Open Safari on your phone. Open Chrome on your laptop. Visit the same website. You are now browsing two structurally different versions of the web, and nothing on screen will tell you so.
Safari has blocked third-party cookies by default since 2020. Firefox does the same through Enhanced Tracking Protection. Chrome does not. Which privacy architecture governs your session depends entirely on which browser icon you tapped. Most people made that choice by habit or device default, with no sense that it was an architectural decision at all.
The numbers look lopsided at first. Chrome holds roughly 65% of global browser traffic. Safari and Firefox together account for around 20%. But geography reshapes the math considerably. In the United States, Chrome drops to about 49% and Safari rises to 32%. On mobile, where Safari ships as the default on every iPhone, the two regimes approach parity. For any business targeting American consumers, both architectures matter roughly equally.
The split exists because the original cookie mechanism was never replaced. Safari and Firefox decided the tracking consequences were unacceptable and stripped the capability. Chrome kept it. Two systems now run simultaneously, with no convergence on any roadmap.
The same campaign, targeting the same audience, runs on fundamentally different logic depending on which browser loads the page. Chrome traffic still supports behavioral targeting through third-party cookies. Safari and Firefox traffic requires alternatives: contextual targeting, first-party data, server-side tracking, modeled conversions. Research from ad verification firms shows contextual ads performing within 5–8% of behavioral targeting on click-through rates. Close enough to work. Different enough to require its own infrastructure.
The same user, visiting the same page, is worth different amounts to publishers depending on their browser. Some estimates put advertising on cookie-restricted browsers at 50–60% less revenue than on Chrome with cookies enabled. Same person, same page. The difference is entirely architectural.
Attribution models built on cookie-based tracking systematically undercount conversions from Safari and Firefox. Those users still convert, but the measurement system loses sight of them entirely.
Marketing teams compensate by layering statistical models, server-side tracking, and data clean rooms on top of the gaps. Every layer adds cost and complexity. Every layer exists because the underlying architecture split and nobody stitched it back together.
What keeps this durable is that nobody needs to agree on a resolution. Safari can keep blocking. Chrome can keep allowing. Both browsers work fine for users. The fracture is absorbed entirely by the infrastructure behind the page, by the teams building parallel systems for two versions of the same web. The cookie persisted long enough to split the web along its seam, and the seam shows no sign of closing.