Google Search: A $1 Trillion Exaptation
How Google's founders expanded the function of the Web's links from navigation to search, without setting out to build a search engine. EGS Case Study.
By Shripal Shah, DBA — Founder, Exaptation AI
Google's founding success is based on an exaptation; however, this analysis is not reflected in strategic management and innovation science literature. As Larry Page later admitted, “Amazingly, I had no thought of building a search engine.”
In this case, the resource was the weblinks between the pages of the World Wide Web, a resource Google did not own. Their original function was navigation. Page and Brin expanded it to search.
In biology, an exaptation is a trait that evolved for one purpose and later took on a different one. In business, it is the moment an existing asset or resource starts doing a valuable job it was never designed to do.
EGS Case Study No. 2 · Google Search · v1.0exaptationai.com
The same links, read two ways
The Web’s links were designed for navigation. At Stanford, Larry Page and Sergey Brin found a second function in the same links: ranking pages for search.
First functionCERN, 1989
Navigation
Designed at CERN: a link takes a reader from one page to the next, and on to the next.
Keep just the links.
Weigh each link
like a citation.
Second functionStanford, 1996
Search
Found at Stanford: counted like citations, the same links show which pages matter most.
Canonical case, 1989–2004
- 1989Web proposed at CERN
- 1991First public website
- 1995Page and Brin meet at Stanford
- 1996BackRub crawler, first search tests
- 1998Google Inc. incorporated
- 2001PageRank patent granted
- 2004Google IPO
Value captured
$1.27 trillion
Google Search & other revenue, 2017–2025, per Alphabet’s annual reports
Framing limits: shows the canonical EGS case (1989–2004), in which links gained a second function without losing the first. Later Google products are outside the case, and not all of Google’s success is attributable to it. Networks are illustrative; right-panel page sizes are PageRank scores of the sample links. Sources: Shah (2024), EGS Case 4; Brin & Page (1998); Alphabet annual reports, 2017–2025.
Case Context: “What If We Downloaded the Contents of the Web, and Kept Just the Links?”
The World Wide Web was initially designed to solve a more contained, proximal problem. In 1989, Tim Berners-Lee was a fellow at the European Organization for Nuclear Research (CERN) in Geneva, responsible for developing an information management system to retain essential data that lived in distributed, discrete places, making it difficult to access and share across research groups. Berners-Lee's invention proposed the protocol logic for the Web and created the world's first webpage.
When the initial Web protocols were designed, their function was to host, retrieve, and share information in text format via webpages. As websites exploded in number from 10 in 1992 to 23,500 in 1995, users navigated the Web using site addresses called URLs and hyperlinks. The initial function of the hyperlinks was navigation from one webpage to another through clicks. Without the links, webpages would be like islands separated by sea, reachable only by typing in each full URL.
By 1996, portals and search engines like Yahoo, Excite, AltaVista, Lycos, and Ask Jeeves allowed over 75 million users to search a Web that had grown to over a quarter million websites. Web search was an existing and growing need. However, the results were suboptimal because they focused on keywords, the text on the page, the title of the URL, and other such factors.
Silicon Valley was the place to be for the fast-growing Web ecosystem, and Stanford University was well situated to find advisors and collaborators open to innovative ideas, explorations, and experiments. That is where the links found a second function.
Case Process: From BackRub to PageRank
Larry Page was a Ph.D. recruit at Stanford when he met Sergey Brin during orientation week in 1995. Page decided to study the Web for his doctoral research, with particular interest in its network structure because it could be studied as a network graph: a map of pages and the links connecting them.
Page posed a simple question to begin the research: What if we could download the entire contents of the Web, and keep just the links? He had observed that you could see which pages linked to which other pages, but no one could determine who was linking back to each page. Imagine if modern social media platforms were designed like this: unidirectional linking would be akin to only seeing whom you are following but not knowing who is following you. Page intuited that collecting these backlinks for each webpage could generate new value, but no one knew how it would happen.
They built a crawler, a computer program that could index the links by starting with one webpage and following all the links to other pages. This research project was called BackRub, as it could help catalog all the backlinks for a webpage. They started in March 1996 with Page's webpage on Stanford's Computer Science department website, and within a few months, the crawler had indexed 24 million unique URLs and nearly 100 million links between them.
The most essential logical breakthrough came from a different but known academic practice. Page's father was a computer science professor at Michigan State University, and the notion of citations is the connective thread of academic research and publishing. In academic publishing, for instance, the higher the number and credibility of citations, the higher the quality of the paper. To Page, the Web was loosely based on the same premise, and a link was, in effect, a citation. Page and Brin thought they could borrow this quality assessment method from the academic domain and apply it to their research. In their case, the higher the number, quality, credibility, and relevance of the webpages linking to your page, the higher the quality of your webpage. In the words of Brin and Page, “Important pages tend to link to important pages.”
When Page and Brin applied the citation methodology to the indexed Web data on BackRub, they noticed much better search results for any query they entered. They had no initial intention of creating a Web search tool. Improvements in search quality led them to test the first version in Fall 1996.
If we imagine Google as a car company, then BackRub was like an intentionally open-ended research project that unexpectedly developed an innovative engine. The PageRank algorithm was like the resulting engine, a step-change improvement from competing engines in the market. Google Search was the consumer-facing car that delivered unexpected new economic value. But astute observers can see a flaw even in this simple analogy: Engineers doing R&D to build an innovative engine already know what they are aiming for. Meanwhile, Page & Brin didn’t have the specific end-goal in mind; it emerged as the project kept going. That is a distinction with a difference.
Case Results: Google's Founding Success Is Based on an Exaptation
After Page and Brin offered the trial within the Stanford community, word spread quickly. The drastic improvement in Web search technology caught the attention of other faculty members, angel investors, and venture capital firms close to the university. The results were so much better than competing products that they considered offering a license to their algorithm to companies like Yahoo and Excite. After receiving some initial capital injection, Google Inc. was officially incorporated in 1998 to graduate from the university lab to a garage-based startup.
It took a long time for Page and Brin to progress from the spark of an idea in 1995, to the creation of a crawler that indexed the Web in 1996, to improving the quality of the search results in 1997, to the formation of Google in 1998, to being granted the US patent in 2001, and ultimately to an IPO in 2004. At the time, few people thought that search would be a significant entry point for people's everyday experiences with the Web, since browsers and portals were the dominant ways to surf.
The initial search service was free. Google Search has generated more than $1 trillion in revenue since the company attached the search function to the advertising business model. Alphabet's annual reports, which break out the figure from 2017 onward, show about $1.27 trillion in Google Search & other revenue from 2017 through 2025, including $224.5 billion in 2025.
Despite not intentionally designing for it, Google expanded the function of the weblinks resource, from navigating the new digital superhighway to assisting in the search for quality websites.
| Google Search: the case at a glance | |
|---|---|
| The resource | Weblinks: the hyperlinks between webpages, part of the Web protocols Tim Berners-Lee designed at CERN (1989) |
| First function | Navigation: moving from webpage A to webpage B with a click |
| The conditions | A Web more than doubling every year; keyword-based search with weak results; curiosity, talent, and capital around Stanford |
| The question | “What if we could download the entire contents of the Web, and keep just the links?” |
| The search | BackRub (1996): a crawler that indexed ~24M URLs and ~100M links, plus a ranking method borrowed from academic citations |
| Second function | Web search: ranking pages by the number and quality of the links pointing to them (PageRank) |
| The domain | Unchanged: the Web, for both functions |
| The result | Google Inc. (1998), IPO (2004); ~$1.27T in Google Search & other revenue, 2017–2025 |
What the Google Search Case Demonstrates
1. Resource Access v Ownership: The resource does not have to be yours. The people who designed the Web were not involved in the specific case of Google Search. Newcomers became active as the Web grew and found a new function in a resource they did not own.
2. Fluid Mindset: A simple question can start the search. “What if we downloaded the contents of the Web, and kept just the links?” was the critical thought that led to initial, open-ended experimental research.
3. Exploratory Search & Discovery: Open-ended experiments can find what others miss. None of the existing search firms had explored backlinks. Uber searched with a specific mandate; Page and Brin explored new ideas and experimented without specific end goals. Both found an exaptation.
4. Innovative Insight from Another Domain: The breakthrough was borrowed. Beyond mathematical wizardry, Page and Brin borrowed a preexisting method of determining relevance from a distant domain (academic literature citation method) and brought it into the new digital world. It took the cognitive flexibility to reach across disciplines to find useful practices.
5. Scaling Markets are Opportunities: A new function can emerge in the same market. The application domain, the arena where the resource is used, remained the same with both the navigation and search functions: it was the fast-growing Web. It is a reminder that a function is defined by what the resource is doing, which can be distinct from where it is applied.
6. Design Follows Quickly: The first function stays, and the value comes later. The original navigation function was not lost as weblinks began serving the new search function; a second function was added. After it emerged in their lab, the early Google team adaptively fine-tuned the new capability to meet user needs. The advertising business model came later.
What Else Is Hiding Within Massive, In-Use Resources?
At Google I/O in May 2026, the company said it has 13 products with more than a billion users each, five of them (Search, Gmail, Android, Chrome, and YouTube) with more than three billion. While later products like Google AdWords, Google Maps, Android, and Gmail are not part of this case, we can reasonably state that they would not exist in their current form without the globally scaled search product and its monetizing power. However, we should be cautious not to attribute all of Google's success to the original search innovation. That would create an attribution error.
More crucially, how many hidden exaptations lurk within this massive, global resource? And could the firm intentionally search for future exaptations?
The same question applies to what Google has retired. The open-source Killed by Google list records 300+ discontinued Google products, services, and devices. For EGS research, it is an unusually well-documented inventory of resources that were built, used, and then set aside (aka legacy, sunset and stranded assets). We are beginning to study it, and we welcome readers who worked on those products, or who want to help research them, to get in touch at shripal@exaptationai.com.
Page and Brin built a search engine for the Web's information. At Exaptation AI, we are building the search engine for every asset's undiscovered functions. Growth, in our view, is a search problem.
Next in the series: Nvidia’s 20-year journey to becoming world’s largest company.
Sources
Alphabet Inc. (2026). 2025 Annual Report. Revenue by type, 2023–2025. Earlier years: Form 10-K, fiscal 2019, and subsequent annual reports.
Battelle, J. (2005). The Search: How Google and Its Rivals Rewrote the Rules of Business and Transformed Our Culture. Portfolio.
Berners-Lee, T. (1989). Information Management: A Proposal. CERN.
Brin, S., & Page, L. (1998). The anatomy of a large-scale hypertextual Web search engine. Computer Networks and ISDN Systems, 30(1–7), 107–117.
Google (2026, May 19). 13 products with more than 1 billion users [Post on X], Google I/O 2026. See also Neowin coverage.
Internet Live Stats. Total number of websites.
Isaacson, W. (2015). The Innovators: How a Group of Hackers, Geniuses, and Geeks Created the Digital Revolution. Simon & Schuster.
Johnson, S. (2011). Where Good Ideas Come From: The Natural History of Innovation. Penguin.
Levy, S. (2011). In the Plex: How Google Thinks, Works, and Shapes Our Lives. Simon & Schuster.
McPherson, S. S. (2011). Sergey Brin and Larry Page: Founders of Google. Twenty-First Century Books.
Ogden, C., and contributors. Killed by Google.
Shah, S. (2024). Exaptation Growth Strategy. Doctor of Business Administration dissertation, Pepperdine Graziadio Business School. EGS Case 4: Google — Navigation to Search.