WEBVTT - When Did Google Stop Being a Search Company? 0:00:04.440 --> 0:00:12.360 Welcome to tech Stuff, a production from iHeartRadio. Hey there, 0:00:12.360 --> 0:00:16.040 and welcome to tech Stuff. I'm your host, Jonathan Strickland. 0:00:16.079 --> 0:00:19.880 I'm an executive producer with iHeart Podcasts and how the 0:00:19.920 --> 0:00:24.599 tech are you. I've often said that Google is not 0:00:25.079 --> 0:00:30.240 a search engine company. It's an advertising company. And of 0:00:30.240 --> 0:00:33.320 course these days, when I say Google, you could argue 0:00:33.320 --> 0:00:35.479 that I really mean Alpha, that you know, the parent 0:00:35.520 --> 0:00:39.199 company that owns all the various Google companies. But to 0:00:39.200 --> 0:00:43.599 be clear, I am not the only person who says this, right, 0:00:43.640 --> 0:00:45.400 I don't mean to say that I'm some sort of 0:00:45.400 --> 0:00:47.879 guru who I did not come up with it. I'm 0:00:47.920 --> 0:00:50.800 not the first person to say it. I also say it. 0:00:50.880 --> 0:00:54.320 That's all. I agree with those who say this. But 0:00:54.400 --> 0:00:57.120 here's something that I find interesting. So, once upon a time, 0:00:57.760 --> 0:01:02.360 Google was a search company. It was not an advertising company. 0:01:02.720 --> 0:01:07.440 In the very early days, Google didn't generate revenue through advertising. 0:01:08.200 --> 0:01:12.800 So let's set the clocks back quite a bit. In 0:01:12.880 --> 0:01:16.039 nineteen ninety five, a guy named Larry Page was mulling 0:01:16.080 --> 0:01:18.840 over the idea of enrolling in Stanford University in order 0:01:18.840 --> 0:01:21.120 to go to grad school. He was at the campus 0:01:21.360 --> 0:01:22.880 in order to get the lay of the land, and 0:01:22.920 --> 0:01:25.240 there was a student who was assigned to him to 0:01:25.319 --> 0:01:29.200 show him around, a Stanford student named Sarah gay Brinn, 0:01:29.440 --> 0:01:32.839 and the story goes that the two got along kind 0:01:32.880 --> 0:01:36.560 of like oil and water. They apparently didn't agree on 0:01:36.640 --> 0:01:41.480 practically anything, but a year later, when Larry Page was 0:01:41.560 --> 0:01:44.640 in the graduate program, the two decided that they would 0:01:44.680 --> 0:01:47.520 work together on what would turn out to be a 0:01:47.640 --> 0:01:52.320 truly huge project. So their goal was to build a 0:01:52.360 --> 0:01:55.680 new kind of web search engine. So at the time, 0:01:56.000 --> 0:01:59.520 the dominant search engines that were still working off the 0:01:59.720 --> 0:02:05.760 very young Worldwide Web primarily used keywords as their guide. 0:02:06.080 --> 0:02:08.360 So you would pop into a search engine. The one 0:02:08.440 --> 0:02:10.639 I used all the time was web Crawler, but there 0:02:10.639 --> 0:02:12.920 were lots of them, and you would type whatever it 0:02:12.960 --> 0:02:15.360 was you were searching for, like let's say it's the 0:02:15.360 --> 0:02:19.800 word tractor, for example. The search engine would return results 0:02:20.280 --> 0:02:24.480 of web pages that featured that keyword somewhere in the page, 0:02:24.919 --> 0:02:28.480 and they might have different ways of prioritizing the web pages. 0:02:28.480 --> 0:02:31.720 Maybe if the keyword appears in the title of the page, 0:02:31.760 --> 0:02:32.799 it might rank higher. 0:02:32.800 --> 0:02:33.480 That kind of thing. 0:02:33.919 --> 0:02:37.120 But the problem was there was never a guarantee that 0:02:37.160 --> 0:02:40.320 the search results you got would actually be relevant to 0:02:40.400 --> 0:02:44.799 the search query that you put in. They might include 0:02:45.240 --> 0:02:48.400 the search term, but it may not turn out to 0:02:48.400 --> 0:02:52.160 be useful. Maybe the page just mentions tractors off handedly, 0:02:52.600 --> 0:02:55.919 like as a joke or as a reference, and it's 0:02:55.960 --> 0:02:58.280 not actually about tractors, it just happens to have the 0:02:58.320 --> 0:03:02.000 word tractors in it. Or maybe it's a web page 0:03:02.040 --> 0:03:05.840 run by a real scumbag who just crams the bottom 0:03:05.919 --> 0:03:09.400 of the page with a ton of random keywords, ones 0:03:09.440 --> 0:03:13.240 that they've determined are really popular, and it's on an 0:03:13.280 --> 0:03:16.520 effort to capture as much traffic from search engines as possible. 0:03:16.560 --> 0:03:19.839 They don't care that this isn't what you were looking for. 0:03:20.280 --> 0:03:23.800 They just want to get you onto that page. The 0:03:23.840 --> 0:03:27.080 search experience could be really a frustrating one. It often 0:03:27.160 --> 0:03:30.280 involved users going to a page, taking a quick look, 0:03:30.320 --> 0:03:33.800 saying oh, this isn't what I need, backing out, and 0:03:33.840 --> 0:03:37.400 then doing it all over again. And it wasn't very 0:03:37.520 --> 0:03:41.200 satisfying or helpful page. And Brend wanted to create an 0:03:41.240 --> 0:03:44.200 engine that would return much better search results. 0:03:44.520 --> 0:03:45.600 So how would you do that? 0:03:45.960 --> 0:03:49.480 Well, their solution was to take a broader approach to search. 0:03:49.760 --> 0:03:53.040 Their search engine would assign a rank a page rank 0:03:53.160 --> 0:03:57.119 if you will to each result, and higher ranks would 0:03:57.160 --> 0:03:59.960 appear higher in the search results page and lots of 0:04:00.240 --> 0:04:04.320 factors would go into determining what page rank a page 0:04:04.360 --> 0:04:07.360 would receive. For example, let's say that you've got a 0:04:07.400 --> 0:04:10.120 web page and you notice that there are tons of 0:04:10.280 --> 0:04:13.760 incoming links pointing to this web page. Well, that could 0:04:13.800 --> 0:04:16.800 end up being a boost to that page's rank, because 0:04:16.839 --> 0:04:19.000 the thought is, if a lot of other web pages 0:04:19.000 --> 0:04:22.159 are pointing to this page, it must be good, right, 0:04:22.240 --> 0:04:24.440 if you've got a lot of different pages. Of course, 0:04:24.440 --> 0:04:26.360 you can game that too, right, You could create a 0:04:26.400 --> 0:04:28.880 whole network of pages that just link to each other 0:04:29.320 --> 0:04:31.400 in an effort to try and boost page eranks. So 0:04:31.480 --> 0:04:35.839 another part of this ranking system would take into account 0:04:35.839 --> 0:04:40.680 the quality of those incoming links, the idea being well, 0:04:41.200 --> 0:04:44.760 if the incoming links are coming from really established websites, 0:04:44.800 --> 0:04:48.880 maybe they belong to a known entity like an important company, 0:04:49.480 --> 0:04:51.799 or maybe it's a web page that has been around 0:04:51.800 --> 0:04:53.880 for a while and has sort of proven itself, any 0:04:53.960 --> 0:04:57.440 links that that page has are going to be worth more. 0:04:57.680 --> 0:05:00.760 Plus those pages themselves, the ones that been around or 0:05:00.800 --> 0:05:03.200 from known entities, those would have a higher rank because 0:05:03.240 --> 0:05:06.240 they were thought of as just being higher quality. Now, 0:05:06.279 --> 0:05:08.880 this is just a couple of examples of factors that 0:05:08.880 --> 0:05:11.200 would determine a web page's page erank. 0:05:11.560 --> 0:05:12.600 As it turned out, there were. 0:05:12.560 --> 0:05:16.359 Lots of them, and the formula would change frequently, and 0:05:16.400 --> 0:05:19.760 it was always kept secret so that pages couldn't try 0:05:19.760 --> 0:05:22.400 to game the system. Although they didn't stop anyone from trying, 0:05:22.920 --> 0:05:24.640 it just meant it was hard to do because you 0:05:24.680 --> 0:05:29.159 didn't exactly know which factors were the most important. The 0:05:29.200 --> 0:05:32.800 PageRank approach meant that you were far more likely as 0:05:32.839 --> 0:05:35.599 a user to come across a search result that was 0:05:35.640 --> 0:05:38.440 relevant to whatever it was you were actually searching for. 0:05:38.920 --> 0:05:39.080 Now. 0:05:39.120 --> 0:05:44.040 Originally, the pair called their search engine BackRub. I think 0:05:44.440 --> 0:05:47.039 we could all agree that it was a fortunate thing 0:05:47.080 --> 0:05:49.880 that they decided to change the name to Google. The 0:05:49.920 --> 0:05:53.000 two received a healthy one hundred thousand dollars check from 0:05:53.000 --> 0:05:57.599 an investor in nineteen ninety eight, and they launched Google Incorporated, 0:05:58.000 --> 0:06:01.640 and for the first couple of years, Google existed primarily 0:06:01.680 --> 0:06:06.360 off of influxes of cash coming from investors. Everyone could 0:06:06.360 --> 0:06:10.279 recognize that Google was providing a valuable service, one that 0:06:10.400 --> 0:06:13.599 was becoming increasingly important as the web was getting more 0:06:13.960 --> 0:06:16.880 complex and larger, and it was doing so at a 0:06:16.960 --> 0:06:20.040 level that really set Google apart from the competition. Google 0:06:20.120 --> 0:06:22.400 was quickly making a name for itself as being the 0:06:22.440 --> 0:06:25.880 best of the web search engines at that time, but 0:06:26.040 --> 0:06:28.960 investors were also starting to get a little concerned, and 0:06:29.040 --> 0:06:31.359 the reason for that was that the powerful Google search 0:06:31.360 --> 0:06:35.080 engine didn't have an equally powerful way to generate revenue. 0:06:35.480 --> 0:06:35.640 Now. 0:06:35.680 --> 0:06:38.920 Initially, one strategy was to try and form partnerships with 0:06:39.040 --> 0:06:42.600 portal sites. And these were websites that acted kind of 0:06:42.640 --> 0:06:45.080 like a homepage, like this is the page that you 0:06:45.120 --> 0:06:49.040 would see when you started up your browser. This was 0:06:49.120 --> 0:06:50.800 kind of a gateway to the rest of the web. 0:06:50.839 --> 0:06:54.080 They might host a few articles. Maybe some of those 0:06:54.160 --> 0:06:57.480 articles were written in house by an editorial staff that 0:06:57.520 --> 0:07:01.160 worked for the website. Maybe some of the articles were 0:07:01.160 --> 0:07:03.479 really generated by a third party and they were just 0:07:03.520 --> 0:07:07.400 being linked to on this landing page. And many of 0:07:07.440 --> 0:07:11.520 them would also include a search tool. So Google would 0:07:11.560 --> 0:07:15.520 try and make deals with these different sites to be 0:07:15.760 --> 0:07:18.640 that search tool. But Google with small potatoes in those days, 0:07:18.920 --> 0:07:21.200 so these deals were few and far between, and they 0:07:21.240 --> 0:07:25.200 certainly weren't enough to make Google look like it was 0:07:25.240 --> 0:07:29.680 going to be arousing success financially, So clearly Google would 0:07:29.720 --> 0:07:31.880 have to come up with a different approach if the 0:07:31.920 --> 0:07:34.800 company was to stay in business, and someone else had 0:07:34.800 --> 0:07:37.160 to come up with an idea that, with a little tweaking, 0:07:37.600 --> 0:07:42.080 would work just fine, and then Google would take that idea. 0:07:42.160 --> 0:07:45.080 To say they stole it would probably be too harsh, 0:07:45.360 --> 0:07:49.080 but they certainly emulated an idea that someone else came 0:07:49.160 --> 0:07:52.400 up with for a totally different site. Now that's someone 0:07:52.440 --> 0:07:55.840 else was a guy named William Bill T. 0:07:56.160 --> 0:07:56.640 Gross. 0:07:57.040 --> 0:07:59.920 He graduated from the California Institute of Technology with the 0:08:00.120 --> 0:08:03.440 degree in mechanical engineering, which would not make you think 0:08:03.480 --> 0:08:06.760 of someone who would go on to create a web 0:08:06.800 --> 0:08:09.560 search engine, but he became known as something of a 0:08:09.600 --> 0:08:13.800 serial entrepreneur. He would launch lots of different companies across 0:08:13.840 --> 0:08:17.720 his career as soon as he was graduating, essentially, including 0:08:17.760 --> 0:08:21.040 a video arcade and a couple of software companies, and 0:08:21.080 --> 0:08:25.200 an audio equipment company and more. But in the late nineties, 0:08:25.560 --> 0:08:28.400 Gross wanted to solve the same problem that Page and 0:08:28.440 --> 0:08:31.040 Brinn had been working on or would be working on, actually, 0:08:31.080 --> 0:08:33.920 because Gross's solution would come out first. He wanted to 0:08:34.000 --> 0:08:36.720 be able to create a search engine that would return 0:08:36.840 --> 0:08:40.240 better search results than a lot of the competitors that 0:08:40.280 --> 0:08:42.880 were on the market. But he took a different approach 0:08:42.880 --> 0:08:45.720 from what Brin and Page were looking at. So instead 0:08:45.760 --> 0:08:49.319 of building out a system meant to assess the value 0:08:49.440 --> 0:08:53.080 of every page that's on the Internet, he figured he 0:08:53.120 --> 0:08:57.280 could deliver better search results by having companies pay him 0:08:57.440 --> 0:09:00.760 in order to have their search result rank higher on 0:09:00.800 --> 0:09:03.760 the list. So here's the basic idea. Now, as I 0:09:03.800 --> 0:09:06.400 mentioned earlier, one of the big problems with early search 0:09:06.400 --> 0:09:09.360 engines is that a lot of web administrators were trying 0:09:09.400 --> 0:09:12.240 to game the system. So it really wasn't unusual for 0:09:12.360 --> 0:09:16.400 web designer to just stuff a page full of keywords 0:09:16.400 --> 0:09:20.720 that didn't actually apply to the page's content itself. So 0:09:20.840 --> 0:09:23.160 clever designers would do this in a way that wasn't 0:09:23.200 --> 0:09:25.720 obvious to the visitor, right because otherwise you would just 0:09:25.840 --> 0:09:28.080 like I said, you would scroll down a page and 0:09:28.080 --> 0:09:31.000 you see like a word jumble of a bunch of 0:09:31.240 --> 0:09:34.400 unrelated words at the end of a page, and it 0:09:34.480 --> 0:09:39.160 just looked like a mess. Clever web administrators would choose 0:09:39.200 --> 0:09:41.880 to make the text the same color as the page's 0:09:41.920 --> 0:09:44.439 background at the bottom, so you wouldn't even see it 0:09:44.480 --> 0:09:46.959 like it might be that the page seems unusually long 0:09:47.040 --> 0:09:50.080 with no content toward the bottom of it, but in 0:09:50.120 --> 0:09:52.520 fact there would be content. It's just it was the 0:09:52.559 --> 0:09:55.440 same color as the background, so you couldn't see it. 0:09:55.720 --> 0:09:57.240 However, while you. 0:09:57.160 --> 0:10:00.600 Couldn't see it, a search engine could. The search engine 0:10:00.600 --> 0:10:03.200 would still see the keywords there even though you wouldn't, 0:10:03.320 --> 0:10:06.320 and that's why this page that doesn't seem to be 0:10:06.360 --> 0:10:08.319 related at all to the thing you were searching for 0:10:08.360 --> 0:10:12.240 would pop up in search results. Meanwhile, you had legitimate 0:10:12.320 --> 0:10:16.600 sites out there, some connected to established brands, that would 0:10:16.720 --> 0:10:19.079 very much like to rank high in search in order 0:10:19.080 --> 0:10:21.400 to drive more traffic to their sites, but with all 0:10:21.440 --> 0:10:24.600 these junk sites out there, that could be hard to do. 0:10:25.120 --> 0:10:27.920 So they might be in a position where they're able 0:10:27.960 --> 0:10:30.600 to and willing to pay to jump to the head 0:10:30.600 --> 0:10:32.760 of the line if it meant that folks would actually 0:10:32.840 --> 0:10:37.320 find their stuff online. And so Gross launched a company 0:10:37.360 --> 0:10:42.640 called go to dot com Incorporated. His strategy included a 0:10:42.760 --> 0:10:46.840 pay per click arrangement, which is exactly what it sounds like. 0:10:47.120 --> 0:10:50.560 The advertiser agrees to pay a certain amount every time 0:10:50.640 --> 0:10:55.320 someone clicks on the ad, so typically you aggregate clicks 0:10:55.360 --> 0:10:58.079 by the thousands, so each click is really just worth 0:10:58.160 --> 0:11:02.280 you usually fractions of a penny. Gross's logic was that 0:11:02.400 --> 0:11:05.960 the garbage sites out there couldn't and wouldn't pay for 0:11:06.080 --> 0:11:10.240 positioning because their entire strategy relied on using keywords in 0:11:10.360 --> 0:11:13.600 bulk that had no application to the actual site. So 0:11:13.800 --> 0:11:15.800 if they wanted to try and do that the same 0:11:15.840 --> 0:11:18.480 way with Gross's strategy, they would have to pay for 0:11:18.600 --> 0:11:21.280 all these different keywords that didn't actually apply to them, 0:11:21.800 --> 0:11:24.440 and it would bankrupt them because people would be clicking 0:11:24.520 --> 0:11:27.400 on those sites over and over again on the search results, 0:11:27.559 --> 0:11:30.880 and they'd have to pay every time that happened, or 0:11:30.920 --> 0:11:34.240 at least every time every thousand times it happened. So 0:11:35.200 --> 0:11:37.040 you would be paying go To for all this traffic. 0:11:37.080 --> 0:11:39.360 You wouldn't be making that money back through the ad 0:11:39.400 --> 0:11:42.600 revenue you generated on your site. There's no way you'd 0:11:42.640 --> 0:11:46.000 be making more money through say, web advertising, than you 0:11:46.040 --> 0:11:48.640 were paying out to go To. So whamo, you end 0:11:48.720 --> 0:11:51.439 up with a search engine that delivers better results, or 0:11:51.440 --> 0:11:54.599 at least more relevant results. Since you can't guarantee that 0:11:54.640 --> 0:11:57.280 the entities that actually do pay for their positioning are 0:11:57.320 --> 0:11:59.520 in fact the best ones out there, but chances are 0:11:59.559 --> 0:12:03.079 there better than the junk sites. Gross launched goo to 0:12:03.240 --> 0:12:05.760 dot com in nineteen ninety eight. In fact, he launched 0:12:05.760 --> 0:12:08.120 it before Page and brend got Google off the ground. 0:12:08.320 --> 0:12:10.800 And while go to only saw modest success in the 0:12:10.800 --> 0:12:14.720 grand scheme of things, that revenue generation model stood out, 0:12:15.200 --> 0:12:18.360 and when it came time to determine a way for 0:12:18.520 --> 0:12:22.920 Google to make money, the go to became well, well, 0:12:22.920 --> 0:12:28.000 it became the go to model. So Google would introduce 0:12:28.040 --> 0:12:31.400 a pair of strategies in two thousand that would bring 0:12:31.440 --> 0:12:35.280 the company into the advertising space. Now, according to a 0:12:35.400 --> 0:12:39.560 Google SEC filing from two thousand and four, quote, in 0:12:39.600 --> 0:12:42.800 the first quarter of two thousand, we introduced our first 0:12:42.840 --> 0:12:47.599 advertising program through our direct sales force. We offered advertisers 0:12:47.600 --> 0:12:51.160 the ability to place text based ads on our websites 0:12:51.400 --> 0:12:54.600 targeted to our users search queries under a program called 0:12:54.640 --> 0:12:58.480 Premium Sponsorships. Advertisers paid us based on the number of 0:12:58.559 --> 0:13:01.760 times their ads were displayed on users search results pages, 0:13:02.080 --> 0:13:05.280 and we recognized revenue at the time these ads appeared 0:13:05.640 --> 0:13:06.320 end quote. 0:13:06.760 --> 0:13:08.079 So with this. 0:13:08.080 --> 0:13:11.200 Strategy, Google would collect a certain amount of money every 0:13:11.240 --> 0:13:15.160 time someone saw a particular ad in their search results. 0:13:15.400 --> 0:13:18.600 Clicking on the ad wasn't necessarily a requirement. Just getting 0:13:18.600 --> 0:13:20.760 the ad in front of folks was enough to warrant 0:13:20.920 --> 0:13:24.360 to the charge of the ad. As for how much 0:13:24.400 --> 0:13:27.920 the advertisers were actually paying, I saw one site where 0:13:27.960 --> 0:13:31.000 a client said their rate was around forty dollars per 0:13:31.080 --> 0:13:36.400 one thousand impressions or views, so every thousand times this 0:13:36.520 --> 0:13:40.320 ad was seen they would pay out forty dollars. And 0:13:40.400 --> 0:13:43.360 Google was rapidly rising in popularity, and when you think 0:13:43.400 --> 0:13:46.080 of all the different potential ads the company could sell 0:13:46.120 --> 0:13:50.840 against different search terms, it really starts to add up. 0:13:51.679 --> 0:13:55.120 The premium sponsorship ads would appear at the top of 0:13:55.200 --> 0:13:58.200 the search results, so the very first search result would 0:13:58.200 --> 0:14:01.559 actually be an AD, so there was a decent chance 0:14:01.640 --> 0:14:04.880 that unobservant folks would just click on the ad, assuming 0:14:04.880 --> 0:14:06.920 it was the best result for their query. Because it 0:14:07.000 --> 0:14:10.040 was at the top of the search results page. Google 0:14:10.120 --> 0:14:13.760 did have the foresight to place a shaded background behind 0:14:13.880 --> 0:14:17.960 the link, which set it apart from the actual search results. 0:14:18.640 --> 0:14:21.040 I suppose folks of the company understood that unless they 0:14:21.040 --> 0:14:24.960 made the effort to differentiate the paid advertising spot from 0:14:25.000 --> 0:14:28.560 the legitimate return results, they could get into some real 0:14:28.640 --> 0:14:33.359 trouble there with regulators like the FTC, which absolutely demanded 0:14:33.360 --> 0:14:38.600 that search engine companies make differentiation between sponsored results and 0:14:38.680 --> 0:14:39.480 search results. 0:14:40.000 --> 0:14:41.760 All Right, we're gonna talk. 0:14:41.600 --> 0:14:45.720 More about how Google did that, plus other strategies Google 0:14:45.800 --> 0:14:48.600 got into in the advertising space, but first we actually 0:14:48.680 --> 0:14:51.320 need to take a break for our own advertisers. 0:14:51.600 --> 0:15:02.280 We'll be right back, okay. 0:15:02.360 --> 0:15:05.680 So when we broke off for ads, we were just 0:15:05.720 --> 0:15:11.800 talking about how Google needed to separate their sponsored results 0:15:11.840 --> 0:15:14.800 from their actual search results, and to follow along that 0:15:15.040 --> 0:15:16.040 for a second. 0:15:16.160 --> 0:15:17.480 This is going to just be a little bit of 0:15:17.520 --> 0:15:18.040 a tangent. 0:15:18.440 --> 0:15:21.320 Over the years, Google has changed up how it tags 0:15:21.440 --> 0:15:24.880 or labels ads versus search results. So example, in two 0:15:24.880 --> 0:15:27.360 thousand and seven, the company changed the background color from 0:15:27.400 --> 0:15:31.640 being blue to yellow. In twenty ten, it went to 0:15:31.720 --> 0:15:35.040 a very kind of light green background. Actually in the 0:15:35.080 --> 0:15:37.680 example I saw it looked more teeled to me. But 0:15:37.720 --> 0:15:40.040 then it went back to yellow. And going back to 0:15:40.120 --> 0:15:42.560 yellow became kind of a thing. Over the following years, 0:15:42.560 --> 0:15:45.480 Google would experiment and changed to different colors, but would 0:15:45.480 --> 0:15:48.520 come back to yellow over and over again. In twenty thirteen, 0:15:48.960 --> 0:15:51.680 Google went with a very pale yellow, which, depending upon 0:15:52.080 --> 0:15:53.920 how your settings are on your mind or you may 0:15:53.920 --> 0:15:56.440 not even be able to see, but that same year. 0:15:56.520 --> 0:15:59.760 That was when the US Federal Trade Commission really issued 0:15:59.760 --> 0:16:02.720 a war learning to search engines in general about a 0:16:02.840 --> 0:16:07.840 quote decline in compliance end quote with those rules that 0:16:08.040 --> 0:16:10.120 all search engines were supposed to follow, that they were 0:16:10.160 --> 0:16:13.120 supposed to make it clear that certain entries were ads 0:16:13.200 --> 0:16:16.880 versus legitimate search results. So Google would end up placing 0:16:16.920 --> 0:16:21.160 an AD tag next to ads in search results to 0:16:21.200 --> 0:16:24.840 make it more clear that they were in fact advertisings. However, 0:16:24.920 --> 0:16:28.520 this AD tag actually appeared below the hyperlink text, so 0:16:28.560 --> 0:16:31.480 you'd get the hyperlink text first, and then that little description, 0:16:31.640 --> 0:16:34.040 that block of text and stuff that would describe the 0:16:34.120 --> 0:16:36.480 link that appears below it. Next to that, you would 0:16:36.520 --> 0:16:39.560 have this little AD tag. The company switched up how 0:16:39.600 --> 0:16:42.480 that looked a few times. They would use different shaded 0:16:42.520 --> 0:16:45.320 backgrounds on the tag and whatnot in an effort to 0:16:45.320 --> 0:16:48.760 comply with the rules, while you know, not complying too hard, 0:16:49.080 --> 0:16:52.680 because ideally you still get people clicking on those ads, 0:16:53.080 --> 0:16:56.000 and it's easier to do that if the people don't 0:16:56.040 --> 0:16:58.080 notice at first that it is an AD, which has 0:16:58.160 --> 0:17:00.160 certainly happened to me. I know, I've been guilty of 0:17:00.160 --> 0:17:02.880 it a few times. So I did a search just 0:17:03.000 --> 0:17:06.120 now to see what looks like these days, And depending 0:17:06.119 --> 0:17:09.040 on what I'm searching for, it may bring up different things. 0:17:09.080 --> 0:17:11.560 Like I searched for a refrigerator, and in that case, 0:17:11.960 --> 0:17:14.920 it brought up a bunch of sponsored links to where 0:17:14.960 --> 0:17:17.960 I could buy a new refrigerator. But then I did 0:17:18.000 --> 0:17:22.119 a different search and what I got were a couple 0:17:22.160 --> 0:17:25.320 of search results at the top that had the label 0:17:25.480 --> 0:17:29.680 sponsored appearing before the hyperlink. So you had a little 0:17:29.760 --> 0:17:33.080 highlighted sponsored and then the hyperlink and then the text. 0:17:33.119 --> 0:17:36.120 And I feel like that's fairly clear that it's an 0:17:36.200 --> 0:17:39.760 advertisement as opposed to just a regular search result. So 0:17:40.040 --> 0:17:43.359 props to Google in that case, because I think that's 0:17:43.400 --> 0:17:46.920 a reasonable approach. It's not beating you over the head 0:17:46.920 --> 0:17:50.800 with it, but I feel like it's noticeable. Anyway, Premium 0:17:50.880 --> 0:17:55.199 sponsorships really started Google's ad revenue journey, but as mentioned, 0:17:55.200 --> 0:17:58.760 Google's direct sales team was responsible at this time for 0:17:58.800 --> 0:18:02.120 selling that adspace, and that means that the sales team 0:18:02.240 --> 0:18:05.640 was reaching out to advertisers with this offer and trying 0:18:05.680 --> 0:18:08.280 to convince them to agree to plunk down cash in 0:18:08.320 --> 0:18:12.080 return for thousands of impressions per ad So it was 0:18:12.119 --> 0:18:15.280 a strong start, but not nearly the strategy that would 0:18:15.280 --> 0:18:18.480 propel Google into becoming one of the largest, most powerful 0:18:18.520 --> 0:18:22.159 tech companies out there. Instead, that development would wait a 0:18:22.160 --> 0:18:25.960 whole two or three months because that same year, in 0:18:26.000 --> 0:18:30.360 two thousand, Google would not just introduce premium sponsorships, they 0:18:30.400 --> 0:18:33.520 also introduced a little offering that at the time was 0:18:33.560 --> 0:18:39.840 called AdWords. So unlike premium sponsorships, prospective advertisers could actually 0:18:39.880 --> 0:18:42.879 sign up for AdWords at a link, you know, self 0:18:42.920 --> 0:18:45.919 service style. They didn't have to be reached by a 0:18:45.960 --> 0:18:49.240 direct sales representative. They would just go to a link 0:18:49.280 --> 0:18:51.840 and say, okay, yes, I want to purchase some AdWords. 0:18:52.119 --> 0:18:53.120 And that wasn't the only. 0:18:52.920 --> 0:18:56.760 Difference between AdWords and premium sponsorships. Another was that adword 0:18:56.800 --> 0:19:00.720 results appeared not at the top of search results. Instead, 0:19:00.720 --> 0:19:03.760 it would appear to the right of search results. Originally 0:19:03.960 --> 0:19:06.080 in what we used to call, or still call, I guess, 0:19:06.119 --> 0:19:08.600 the right hand rail. I just don't hear much about 0:19:08.600 --> 0:19:11.359 it anymore because I'm not in those meetings anymore, which 0:19:11.400 --> 0:19:14.159 is fine by me. But the right hand rail is 0:19:14.280 --> 0:19:16.200 you can think of that as like the right margin 0:19:16.480 --> 0:19:19.679 of a web page. So in the old web banner days, 0:19:19.760 --> 0:19:23.040 not that web banners are gone, but they certainly aren't 0:19:23.080 --> 0:19:26.240 the dominant form of advertising anymore. In the old web 0:19:26.280 --> 0:19:28.640 banner days, there were a few places where you would 0:19:28.640 --> 0:19:32.160 typically put a web banner, and you might charge different 0:19:32.200 --> 0:19:35.720 amounts depending upon the location. So the really popular one 0:19:35.800 --> 0:19:37.560 was at the top of the page, right like at 0:19:37.560 --> 0:19:39.000 the very top of the page you have a web 0:19:39.160 --> 0:19:42.639 banner ad that advertises for something, But another would be 0:19:42.680 --> 0:19:45.080 in the margins, either along the left or the right. 0:19:45.640 --> 0:19:47.640 Very rarely you might put one at the bottom. Those 0:19:47.680 --> 0:19:50.800 were often the least valuable ones, or at least the 0:19:51.000 --> 0:19:54.879 least expensive ones, because the common thought was that people 0:19:54.920 --> 0:19:56.399 just didn't scroll all the way to the bottom of 0:19:56.440 --> 0:20:00.040 the page. But anyway, the right rail is where the 0:20:00.080 --> 0:20:03.920 adword results originally would appear, So starting in August two 0:20:03.960 --> 0:20:06.600 thousand or so, searching for something on Google would bring 0:20:06.680 --> 0:20:09.840 up premium sponsorship ads at the top of your results 0:20:10.000 --> 0:20:12.760 and a couple of AdWords ads set off to the right. 0:20:13.200 --> 0:20:17.080 The AdWords ads also had a sponsored link label on them, 0:20:17.160 --> 0:20:20.360 so they did indicate to the user that they were advertising. 0:20:20.400 --> 0:20:23.959 They were not an official search result, so that label 0:20:24.200 --> 0:20:27.200 was there from the beginning. From the advertiser side, buying 0:20:27.240 --> 0:20:31.280 AdWords was pretty straightforward. The advertiser identified which keyword or 0:20:31.359 --> 0:20:36.040 keywords they wanted, They then submitted the ad and bammed. 0:20:36.080 --> 0:20:39.280 Their ad would show up wherever anyone searched for that 0:20:39.320 --> 0:20:41.960 particular keyword. It was pretty simple stuff, and of course 0:20:41.960 --> 0:20:45.000 they also paid a certain amount of money. They were 0:20:45.080 --> 0:20:47.320 kind of doing a pay per click as opposed to 0:20:47.760 --> 0:20:51.000 pay per impression, meaning that in this case, it wasn't 0:20:51.119 --> 0:20:53.560 enough that someone saw the ad, they needed to have 0:20:53.680 --> 0:20:56.680 clicked on it for it to count toward how much 0:20:56.760 --> 0:21:00.879 Google would charge the advertiser. Things held steady for a 0:21:00.920 --> 0:21:04.560 couple of years, but in two thousand and three, Google 0:21:04.680 --> 0:21:07.280 made a decision that would shake things up. They decided 0:21:07.320 --> 0:21:12.720 to phase out premium sponsorships and actually replace them with AdWords, 0:21:13.480 --> 0:21:16.840 and the direct sales angle wouldn't be necessary anymore, at 0:21:16.920 --> 0:21:20.159 least not for this, and AdWords would migrate from the 0:21:20.200 --> 0:21:24.679 right hand rail to being above the search results. So essentially, 0:21:24.680 --> 0:21:27.159 they were saying the AdWords approach seemed to be the 0:21:27.200 --> 0:21:30.600 most successful, most lucrative, the one that was giving the 0:21:30.640 --> 0:21:34.760 most satisfaction to advertisers as well as to Google, so 0:21:35.080 --> 0:21:36.639 that was what they were going to switch to. They 0:21:36.640 --> 0:21:39.280 were going to make that the dominant form of ad 0:21:39.320 --> 0:21:44.359 revenue generation. In addition, adwards clients would receive charges based 0:21:44.359 --> 0:21:46.879 on those number of clicks the ads got, Like I said, 0:21:46.880 --> 0:21:49.359 not the number of impressions, which is nice, right, because 0:21:49.760 --> 0:21:53.760 someone just seeing your ad that might not mean very much, 0:21:54.200 --> 0:21:56.720 but if they click on your ad, that's a much 0:21:56.760 --> 0:22:00.240 bigger deal. However, this also meant that Google would like 0:22:00.440 --> 0:22:04.199 the cost of those ads. They wouldn't charge that forty 0:22:04.280 --> 0:22:08.800