Learning from Yahoo’s mistakes

Nicholas Carlson wrote a fantastic book about Yahoo entitled “Marissa Mayer and The Fight to Save Yahoo.” Below I have paraphrased what I found to be valuable sections, verbatim; the full book is definitely worth the read.

Allow employs to give feedback, anonymously. The questions were submitted during the week leading up to the FYI through an application called “Yahoo Moderator” on Yahoo’s internal network. Everyone in the company could see questions after they were submitted, and employees would vote on which questions they wanted Mayor to answer that week.

Don’t cut entire divisions. Project Alpha sought to reduce Yahoo’s workforce by cutting whole divisions from the company rather than examining the work of each employee in each group and identifying the poor performers who should go and the high performers who should stay, even if that meant moving to another group. When Mayer heard that, she couldn’t believe it.

Don’t rate employees like this. What they had begun to hate about the Marissa Mayer regime was her awful employee rating system — the QPRs. In an effort to flush out some of Yahoo’s poor performers and cut some cost in the process, Mayer had rolled out a system where managers had to grade their employees on a fixed curve. Even on top-performing teams, someone always had to get a poor score. Poor and sometimes even average scores made it nearly impossible to transfer jobs, earn full bonuses, or get promotions. Because someone would have to be ranked worst even on teams full of all-star performers, the most talented employees refused to work together. Because employees were not judged on their own work, but on how well they did relative to their peers, they would actively seek to undermine each other. Workers would prioritize tasks that got them closer to their personal goals over doing anything else. This made sense. Collaborating and helping out on a project that wasn’t going to get you closer to an “exceeds” rating was just a stupid thing to do.

The worst part was, every quarter, managers would guide their teams toward collective goals, and then, even if all of those goals were met, they had to single out a few people and tell them they had missed expectations. Somebody always had to “occasionally miss.” Even if no one ever missed.

Mayer wasn’t necessarily responsible for the climbing stock price, Alibaba was. She joined Yahoo just before its stock price started soaring due to the company’s investment in a booming Chinese startup called Alibaba. Investors piled into Yahoo with little care for how well the core business Mayer was running performed on a quarterly basis. In August of 2005, the companies announced the deal “Yahoo gave Alibaba $1 billion in cash, a license to use Yahoo technology, and Yahoo China in exchange for a 44% stake in Alibaba.”

.

The birth of the internet. In April 1963, computer scientist J.C.R. Licklider wrote a memo proposing an “intergalactic Computer Network.” That October, he went to work for the US Department of Defense. There, Licklider persuaded two colleagues, Ivan Sutherland and Bob Taylor, to build a prototype network. Sutherland and Taylor had one running by the end of 1969. By 1971, fifteen computers were connected to the network, by then called ARPANET.

The birth of Yahoo. Finally, one of those two students. David Filo, made a list of all the sites he liked. He shared it with his friend Jerry Yang who also made a list. Filo combined the lists and in early 1994, Yang published the list on a web page hosted for him by Stanford. By September 1994, the site had two thousand links on its site and was getting fifty thousand visits per day. By then, Yang had begun organizing the links on the site into nineteen different categories, from Computers to Art to Business.

Perhaps, the only reason Yahoo ever became a business at all was what happened in October 1994. That’s when Marc Andreessen’s startup, Netscape, launched an even more consumer-friendly version of the Mosaic browser, called Navigator. Netscape Navigator linked directly to Yahoo.com from a button at the top of the application labeled “directory.” As millions of people tried the World Wide Web for the first time, they went straight to Yahoo to figure out where to go.

Yahoo held its initial public offering on April 12, 1996. The 151 percent IPO spike was $848 million. That meant the company was worth $808 million more than the valuation at which Masayoshi Son and Softbank bought in three months before.

How Mark Cuban got rich. Yahoo paid a crazy high price of $5.7 billion for a startup called Broadcast.com. It sent Yahoo’s stock up 22 percent the next day.

Yahoo could have bought eBay for nothing. In 1999, Yahoo had been close to acquiring online auctions start up eBay. eBay’s board and Yahoo’s board approved the deal, but it fell apart when Mallet demanded that eBay’s CEO, Meg Whitman, report to him.

If you have do cuts, do them deep and do them once. There were two problems. The first was that a huge percentage of the billion dollars came from ads sold to dot-coms; dot coms that Decker believed did not have defensible business models. The cuts were deep. Four hundred people lost their jobs. But the cuts were swift and once they were over, Yahoo’s employees knew they were safe.

Listen to your customers to decide where to focus. It was the height of the Jeff Mallett era, when he was using “click-finding” to decide in which “pods” he wanted to invest to build products and services for users coming to Yahoo.com. When Mallett found an activity Yahoo users liked to do, he would decide whether or not Yahoo should build a product around that activity, buy a company that made that product, or partner with one that made it. For example, in the case of email, Mallett had decide to buy. So Yahoo bought a company called Four11 for $94 million and used its technology to launch Yahoo Mail.

Yahoo could have bought Google. Back in 1997, when Google was still a Stanford thesis project called BackRub, its creator, Larry Page, wanted to sell it for a million dollars so he could finish his PhD and become a professor. Page met with Jerry Yang and David Filo. Everyone got along, but Yahoo passed.

How Google destroyed Yahoo in search. The Google logo on every Yahoo search was free advertising. It let consumers know why Yahoo search had suddenly gotten better: this thing called Google. Millions of web users went to Google.com and never came back.

AdWords quickly proved to be an effective tool for marketers and this made it lucrative for Google. AdWords got a much higher click-through rate than any other ads on the Internet. Yahoo never really had a chance to buy Google in 2002. That moment in 1997 was it.

There are a million reasons why Yahoo lost to Google in search, but there’s also just one reason why: Yahoo put the ads on its search results pages in the wrong order. Google put the ads on its pages in the right order. Ads were typically sold on an “impressions” basis — actually, on a per-one-thousands-impressions basis, also know as a CPM (cost per mille). They were sold that way because TV, radio, and print ads were sold that way. Then, in 1998, and entrepreneur named Bill Gross created GoTo.com, which would become Overture. Instead of charging advertisers on a per-impression basis, he thought it would be a better idea to charge them on a cost-per-click (CPC) basis. He was right. Overture, and then Yahoo, would arrange the ads on the search results page from top to bottom in the order of which advertisers were willing to pay the most per click. It was a straight auction for every keyword. To the highest bidders went the top, most valuable slot.

Google’s system for determining the arrangement was more complicated. It optimized the order for yield. Sometimes, it would put an ad at the very top of a page from an advertiser with a lower bid, because the ad was much more likely to be clicked on by users. Google’s logic was sound: an ad that pays $0.55 per click is more valuable than an ad that gets $1.00 per click if it gets clicked on twice as much. Google would also do things like award the top slot to advertisers who bid the second-highest amount. That encouraged advertisers to bid high, knowing that they would never be caught far outbidding the marketplace. Google also played around with which ads it would show against which queries, never being as basic as Overtures’s strict auctions.

Ads are a winner take all market. One reason that happened was that as Google added market share, search marketers decided to quit splitting their budgets between Yahoo and Google and concentrate all their bidding power in one market — Google’s. That further drove yield, further enabling Google to buy market share, and on it went.

Google’s market share would grow so large that search marketers would decide they couldn’t afford to waste money buying ads with other search engines. They would decide to use that budget only on Google ads to make sure they could bid as high as possible. Google would own an effectively monopoly on the Internet’s only real business model.

Yahoo could have bought Facebook. Yahoo had thought about buying Facebook since Zuckerberg created it in a Harvard dorm room in 2004. That year, a Yahoo M&A executive named Make Marquez met with Zuckerberg and Facebook president Sean Parker about a possible deal. Facebook’s first outside investor, Peter Thiel, thought Zuckerberg would be insane to walk away from Yahoo. Zuckerberg eventually said he would do the deal but only if Yahoo offered $1 billion. “We’re going to have to do this deal at $850 million, not the billion you wanted,” Semel told Zuckerberg. Zuckerberg was quiet. He looked disappointed. Nilsson felt the tension rise in the room. It felt awkward. No one on his team wanted to retrade the deal like this. It felt like reneging.

Facebook was hardly the only startup Yahoo nearly acquired during the Semel years, didn’t, and then watched grow into a giant. Yahoo looked at LinkedIn and Twitter, two companies worth $25 billion in 2014. Yahoo probably could have had YouTube.

Another reason why Ad-tech is so challenging. All those users clicking around Facebook meant Facebook had a ton of display advertising inventory to sell. Facebook’s growth flooded the display advertising market with inventory. Demand from marketers could not keep up with all the new supply — budgets were not increasing at the pace of Facebook’s growth — and the average price the industry, including Yahoo, was able to charge declined steeply.

1 genius developer is worth 1,000 average developers. Soon after joining, Singhal decided the code Google used for figuring out how to rank its search results needed a major overhaul. It had been written by Sergey Brin, and it was very sloppy. Singhal rewrote the whole thing in two months, adding huge improvements to relevancy and speed.

The shelf-life of tech is short, especially software. Andreessen talked about the difference between technology companies and “normal” companies. He said the output of normal companies is their product: cars, shoes, life insurance. In his view the output of technology companies is innovation. Whatever they are selling today they will be selling something different in five years. If they stop innovating, they die.