# Larry Page

> 1973– · Computer Scientist, Co-founder of Google
>
> **Recorded contribution:** Co-founded Google; PageRank algorithm

## How to use this dossier

Read for a causal chain, not a hero story: inherited problem → contribution → mechanism → downstream capability → limit. Then close the page and complete the reconstruction exercise from memory.

## 1. Historical orientation

While a Stanford doctoral student, Larry Page investigated the Web's link graph and, with Sergey Brin, developed the PageRank method and a search prototype that became Google. Page later served as Google's founding CEO, product leader, and Alphabet CEO. His historical role combines an influential ranking idea with the institution that deployed Web search at global scale.

## 2. The problem inherited

The rapidly growing Web contained too many pages for directories and simple keyword matching; search needed a scalable signal of relative authority and relevance.

## 3. The central contribution

Page co-developed PageRank, treating links as weighted endorsements in a recursive graph computation, and co-founded Google to integrate ranking with crawling, indexing, and serving.

## 4. Reconstruct the mechanism

1. Crawl pages and construct a directed graph from hyperlinks.
2. Distribute each page's current rank across its outgoing links.
3. Add a random-jump probability so the process remains well-behaved and does not become trapped.
4. Iterate toward a stable score and combine it with query-specific signals for retrieval.

## 5. What changed downstream

- Link analysis improved early Web search quality and helped Google grow into critical information infrastructure.
- Search ranking became a powerful mediator of visibility, advertising, knowledge access, and market power.

## 6. Attribution, limits, and uncertainty

- PageRank was co-authored with Brin and built on citation analysis and prior graph-ranking ideas; Google search has always used many additional signals and teams.
- Links can be manipulated, authority can reproduce existing power, and ranking objectives embed commercial and social choices.

## 7. Reconstruction lab

Calculate PageRank by hand on a four-page graph for five iterations. Add one link farm node and quantify how the ranking changes before and after damping. Compute one iteration without damping and inspect any rank sinks or cycles. Then vary the damping factor and state the behavioral assumption represented by a random jump. Separate the mathematical signal from the complete search system: crawling determines what can be seen, indexing determines what can be retrieved, and ranking orders candidates. Finally, list two relevance or fairness questions that a link graph alone cannot answer. Use labeled relevance judgments to test the rankings, making explicit that an elegant graph equation still needs an independently defined measurement target.

## 8. Evidence trail

- [The Anatomy of a Large-Scale Hypertextual Web Search Engine](http://infolab.stanford.edu/~backrub/google.html) — Stanford University
- [The PageRank Citation Ranking](http://ilpubs.stanford.edu:8090/422/1/1999-66.pdf) — Stanford InfoLab
- [Larry Page](https://abc.xyz/investor/founders-letters/) — Alphabet

---

*Research checked 2026-08-09. Dates, roles, and claims about living people are historical snapshots. Linked sources remain the authority; this dossier is original instructional synthesis.*