 2023. Merriam-Webster made “slop” its 2025 word of the year.

Now the other half of the data. 86 percent of articles ranking in Google Search are human-written. 82 percent of articles cited by ChatGPT and Perplexity are human-written. When AI articles do appear, they rank lower. And the share of AI articles has plateaued since mid 2024, as publishers learned that volume does not buy visibility.

That is the Renaissance curve replaying on a two-year clock. In 1460 a printed book was a marvel. By 1545 it was a nuisance. AI output had its 1460 moment in 2023. It is 1545 now. AI images, AI hooks, AI blogs were a marvel for about eighteen months. Today they are the baseline, and a baseline is worth exactly zero in an auction for attention.

This has happened at least three times since Gutenberg, and the pattern never changes.

In 1985, PageMaker and the LaserWriter made everyone a publisher. The result was the ransom-note effect (flyers in eleven fonts), and professional designers became more valuable, not less.

In 1888, Kodak promised you press the button and they do the rest. Painting did not die. It went where the camera could not follow, into Impressionism and abstraction. Expect deep-pocket brands to do the same: real photography, hand-built campaigns, visible human craft, as a status signal in a synthetic feed.

In 2008, the App Store launched with about 500 apps and anything decent got downloaded. Today it holds close to two million, and discovery is an industry with its own acronym. Nobody in mobile says build it and they will come anymore.

The pattern: a tool makes production free, a flood follows, the default output drops to zero value, and the scarce skill becomes being chosen. A new profession appears every time: the editor, the designer, the app-store optimizer. This time it is not a new profession. It is marketing, getting its job back.

Renaissance means rebirth, so name what died. Marketing had a craft era. In 1959, DDB ran a Volkswagen ad with a tiny car in a field of white space under two words, Think Small, when every other car ad shouted. Ad Age later ranked it the best campaign of the twentieth century. That was marketing as judgement: positioning, voice, the nerve to look different.

Then came the dashboard age. Marketing became bid management, CTR watching, template filling. It got de-skilled, and de-skilled work is the first to be cut. That is the dark age the C-suite myth grew up in.

AI eats the dashboard work. What is left is the old craft, and it is scarce again. Positioning that a model cannot generate because it requires a choice. A voice the weights did not hand you. Distribution deals, unusual channels, the measurement discipline to know whether you are actually distinct rather than assuming it. The medieval scriptorium produced anonymous copyists. The Renaissance produced named masters. AI is sorting us into the same two classes, and the sorting has already started.

Dürer put his mark on everything and crossed the Alps to defend it. That is the job description now.

One number this week. Pull 12 months of your content and campaign output volume, and 12 months of your customer acquisition cost. If output went up and CAC did not come down, you have been paying the slop tax: you added abundance, not advantage. That gap, measured monthly, is the only honest scorecard for whether your marketing is in the dashboard age or the craft age.

AI made the product cheap. It made being chosen expensive.

In God we trust. All others must bring data.


- Ashish