This article is a translation of Paul Graham's essay " What happened to Yahoo
When I started working at Yahoo after they bought our startup in 1998, this company seemed to be the center of the world. It seemed to be the next big thing. It seemed like it would be what Google was.
What went wrong? The problems that pulled the company back appeared long ago, in fact from the very foundation of the company. These problems were already clearly visible when I got there in 1998. Yahoo had two problems that Google did not have: easy money and a reluctance to be a technology company.
When I first met Jerry Yang, we had different opinions about why we met. He thought that we met so that he could personally check us before buying. I thought we met so that we could show him our new technology, the Revenue Cycle. It was a way of sorting product search results. Sellers offered a percentage of sales that they were willing to pay for traffic, but the result was not sorted by this percentage, but by the percentage multiplied by the average amount of goods that the user would buy. It was similar to the algorithm that Google now uses to sort advertisements, but this happened in the spring of 1998, when Google was not yet founded.
The profit cycle was the best way to sort the search results for products, in the sense that it sorted in the order in which Yahoo would get the maximum profit from each link. But this was not the only advantage. Ranking search results by user behavior makes searching better. Users improve the search engine: you can start sorting the results based on a simple text match, but over time, when users buy more, the search engine improves every time.
But Jerry didn't care. I was a little confused. I showed him the technology that allowed us to get the maximum profit from search traffic, but he did not care. I could not understand - whether I explained this to him badly, or whether he simply did not want to show his surprise.
I did not know the answer until I started working at Yahoo. None of my assumptions were true. The reason Yahoo didn’t care about the technology that maximized the traffic profit was because advertisers were already overpaying for it. If they simply received the actual cost, they would receive less.
Now it is hard to believe, but at that time a lot of money was made on banner advertising. Advertisers wanted to pay ridiculous amounts for banner advertising. The Yahoo sales team decided to use this source of income. Led by a large and terribly impressive man named Anil Singh, marketers from Yahoo flew into Procter & Gamble’s office and came back with millions of orders to display banners.
Prices seemed low compared to print ads, which advertisers compared with no other comparison. But they were high compared to how much they really cost. These big, dumb companies provided a dangerous source of income. But there was another, more dangerous source: other Internet startups.
In 1998, Yahoo was at the top of the pyramid. Investors are very interested in the Internet. One of the reasons this happened was Yahoo's profit growth. And they invested in Internet startups. Then startups used this money to buy ads on Yahoo, which allowed them to get traffic. This led to even greater revenue for Yahoo, which again convinced investors that it was worth investing in the Internet. When I once realized this, I jumped like Archimedes in my bathroom, only instead of “Eureka!” He cried “Sales!”.
Internet startups and Procter & Gamble advertised their brand. They did not think about the target audience. They just wanted a lot of people to see their ads. Traffic has become a thing that can be obtained from Yahoo. No matter what traffic. [one]
It did not only Yahoo. All search engines did this. That is why they wanted people to call them portals, not search engines. Despite the real meaning of the word portal, they meant that this is a site where users could find what they want on the site itself, instead of switching to other sites, as in search engines.
I remember talking to David Filo in 1998 or early 1999 that Yahoo should buy Google, because I and most of the other programmers in the company used it, not Yahoo, to search. He told me that you should not worry about it. The search was only 6% of our traffic, and we grew by 10% every month. It was hardly worth improving it.
I did not say that search traffic is worth more than all the other. I said, "Oh, good." Because I myself did not understand how much search traffic costs. I'm not even sure that Larry and Sergey understood this. If they understood, Google would hardly have spent the effort on enterprise search.
If things were different, Yahoo executives might understand how important the search is. But from the truth, they were separated by the hardest obstacle - money. While customers were writing solid checks for banner advertising, it was difficult to take the search seriously. Google is not confused.
But Yahoo had another problem that prevented it from changing direction. They were knocked out of balance from the very beginning because of their unwillingness to be a technology company.
When I started working at Yahoo, the strangest thing I learned was that they insisted on what they should be called a media company. If you were in their offices, they looked like a software company. The jobs were filled with programmers who wrote the code, managers who thought about the list of features and completion dates, technical support operators (yes, there were technical support operators), who told users to restart the browser, and so on, as in any IT company. Why did they call themselves a media company?
One of the reasons is the way they earn: selling advertising. In 1995, it was difficult to imagine a technology company that made money that way. Technology companies made money from selling software. Media companies sold ads. So they must be a media company.
Another big factor was the fear of Microsoft. If anyone at Yahoo allowed them to be a technology company, the next thought would be that Microsoft would destroy them.
For those who are much younger than me, it’s hard to imagine what fear Microsoft caused in 1995. Imagine a company several times more powerful than Google, but much more aggressive. Fear of them was absolutely justified. Yahoo saw them destroy the first Internet company, Netscape. It would be logical to assume that if they tried to become like Netscape, they would share their fate. How could they guess that Netscape was the last victim of Miscrosoft?
It was wiser to claim the role of a media company to confuse Microsoft. But unfortunately, Yahoo tried to be a media company. For example, project managers at Yahoo were called producers, and various parts of the company were possessions. What Yahoo really needed was to be a technology company, but they tried to be something else, and it all ended up being neither one nor the other. That is why Yahoo has never been a company with a clearly defined personality.
The worst consequence was that they did not take programming seriously. Microsoft (in those days), Google, and Facebook had a programmer-oriented culture. But in Yahoo, programming was treated like a product. At Yahoo, end-user software was controlled by project managers and designers. The work of programmers was to take the work of designers and project managers and bring it to the final stage, turning it into code.
One of the obvious results of this practice was that when Yahoo created something, these things were often not very good. But it was not the worst problem. The worst problem was that they hired bad programmers.
Microsoft (in those days), Google and Facebook actively hired the very best programmers. Yaho did not do this. They preferred good programmers to bad ones, but they were not focused on hiring the smartest people the way the winners did. And if you take into account the competition for programmers at the time, during the Bubble, it is not surprising that the quality of their programs was uneven.
In technology, if you have bad programmers, you died. I can’t think of an example in which a company drowned in technical mediocrity and then came to life. Good programmers want to work with other good programmers. As soon as the quality of programmers in your company falls, you enter a spiral from which it is impossible to get out. 
Yahoo entered this spiral early. If Yahoo had a time when it attracted talents with the same force as Google, that time was over by 1998, when I started working there.
The company seemed to have aged prematurely. In most technology companies, power is transferred to people in suits and middle managers. At Yahoo, they seemed to deliberately speed up the process. They did not want to be a group of hackers. They wanted to be people in costumes. The media company should be managed by people in costumes.
When I first visited Google, there were about 500 people working there, just as much as at Yahoo when I got there to work. But the difference in work was noticeable. It was still a programmer-oriented culture. I remember how I talked in the cafeteria with programmers about the problem of manipulating search results (now known as SEO), and they asked: “What should we do?” Yahoo programmers would not ask this question. They had no reason to do it - their job was to do what the managers would say. I remember returning from Google, thinking: “Wow, this is still a startup.”
We cannot learn much from Yahoo's first fatal error. Most likely, you just need to hope that the company will avoid this problem and will not rely on an unreliable source of income. But startups can learn a good lesson from the second mistake. In the software business, you cannot afford not to have a programmer-oriented culture.
Perhaps the most impressive example of a programmer-oriented culture that I heard was brought by Mark Zuckerberg at Startup School in 2007. He said that before Facebook had hired programmers even for positions that do not require programming, such as HR and marketing.
So which companies need a programmer-oriented culture? What companies are involved in the software business? As Yahoo’s experience shows, the scope of application of this rule is much wider than most suppose. The answer is: any company that needs good software.
How can great programmers want to work in a company that doesn’t have a programmer-oriented culture while other companies have it? I can think of two reasons: they are well paid, or the field of activity itself is interesting, and in this area there are no companies with such a culture. In no other way can you attract good programmers to work in a culture oriented to people in costumes. And without good programmers, it will not be possible to produce good programs, and no matter how many people will work on the problem and how many processes will ensure “quality”.
A programmer-oriented culture often seems irresponsible. That is why people who propose to destroy it often use phrases such as “mature control system”. This is the phrase used in Yahoo. But there are things that are significantly worse than seeming irresponsibility. Defeat, for example.
 We came closest to targeting when I worked there, when we created pets.yahoo.com, in order to provoke a war between three startups that were engaged in animal feed, for the place of general sponsor.
 Theoretically, you can get out of the spiral by buying good programmers, instead of hiring them. You can get programmers who would never have come to you to get a job by buying their start-ups. But companies that are smart enough to do this do not need it.