WEBVTT - CZM Rewind: The Man That Destroyed Google Search 0:00:02.240 --> 0:00:07.200 All Zone Media. Hi, Better Offline fans, It's me, your 0:00:07.200 --> 0:00:11.200 host ed zetron. Now Here in America, it's Thanksgiving, so 0:00:11.360 --> 0:00:14.000 we're all given the week off. Robert and Sophie have 0:00:14.040 --> 0:00:16.040 told me I'm not allowed to podcast and if they 0:00:16.079 --> 0:00:18.439 find out, they're going to once again put me in 0:00:18.440 --> 0:00:21.200 the carnival prison. And once again I was sent a 0:00:21.239 --> 0:00:23.800 lo quality jpeg of a gun by Robert. Unclear what 0:00:23.840 --> 0:00:26.599 it means. Nevertheless, this is one of my favorite episodes 0:00:26.640 --> 0:00:29.479 I've ever recorded. It's about a real shithead who destroyed 0:00:29.480 --> 0:00:31.960 a product I love, you love and we both are 0:00:32.000 --> 0:00:34.879 doomed to use every day. Now. If you're not in America, 0:00:34.960 --> 0:00:37.640 and this is a big problem, I don't know what 0:00:37.720 --> 0:00:40.880 to tell you. The podcast doesn't cost you any money, unless, 0:00:40.880 --> 0:00:42.760 of course, you pay five dollars a month for our 0:00:42.800 --> 0:00:46.600 cooler Zoe Media, in which case I apologize deeply. I 0:00:46.640 --> 0:00:50.479 will be sleeping not at all this week as penance 0:00:50.479 --> 0:00:52.600 for what I've done to you. For the rest of 0:00:52.600 --> 0:00:59.720 you in America, no penance necessary, Please enjoy this episode. Hello, 0:01:00.000 --> 0:01:11.480 welcome to Better Offline. I'm your host ed zig tron. Well, 0:01:14.240 --> 0:01:16.720 and in the next two episodes, I'm going to tell 0:01:16.760 --> 0:01:19.640 you the names some of the people responsible for destroying 0:01:19.640 --> 0:01:22.720 the Internet. And I'm going to start on February fifth, 0:01:22.760 --> 0:01:26.959 twenty nineteen, when Ben Gomes, Google's former head of Search, well, 0:01:27.000 --> 0:01:30.000 he had a problem. Jerry Dishler, then the VP and 0:01:30.080 --> 0:01:33.080 GM of Ads at Google, and Shivvan Carterman, then the 0:01:33.160 --> 0:01:36.600 VP of Engineering Search and Ads on Google properties, had 0:01:36.640 --> 0:01:39.280 called something called a code yellow for search revenue due 0:01:39.280 --> 0:01:41.479 to and I quote emails that came out as part 0:01:41.520 --> 0:01:45.199 of Google's anti trust hearing steady weakness in the daily 0:01:45.280 --> 0:01:48.160 numbers and a likeliness that it would end the quarter 0:01:48.320 --> 0:01:53.200 significantly behind in metrics. That kind of unclear for those 0:01:53.280 --> 0:01:57.120 unfamiliar with Google's internal kind of scientology esque jargon, which 0:01:57.120 --> 0:02:00.880 means most people, let me explain, a good yellow isn't 0:02:00.920 --> 0:02:04.040 a terrible need to piss or some sort of crisis 0:02:04.040 --> 0:02:07.720 of moderate severity. The yellow, according to Stephen Levey's Tell 0:02:07.720 --> 0:02:10.720 All book about Google, refers to and I promise this 0:02:10.760 --> 0:02:12.880 is not a joke, the color of a tank top 0:02:13.080 --> 0:02:15.800 that a former VP of Engineering called Wayne Rosling used 0:02:15.800 --> 0:02:18.960 to wear during his time at the company. It's essentially 0:02:19.000 --> 0:02:22.320 the equivalent of deafcom I and activates as Levy explained 0:02:22.360 --> 0:02:25.320 a war room like situation where workers are pulled from 0:02:25.360 --> 0:02:27.720 their desks and into a conference room where they tackle 0:02:27.760 --> 0:02:30.760 the problem as a top priority. Any other projects or 0:02:30.760 --> 0:02:35.119 concerns are sidelined and independently. I've heard there are other 0:02:35.200 --> 0:02:38.000 colors like purple. I'm not going to get into that, though, 0:02:38.040 --> 0:02:42.040 it's quite boring and irrelevant to this situation. In emails 0:02:42.080 --> 0:02:45.000 released as part of the Department of Justices antitrust case 0:02:45.000 --> 0:02:48.360 against Google, as a previously mentioned, Dishler laid out several 0:02:48.440 --> 0:02:52.799 contributing factors. Search query growth was significantly behind forecast, the 0:02:52.800 --> 0:02:56.120 timing of revenue launches was significantly behind, and he had 0:02:56.120 --> 0:02:59.880 this vague worry that several advertiser specific and sector weaknesses 0:03:00.040 --> 0:03:03.160 existed in search. Now I want to cover something because 0:03:03.400 --> 0:03:05.280 I've messed up, and I really want to be clear 0:03:05.320 --> 0:03:08.919 about this. I've previously and erroneously referred to the code 0:03:09.000 --> 0:03:11.400 yellow as something that Gomes raised as a means of 0:03:11.400 --> 0:03:14.440 calling attention to the proximity of Google's ad side getting 0:03:14.480 --> 0:03:17.560 a little too close to Search. I'm afraid the truth 0:03:17.800 --> 0:03:21.880 is extremely depressing and so much grimmar. The code yellow 0:03:21.960 --> 0:03:24.560 was actually the rumble of the goddamn rot economy, with 0:03:24.680 --> 0:03:27.959 Google's revenue arms sounding the alarm that its golden goose 0:03:28.000 --> 0:03:32.080 wasn't laying enough eggs. Gomes, a Googler of nineteen years 0:03:32.120 --> 0:03:35.680 that basically built the foundation of modern search engines, should 0:03:35.720 --> 0:03:37.800 go down as one of the few people in tech 0:03:37.960 --> 0:03:40.880 that actually fought for an actual principle, and he was 0:03:40.920 --> 0:03:45.280 destroyed by a guy called Prabaka Ragavan, a computer scientist 0:03:45.320 --> 0:03:49.800 class traitor that sided with the management consultancy sect. More confusingly, 0:03:50.080 --> 0:03:52.800 one of their problems was that there was insufficient growth 0:03:52.800 --> 0:03:56.960 in queries, as in the amount of things that people 0:03:56.960 --> 0:03:59.960 were asking Google. It's a bit like if Ford decided 0:04:00.200 --> 0:04:03.000 that things were going poorly because their drivers weren't putting 0:04:03.080 --> 0:04:06.960 enough goddamn miles on their trucks. This whole story has 0:04:07.080 --> 0:04:09.200 personally upset me, and I think you're going to hear 0:04:09.240 --> 0:04:12.600 that in this but going through these emails is just 0:04:12.760 --> 0:04:17.640 very depressing. Anyway. A few days beforehand, on February first, 0:04:17.640 --> 0:04:22.239 twenty nineteen, Kristen Gil, then Google's VP Business Finance Officer, 0:04:22.440 --> 0:04:26.200 had emailed Shashi Thakker, then Google's VP of Engineering Search 0:04:26.240 --> 0:04:29.120 and Discover, saying that the ADS team had been considering 0:04:29.240 --> 0:04:32.279 a code yellow to close the search gap it was seeing, 0:04:33.880 --> 0:04:37.240 vaguely referring to how critical that growth was to an 0:04:37.360 --> 0:04:41.240 unnamed company plan. To be clear, this email was in 0:04:41.240 --> 0:04:44.039 response to Thaker stating that there is nothing that the 0:04:44.080 --> 0:04:46.680 search team could do to operate at the fidelity of 0:04:46.720 --> 0:04:50.359 growth that the ads department had demanded. Shashy Forward did 0:04:50.360 --> 0:04:52.839 the email to Gomes asking if there's any way to 0:04:52.920 --> 0:04:57.080 discuss this with Sandharpashai, Google CEO, and declared that there 0:04:57.120 --> 0:04:58.920 was no way he would sign up for a high 0:04:59.000 --> 0:05:03.679 fidelity businessman for daily active users on search. Sakker also 0:05:03.720 --> 0:05:06.719 said something that I've been thinking about constantly since I 0:05:06.800 --> 0:05:10.200 read these emails, that there was a good reason that 0:05:10.279 --> 0:05:14.479 Google's founders separated search from ads. I want you to 0:05:14.560 --> 0:05:19.040 remember that line for later. A day later, on February second, 0:05:19.080 --> 0:05:22.919 twenty nineteen, Thacker and Gomes shared their anxieties with Nick Fox, 0:05:23.120 --> 0:05:25.960 a vice president of Search and Google assistant, entering a 0:05:26.000 --> 0:05:29.120 multiple day long debate about Google's some lust for growth. 0:05:30.080 --> 0:05:32.800 This thread is a dark window into the world of 0:05:32.839 --> 0:05:35.960 growth focus tech, where the Kerr listed the multiple points 0:05:36.000 --> 0:05:39.159 of disconnection between ADS and search, discussing how the search 0:05:39.200 --> 0:05:43.640 team wasn't able to finally optimize engagement on Google without 0:05:43.640 --> 0:05:47.200 hacking it, a term that means effectively tricking users into 0:05:47.200 --> 0:05:49.599 spending more time on a site, and that doing so 0:05:49.720 --> 0:05:52.599 would lead them to and I quote, abandoned work on 0:05:52.640 --> 0:05:56.720 efficient journeys. In one email, Fox adds that there was 0:05:56.760 --> 0:05:59.560 a pretty big disconnect between what finance and ads wants 0:06:00.160 --> 0:06:05.279 Search was doing. Every part of this story pisses me 0:06:05.320 --> 0:06:09.080 off so much. When Gomes pushed back on the multiple 0:06:09.080 --> 0:06:11.640 requests for growth, Fox added that all three of them 0:06:11.640 --> 0:06:14.320 were responsible for Search and that Search was and again 0:06:14.360 --> 0:06:17.280 I quote, the revenue engine of the company, and that 0:06:17.320 --> 0:06:20.200 bartering with the ads and finance teams was now potentially 0:06:20.279 --> 0:06:24.320 the new reality of their jobs. On February sixth, twenty nineteen, 0:06:24.640 --> 0:06:27.240 Gomes said that he believed that Search was getting too 0:06:27.320 --> 0:06:30.120 close to the money and ended his email by saying 0:06:30.120 --> 0:06:32.599 that he was concerned that growth is all that Google 0:06:32.680 --> 0:06:37.120 was thinking about. On March twenty second, twenty nineteen, Google 0:06:37.200 --> 0:06:40.920 VP of Product Management Darshan Kantac would declare the end 0:06:41.000 --> 0:06:45.000 of the code yellow. The thread mostly consisted of congratulatory 0:06:45.040 --> 0:06:48.720 emails until Gomes made the mistake of responding congratulating everyone, 0:06:48.920 --> 0:06:51.040 saying that the plans architected as part of the code 0:06:51.080 --> 0:06:56.440 yellow would do well throughout the year. Enter probaka Ragavan, 0:06:56.960 --> 0:06:59.800 then Google's head of Ads and the true mastermind behind 0:06:59.800 --> 0:07:03.080 the Yellow, who would respond curtly saying that the current 0:07:03.120 --> 0:07:07.000 revenue targets were addressed by heroic RPM engineering and that 0:07:07.040 --> 0:07:11.679 the core query softness continued without mitigation, a very clunky 0:07:11.680 --> 0:07:14.800 way of saying that despite these changes, query growth was 0:07:14.840 --> 0:07:17.760 not happening at the rate he needed it to. A 0:07:17.840 --> 0:07:21.280 day later, Gomes emailed Fox Andhaker an email he intended 0:07:21.280 --> 0:07:24.080 to center Ragavan. He led by saying that he was 0:07:24.120 --> 0:07:27.560 annoyed both personally and on behalf of the search team. 0:07:27.840 --> 0:07:31.800 In this very long email, he explained in arduous detail 0:07:32.040 --> 0:07:35.440 how one might increase engagement with Google Search, but specifically 0:07:35.520 --> 0:07:39.120 added that they could increase queries quite easily in the 0:07:39.120 --> 0:07:42.760 short term, but only in user negative ways, like turning 0:07:42.760 --> 0:07:47.040 off spell correction or ranking improvements, or placing refinements effectively 0:07:47.080 --> 0:07:49.920 labels all over the page, adding that it was possible 0:07:49.960 --> 0:07:52.119 that there are trade offs here between the different kinds 0:07:52.120 --> 0:07:55.120 of user negativity caused by engagement hacking, and that he 0:07:55.240 --> 0:07:59.480 was deeply, deeply uncomfortable with this, He also added that 0:07:59.520 --> 0:08:02.640 this was the reason he didn't believe that queries, as 0:08:02.640 --> 0:08:06.200 in the amount of the things with people searching on Google, 0:08:06.400 --> 0:08:08.840 were a good metric to measure search, and that the 0:08:08.840 --> 0:08:12.320 best defense against the weaknesses of queries was to create 0:08:12.320 --> 0:08:15.360 compelling user experiences that make users want to come back. 0:08:15.840 --> 0:08:18.960 Crazy idea there, what if the product was good? Not 0:08:19.440 --> 0:08:23.960 good enough? Of prabaka, so little bit of history about 0:08:24.000 --> 0:08:26.960 Google here. They regularly throughout the year do Core updates 0:08:27.040 --> 0:08:30.680 to Search. These are updates that change the algorithm. Let's say, okay, 0:08:30.680 --> 0:08:32.320 we're going to suppress this kind of thing. We can 0:08:32.360 --> 0:08:35.679 elevate this kind of thing. And they are actually the 0:08:35.720 --> 0:08:40.640 reason that search changes. It's why certain sites suddenly disappear 0:08:40.760 --> 0:08:43.120 or reappear. It's why sites get a ton of traffic, 0:08:43.160 --> 0:08:45.320 some don't get any, and so on and so forth. 0:08:45.320 --> 0:08:47.920 But they do a lot of them. The one that's 0:08:47.960 --> 0:08:50.080 really interesting and a little bastard and I went and 0:08:50.120 --> 0:08:53.679 looked through pretty much the last decade of these. The 0:08:53.720 --> 0:08:55.640 one that stood out to me was the March twenty 0:08:55.720 --> 0:08:58.480 nineteen Core update to Search, which happened about a week 0:08:58.559 --> 0:09:01.360 before the end of the code yell, meaning that it's 0:09:01.520 --> 0:09:04.840 very likely that this was a result of Prabaka's bullshit. 0:09:05.840 --> 0:09:08.120 So this was expected to be one of the largest 0:09:08.200 --> 0:09:10.240 updates to search in a very long time, and I'm 0:09:10.320 --> 0:09:14.160 quoting Search Engine Journal there. Yet when it launched, many 0:09:14.200 --> 0:09:17.760 found that the update mostly rolled back changes and traffic 0:09:17.800 --> 0:09:21.160 was increasing to sites that had been suppressed by previous updates, 0:09:21.160 --> 0:09:24.960 like Google Search's Penguin update from twenty twelve that specifically 0:09:25.000 --> 0:09:29.160 targeted spami search results. There were others that were seeing 0:09:29.200 --> 0:09:31.480 traffic as well from an update that happened on the 0:09:31.480 --> 0:09:34.320 first of August twenty eighteen that was a few months 0:09:34.360 --> 0:09:38.360 after Gomes became head of Search. While I'm guessing here, 0:09:38.520 --> 0:09:40.679 I really don't know. I do not work for Google. 0:09:40.720 --> 0:09:43.520 I do not have friends there. I think the timing 0:09:43.559 --> 0:09:46.280 of the March twenty nineteen Core update, along with the 0:09:46.320 --> 0:09:50.199 traffic increases the previously suppressed sites that one hundred percent 0:09:50.200 --> 0:09:54.120 were spamy seo nonsense. I think these suggest that Google's 0:09:54.160 --> 0:09:56.480 response to the Cojello was to roll back changes that 0:09:56.520 --> 0:10:00.559 were made to maintain the quality of search. A few 0:10:00.600 --> 0:10:03.240 months later, in May twenty nineteen, Google would roll out 0:10:03.240 --> 0:10:06.120 a redesign of how ads were shown on Google Search. 0:10:06.960 --> 0:10:10.480 Specifically on mobile, replacing the bright green AD label and 0:10:10.840 --> 0:10:13.320 URL color on ads with a tiny, little, bolded black 0:10:13.360 --> 0:10:15.960 note that said AD in the smallest font you could 0:10:16.000 --> 0:10:19.400 possibly put there, with the link looking otherwise identical to 0:10:19.440 --> 0:10:23.120 a regular search link. I guess that's how they managed 0:10:23.160 --> 0:10:26.120 to start hating their numbers. Hah. And then in January 0:10:26.160 --> 0:10:29.040 twenty twenty, Google would bring this change to desktop, and 0:10:29.080 --> 0:10:32.000 the vergins John Porter would suggest that it made Google's 0:10:32.040 --> 0:10:38.680 ads look just like search results now awesome. Five months later, 0:10:38.920 --> 0:10:41.760 a little over a year after the code yellow situation, 0:10:42.240 --> 0:10:45.880 Google would make Probakar Ragavan the head of Google Search, 0:10:46.480 --> 0:10:49.320 with Jerry Dishler taking his place as the head of Ads. 0:10:50.679 --> 0:10:53.840 After nearly twenty years of building Google Search domes would 0:10:53.840 --> 0:10:57.240 be relegated to the SVP of Education at Google Domes. 0:10:57.400 --> 0:10:59.840 It was a critical part of the original team that 0:11:00.000 --> 0:11:03.479 made Google Search work, who has been credited with establishing 0:11:03.480 --> 0:11:06.560 the culture of the world's largest and most important search engine, 0:11:06.800 --> 0:11:10.560 was chased out by a growth hungry managerial type, several 0:11:10.640 --> 0:11:14.720 of them actually led by Probagar Ragavan, a management consultant 0:11:14.760 --> 0:11:18.280 wearing an engineer costume. As a side note, by the 0:11:18.280 --> 0:11:21.439 way I use the term management consultant there as a pejorative, 0:11:21.760 --> 0:11:24.920 while he exhibits all the same being counting morally young 0:11:24.960 --> 0:11:27.800 guided behaviors of a management consultant. From what I can tell, 0:11:27.920 --> 0:11:30.599 Ragavan has never actually worked in that particular sector of 0:11:30.640 --> 0:11:34.600 the economy. But you know who has. San Dhar Pishai, 0:11:34.800 --> 0:11:38.640 the CEO of Google, who previously worked at McKinsey, arguably 0:11:38.679 --> 0:11:41.920 the most morally abhorrent company that's ever existed, having played 0:11:42.000 --> 0:11:44.400 roles both in the two thousand and eight financial crisis, 0:11:44.480 --> 0:11:46.480 where it encouraged banks to load up on debt and 0:11:46.520 --> 0:11:50.120 floored mortgage backed securities, and the ongoing opiord crisis, where 0:11:50.120 --> 0:11:52.960 it effectively advised Perdue Farmer on how to growth hack 0:11:53.040 --> 0:11:58.160 sales of oxy content, an extremely addictive painkiller. McKinsey has 0:11:58.160 --> 0:12:01.000 paid nearly one billion dollars over several settlements due to 0:12:01.080 --> 0:12:04.720 its work with Perdue. But I'm getting sidetracked, but one 0:12:04.800 --> 0:12:08.559 last point. McKinsey is actively anti labor. When a company 0:12:08.600 --> 0:12:12.360 brings in a McKinsey consultant, they're often there to advise 0:12:12.400 --> 0:12:16.120 on how to cut costs, which inevitably means layoffs and outsourcing. 0:12:16.600 --> 0:12:19.839 McKinsey is to the middle class. What fleshy in bacteria 0:12:20.280 --> 0:12:32.840 is the skin. But back to the emails, which are 0:12:32.880 --> 0:12:37.439 a stark example of the monstrous, disgusting rot economy, the 0:12:37.440 --> 0:12:41.080 growth that all costs, mindset that's dominating the tech ecosystem. 0:12:41.200 --> 0:12:43.800 And if you take one thing away from this episode, 0:12:43.920 --> 0:12:46.880 I want it to be the name Prabakar Ragavan, and 0:12:46.920 --> 0:12:50.160 I'm understanding that there are people responsible for the current 0:12:50.280 --> 0:12:54.520 state of the Internet. These emails, which I really encourage 0:12:54.520 --> 0:12:55.880 you to look up and if you go to where's 0:12:55.880 --> 0:12:58.040 youreaed dot at, you'll be able to see a newslet 0:12:58.160 --> 0:13:01.319 that has links to them. Well, these emails tell a 0:13:01.360 --> 0:13:05.440 dramatic story about how Google's finance and advertising teams, led 0:13:05.440 --> 0:13:08.800 by Ragavan, with the blessing of CEO Sandhar Pashai, the 0:13:08.880 --> 0:13:12.760 McKinsey guy, actively worked to make Google worse to make 0:13:12.800 --> 0:13:16.120 the company more money. This is exactly what I mean 0:13:16.160 --> 0:13:19.560 when I talk about the economy, an illogical, product destroying 0:13:19.600 --> 0:13:23.600 mindset that turns products you love into torturous, frustrating, quasi 0:13:23.640 --> 0:13:26.319 tools that require you to fight the company to get 0:13:26.360 --> 0:13:30.400 the thing you want. Ben Gomes was instrumental in making 0:13:30.440 --> 0:13:33.319 search work both as a product and a business. He 0:13:33.440 --> 0:13:36.160 joined the company in nineteen ninety nine, a time long 0:13:36.200 --> 0:13:39.120 before Google established dominance in the field, and the same 0:13:39.200 --> 0:13:42.000 year when Larry Page and Sergey Britain tried to sell 0:13:42.040 --> 0:13:45.199 the company to Excite for one million dollars, only to 0:13:45.280 --> 0:13:48.280 walk away after Vinnard Coosler and Excite investor and co 0:13:48.320 --> 0:13:51.600 founder of some Microsystems that's now a VC who tried 0:13:51.640 --> 0:13:54.440 to stop people going to a beach in half Moon Bay, Well, 0:13:54.480 --> 0:13:56.280 he tried to low ball them with a seven hundred 0:13:56.280 --> 0:13:59.360 and fifty thousand dollars offer, also known as a one 0:13:59.400 --> 0:14:03.440 hundred square apartment in San Francisco. In an interview with 0:14:03.480 --> 0:14:07.160 Fast Companies Harry McCracken from twenty eighteen, Gomes frayed Google's 0:14:07.200 --> 0:14:10.440 challenge as taking the page erank algorithm from one machine 0:14:10.600 --> 0:14:13.199 to a whole bunch of machines and they weren't very 0:14:13.200 --> 0:14:17.120 good machines at the time. Despite his impact and tenure, 0:14:17.400 --> 0:14:19.480 Gomes had only been made head of Search in the 0:14:19.480 --> 0:14:23.360 middle of twenty eighteen after John Gillanderia moved to Apple 0:14:23.440 --> 0:14:26.800 to work on its machine learning in AR strategy. Domes 0:14:26.840 --> 0:14:29.920 had been described as Google's searches are beloved for his 0:14:29.960 --> 0:14:35.480 ability to communicate across Google's many quite decentralized apartments. Every 0:14:35.520 --> 0:14:39.160 single article I've read about Gomes and his tenure at 0:14:39.160 --> 0:14:42.480 Google spoke of a man deeply ingrained in the foundation 0:14:42.800 --> 0:14:45.720 of one of the most important technologies ever made, a 0:14:45.720 --> 0:14:48.400 man who had dedicated decades to maintaining a product with 0:14:48.520 --> 0:14:51.840 a and I quote Gomes here guiding light of serving 0:14:52.120 --> 0:14:56.440 the user and using technology to do that. And when 0:14:56.480 --> 0:14:59.480 finally given the keys to the kingdom, the ability to 0:14:59.520 --> 0:15:02.880 elevate Google Search even further, he was rap fucked by 0:15:02.880 --> 0:15:06.160 a series of rotten carerists trying to please Wall Street, 0:15:06.360 --> 0:15:09.320 led by Probakar Ragavan. Do you want to know what 0:15:09.360 --> 0:15:12.960 Provacar Ragavan's old job was? What Probacar Ragavan, the new 0:15:12.960 --> 0:15:15.480 head of Google Search, the guy that ran Google Search, 0:15:15.520 --> 0:15:17.960 that runs Google Search right now, that is running Google 0:15:17.960 --> 0:15:20.240 Search into the goddamn ground. Do you want to know 0:15:20.280 --> 0:15:23.400 what his job was? His job before Google, He was 0:15:23.440 --> 0:15:26.440 the head of search for god damn Yahoo from two 0:15:26.520 --> 0:15:30.320 thousand and five through two thy and twelve when he 0:15:30.440 --> 0:15:33.840 joined the company. When Probakar Ragavan took over Yahoo Search, 0:15:34.080 --> 0:15:37.280 they held a thirty point four percent market share, not 0:15:37.480 --> 0:15:40.800 far from Google's own thirty six point nine percent, and 0:15:40.920 --> 0:15:44.480 miles ahead of the fifteen point seven percent that Microsoft's 0:15:44.600 --> 0:15:48.600 MSN Search had. By May twenty twelve, Yahoo was down 0:15:48.600 --> 0:15:51.120 to just thirteen point four percent and had shrunk for 0:15:51.280 --> 0:15:54.520 the previous nine consecutive months, and was being beaten by 0:15:54.600 --> 0:15:58.880 even the newly released Bing. That same year, Yahoo had 0:15:58.920 --> 0:16:02.400 the largest layoffs in its corporate history, shedding two thousand employees, 0:16:02.640 --> 0:16:06.720 or fourteen percent of its overall workforce. The man who deposed, 0:16:06.760 --> 0:16:09.960 Ben Gomes, someone who worked on Google Search from its 0:16:10.120 --> 0:16:12.680 very beginnings, was so shit at his job that in 0:16:12.680 --> 0:16:15.440 two thousand and nine, Yahoo effectively threw in the towel 0:16:15.560 --> 0:16:18.920 on its own search tech, instead choosing to license Being's 0:16:19.040 --> 0:16:21.920 engine in a ten year deal. If we take a 0:16:22.000 --> 0:16:25.920 long view of things, this likely precipitated the overall decline 0:16:25.920 --> 0:16:28.320 of the company, which went from being worth one hundred 0:16:28.320 --> 0:16:30.560 and twenty five billion dollars at the peak of the 0:16:30.600 --> 0:16:33.280 dot com boom to being sold to Verizon for four 0:16:33.320 --> 0:16:36.600 point eight billion dollars in twenty seventeen, which is roughly 0:16:36.640 --> 0:16:40.560 a three thousand square foot apartment in San Francisco. With 0:16:40.680 --> 0:16:43.520 search no longer a priority, in making less money, for 0:16:43.560 --> 0:16:46.600 the company, Yahoo decided to pivot into Web two point 0:16:46.680 --> 0:16:49.560 zero and original content, making sum bats that paid off, 0:16:49.680 --> 0:16:52.360 but far far too many that did not. It spent 0:16:52.480 --> 0:16:55.960 one point one billion dollars on Tumblr in twenty thirteen, 0:16:56.200 --> 0:16:58.400 only for Verizon to sell it for just three million 0:16:58.440 --> 0:17:01.560 dollars in twenty nineteen. Zimbra in two thousand and seven, 0:17:01.560 --> 0:17:04.920 ostensibly to complete with the new Google Apps productivity Suite, 0:17:05.080 --> 0:17:07.240 only to sell it for a reported fraction of the 0:17:07.280 --> 0:17:10.600 original purchase price to VMware a few years later. That's 0:17:10.640 --> 0:17:14.960 not his fault, but nevertheless, Yahoo was a company without 0:17:15.000 --> 0:17:18.200 a mission, a purpose, or an objective. Nobody and I'll 0:17:18.240 --> 0:17:21.000 speculate even though his leading the company really knew what 0:17:21.040 --> 0:17:24.000 it was and what it did. Anyway, just a big 0:17:24.040 --> 0:17:26.520 shout out right now to Kura Swisher, who referred to 0:17:26.560 --> 0:17:31.240 Pradaka as well respected. When he moved from Yahoo to Google. 0:17:31.680 --> 0:17:35.600 He absolutely nailed at Kara bang up job. In an 0:17:35.640 --> 0:17:38.520 interview with ZDNTS Dan Farber from two thousand and five, 0:17:38.840 --> 0:17:42.439 Ragavan spoke of his intent to align the commercial incentives 0:17:42.440 --> 0:17:45.440 of a billion content providers with social good intent while 0:17:45.440 --> 0:17:48.280 at Yahoo and his eagerness to inspire the audience to 0:17:48.280 --> 0:17:53.080 give more data. What anyway before that? It's it's actually 0:17:53.280 --> 0:17:56.240 hard to find out exactly what Ragavan did, though according 0:17:56.240 --> 0:17:59.399 to zd net, he spent fourteen years doing search and 0:17:59.480 --> 0:18:05.359 data mind research ibm MM. In April twenty eleven, The 0:18:05.359 --> 0:18:08.320 Guardian ran an interview with Ragavan that called him Yahoo's 0:18:08.359 --> 0:18:12.720 secret weapon, describing his plan to make rigorous scientific research 0:18:12.760 --> 0:18:16.440 and practice to inform Yahoo's business from email to advertising, 0:18:16.680 --> 0:18:19.879 and how under then CEO Carol Bart's the focus had 0:18:19.920 --> 0:18:23.520 shifted to the direct development of new products. It speaks 0:18:23.520 --> 0:18:28.200 of Ragavan's scientific approach and his steady process based logic 0:18:28.240 --> 0:18:31.400 to innovation that is very different to the common perception 0:18:31.680 --> 0:18:34.680 the ideas and development are more about luck and spontaneity. 0:18:35.000 --> 0:18:37.119 A sentence that I'm only reading to you because I 0:18:37.160 --> 0:18:39.439 really need you to hear how stupid it sounds and 0:18:39.440 --> 0:18:43.959 how specious some of the tech press used to be. Frankly, 0:18:44.040 --> 0:18:47.320 this entire article is ridiculous, so utterly vacuous that I'm 0:18:47.359 --> 0:18:49.919 actually astonished. I don't want to name the reporter. I 0:18:50.000 --> 0:18:54.520 feel bad. What about Ragavan's career made this feel right? 0:18:55.040 --> 0:18:57.920 How has nobody connected these thoughts before I have a 0:18:58.000 --> 0:18:59.879 day job. I run a PR firm, I am a 0:19:00.119 --> 0:19:05.479 blogger with a podcast, and I'm the one who said, yeah, okay, 0:19:05.600 --> 0:19:08.080 drag Uller is now the CEO of the blood Bank. 0:19:08.320 --> 0:19:11.040 Nobody saw this. Nobody saw this at the time. I 0:19:11.119 --> 0:19:13.560 just feel a bit crazy. I feel a bit crazy. 0:19:13.600 --> 0:19:16.840 But to be clear, this was something written several years 0:19:16.880 --> 0:19:20.960 after Yahoo had licensed its search technology to Microsoft in 0:19:21.000 --> 0:19:24.520 a financial deal that the next CEO, Marissa Maya, who 0:19:24.600 --> 0:19:28.800 replaced Barts, was still angry about for years. Ragavan's reign 0:19:28.880 --> 0:19:31.600 as what zd Neat referred to as the search Master 0:19:32.200 --> 0:19:36.320 was one so successful that it ended up being replaced 0:19:36.320 --> 0:19:38.560 by a search engine that not a single person in 0:19:38.600 --> 0:19:43.320 the world enjoys saying out loud. The Guardian article ran 0:19:43.440 --> 0:19:47.480 exactly one year before dramatic layoffs at Yahoo that involved 0:19:47.520 --> 0:19:50.840 firing entire divisions worth of people, and four months before 0:19:50.880 --> 0:19:54.000 Carol Barts would be fired by telephone by then chairman 0:19:54.080 --> 0:19:58.199 Roy Bostock. Her replacement, Scott Thompson, who previously served as 0:19:58.240 --> 0:20:00.720 president of PayPal, would last a whole five months in 0:20:00.760 --> 0:20:04.320 the role before he was replaced by former Google executive 0:20:04.359 --> 0:20:07.000 Marissa Mayer. In part because it emerged he lied on 0:20:07.040 --> 0:20:11.760 his resume about having a computer science degree. Hey, Brobaka, 0:20:11.760 --> 0:20:28.400 did you not notice that? Anyway? Whatever, Barts joined Yahoo 0:20:28.480 --> 0:20:30.800 in two thousand and nine, so about four years into 0:20:31.240 --> 0:20:35.480 Braba Kha's reign of terror. I guess, and she joined 0:20:35.480 --> 0:20:38.360 in the aftermath of its previous CEO, Jerry Yang, refusing 0:20:38.400 --> 0:20:41.600 to sell the company to Microsoft for forty five billion dollars. 0:20:41.960 --> 0:20:44.199 In her first year, she laid off hundreds of people 0:20:44.440 --> 0:20:47.520 and struck a deal that I've mentioned before to power 0:20:47.600 --> 0:20:51.000 Yahoo Search using Microsoft's being search engine tech, with Microsoft 0:20:51.040 --> 0:20:53.440 paying Yahoo eighty eight percent of the revenue it gained 0:20:53.440 --> 0:20:55.880 from searches, a deal that made ya who are couple 0:20:55.960 --> 0:20:58.040 hundred million dollars for handing over the keys and the 0:20:58.080 --> 0:21:01.960 tech to its most high traffic platform. As I previously stated, 0:21:01.960 --> 0:21:05.639 when Brabakhar Ragavan, Yahoo's secret weapon was doing his work, 0:21:05.760 --> 0:21:08.600 Yahoo Search was so valuable that it was replaced by 0:21:08.640 --> 0:21:13.040 Bing its sole value. In fact, I mean, maybe I'm 0:21:13.040 --> 0:21:15.760 being a little unfair. There's a way of looking at 0:21:15.760 --> 0:21:18.760 this that you could say that Yahoo's entire value at 0:21:18.760 --> 0:21:21.119 the end of his career was driven by nostalgia and 0:21:21.160 --> 0:21:25.679 association with days before he worked there. Anyway, thanks to 0:21:25.720 --> 0:21:28.400 the state of modern search. It's actually very, very difficult 0:21:28.440 --> 0:21:31.240 to find much about Ragavan's history. It took me hours 0:21:31.240 --> 0:21:35.000 of digging through Google and at one point being embarrassingly 0:21:36.080 --> 0:21:38.520 to find three or four articles that went into any 0:21:38.560 --> 0:21:41.119 depth about him. But from what I've gleaned, his expertise 0:21:41.160 --> 0:21:44.480 lies primarily in failing upwards ascending through the ranks of 0:21:44.520 --> 0:21:49.080 technology on the momentum from the explosions he's coursed. In 0:21:49.119 --> 0:21:51.879 a wide interview from twenty twenty one, GLAD handler Stephen 0:21:51.960 --> 0:21:54.520 Levy said Ragavan isn't the CEO of Google, he just 0:21:54.600 --> 0:21:57.280 runs the place, and described his addition to the company 0:21:57.359 --> 0:22:00.600 as a move from research to management. Bill Levy calls 0:22:00.680 --> 0:22:03.840 him a world class computer scientist who has authored definitive 0:22:03.840 --> 0:22:06.760 text in the field, which is true. He also describes 0:22:06.840 --> 0:22:10.439 Ragavan as choosing a management track, which definitely tracks with 0:22:10.480 --> 0:22:14.000 everything I found out about him. Ragavan proudly declares that 0:22:14.000 --> 0:22:16.760 Google's third party ad tech plays a critical role in 0:22:16.840 --> 0:22:19.439 keeping journalism alive and a really shitty answer to a 0:22:19.520 --> 0:22:21.840 question that was also made at a time when he 0:22:21.960 --> 0:22:26.359 was in aggressively incentivizing search engine optimized content and a 0:22:26.480 --> 0:22:28.680 year after he'd deposed someone who actually gave a shit 0:22:28.760 --> 0:22:33.600 about search. Under Ragavan, Google has become less reliable and 0:22:33.600 --> 0:22:37.840 it's dominated by search engine optimization and just outright spam. 0:22:38.119 --> 0:22:40.880 And I've said this before, but look, we complain about 0:22:40.880 --> 0:22:43.920 the state of Twitter under Elon Musk and justifiably he's 0:22:43.920 --> 0:22:47.520 a vile, anti, semi racist bigger. We all know this. 0:22:47.640 --> 0:22:51.600 It's fully true. We can say a million times. However, 0:22:51.680 --> 0:22:56.840 I'd argue that Ragavan, by extension Sandhar Pashai, deserve one 0:22:56.920 --> 0:23:02.440 hundred times more criticism. They've on unfathomable damage to society. 0:23:02.480 --> 0:23:05.840 You really can't fix the damage they've been doing and 0:23:05.880 --> 0:23:08.399