WEBVTT 00:00:00.000 --> 00:00:02.720 Imagine just for a second that you are the CEO 00:00:02.720 --> 00:00:06.750 of a massive multinational company. OK. Or maybe 00:00:06.750 --> 00:00:09.009 you're an A -list celebrity, right? You just 00:00:09.009 --> 00:00:11.429 starred in the biggest summer blockbuster of 00:00:11.429 --> 00:00:13.750 the year. You wake up, you pour your morning 00:00:13.750 --> 00:00:16.829 coffee and you sit down with one burning question. 00:00:17.129 --> 00:00:19.390 What is everyone saying about me? Exactly. What 00:00:19.390 --> 00:00:21.969 is the entire world saying about me today? Now, 00:00:21.989 --> 00:00:24.370 you obviously don't just you don't pull out your 00:00:24.370 --> 00:00:26.649 phone and Google yourself. No, you don't have 00:00:26.649 --> 00:00:28.010 the time for that. You don't have the time. And 00:00:28.010 --> 00:00:30.850 honestly, the sheer volume of information out 00:00:30.850 --> 00:00:33.229 there would be completely paralyzing. You'd never. 00:00:33.289 --> 00:00:36.250 Right. It's just too much data. So instead, you 00:00:36.250 --> 00:00:40.329 rely on an invisible, massive, and highly sophisticated 00:00:40.329 --> 00:00:44.390 industry to do that scrolling for you. Welcome 00:00:44.390 --> 00:00:48.909 to today's deep dive. We are exploring the hidden, 00:00:48.969 --> 00:00:52.170 surprisingly high -stakes world of media monitoring 00:00:52.170 --> 00:00:54.149 services. Or if you want to use the historical 00:00:54.149 --> 00:00:56.030 term, press clipping bureaus. Press clipping 00:00:56.030 --> 00:01:00.789 bureaus. I love that. It is a massive, invisible 00:01:00.789 --> 00:01:03.780 infrastructure. Right. And what makes this topic 00:01:03.780 --> 00:01:06.260 so compelling is that while the underlying technology 00:01:06.260 --> 00:01:09.239 has completely transformed, I mean, we're talking 00:01:09.239 --> 00:01:13.040 a shift from literal scissors and glue to highly 00:01:13.040 --> 00:01:15.760 complex artificial intelligence driven scraping 00:01:15.760 --> 00:01:19.379 algorithms. The core human desire driving it 00:01:19.379 --> 00:01:21.680 hasn't changed a bit. It really hasn't. We all 00:01:21.680 --> 00:01:24.840 have this inherent need to understand. ideally 00:01:24.840 --> 00:01:27.439 control our own narrative out in the wild. Okay, 00:01:27.500 --> 00:01:30.379 let's unpack this. For this deep dive, our map 00:01:30.379 --> 00:01:33.280 is the Wikipedia article detailing the history, 00:01:33.519 --> 00:01:35.640 the technological leaps, and the rather intense 00:01:35.640 --> 00:01:38.299 legal battles surrounding media monitoring. Yes, 00:01:38.420 --> 00:01:40.319 the legal side gets very interesting. Oh, it 00:01:40.319 --> 00:01:42.400 really does. And our mission today is to trace 00:01:42.400 --> 00:01:44.540 this evolution. We are going to track how the 00:01:44.540 --> 00:01:47.340 flow of information went from 19th century Parisian 00:01:47.340 --> 00:01:50.180 stage actors buying physical paper cutouts of 00:01:50.180 --> 00:01:52.219 their own reviews. Which is such a great story. 00:01:52.400 --> 00:01:54.280 Right, all the way to the modern... AI -driven 00:01:54.280 --> 00:01:57.180 media intelligence bots of today. Bots that we 00:01:57.180 --> 00:01:59.000 should probably add right up front are currently 00:01:59.000 --> 00:02:01.439 sparking international copyright lawsuits. It 00:02:01.439 --> 00:02:03.439 is quite the journey. We're really looking at 00:02:03.439 --> 00:02:06.400 a fundamental shift in how human beings interact 00:02:06.400 --> 00:02:09.639 with mass media. And as we dig into this, you'll 00:02:09.639 --> 00:02:12.180 see how the methods we use to track information 00:02:12.180 --> 00:02:14.900 actually end up shaping the architecture of the 00:02:14.900 --> 00:02:16.719 information itself. I want to start with the 00:02:16.719 --> 00:02:19.699 basics, though. If we strip away the modern tech 00:02:19.699 --> 00:02:23.620 turk, What exactly is a media monitoring service 00:02:23.620 --> 00:02:26.060 at its core? Because it seems like it could mean 00:02:26.060 --> 00:02:28.159 a lot of different things depending on who is 00:02:28.159 --> 00:02:30.419 paying the bill. At its foundation, it's simply 00:02:30.419 --> 00:02:32.419 a service that provides clients with copies of 00:02:32.419 --> 00:02:34.919 media content that is of specific interest to 00:02:34.919 --> 00:02:37.280 them. Simple enough. But you hit the nail on 00:02:37.280 --> 00:02:39.860 the head regarding the demand. The service has 00:02:39.860 --> 00:02:42.900 to be incredibly malleable. For one client, it 00:02:42.900 --> 00:02:45.419 might be providing raw documentation of every 00:02:45.419 --> 00:02:47.479 single time their company is mentioned. Just 00:02:47.479 --> 00:02:50.330 a data dump. Exactly. But for another, it might 00:02:50.330 --> 00:02:53.069 be a detailed statistical analysis of their brand 00:02:53.069 --> 00:02:55.650 sentiment. It could involve tracking editorial 00:02:55.650 --> 00:02:58.990 opinion across an entire geographic region, or 00:02:58.990 --> 00:03:01.629 simply keeping tabs on what one specific competitor 00:03:01.629 --> 00:03:04.120 is doing in trade journals. And it's not just 00:03:04.120 --> 00:03:07.340 PR news, is it? I imagine advertisers are heavily 00:03:07.340 --> 00:03:09.419 reliant on this kind of tracking as well. Absolutely. 00:03:09.520 --> 00:03:11.159 Tracking advertising content, you know, where 00:03:11.159 --> 00:03:13.659 it's placed, how often it runs, what the competitors 00:03:13.659 --> 00:03:15.979 are spending. That is a huge sector of this industry. 00:03:16.300 --> 00:03:19.400 But to really understand how massive this undertaking 00:03:19.400 --> 00:03:21.960 is today, we have to look at how it started. 00:03:22.180 --> 00:03:25.219 The analog era. Right. And that means going back 00:03:25.219 --> 00:03:27.840 to a time when information was entirely physical. 00:03:28.840 --> 00:03:32.219 We are talking about the mid to late 19th century. 00:03:32.810 --> 00:03:35.229 Paint a picture for us. What does the media landscape 00:03:35.229 --> 00:03:37.710 look like at this point? And who actually realizes 00:03:37.710 --> 00:03:40.270 there's a business opportunity here? Well, put 00:03:40.270 --> 00:03:42.930 yourself in London in the 1850s. Mass media is 00:03:42.930 --> 00:03:45.490 essentially limited to print. But the volume 00:03:45.490 --> 00:03:47.289 of that print is suddenly exploding. Because 00:03:47.289 --> 00:03:49.229 of the telegraph, right? Yes, you have the invention 00:03:49.229 --> 00:03:51.289 of the telegraph, playing of submarine cables. 00:03:51.569 --> 00:03:53.969 News that used to take weeks to arrive by ship 00:03:53.969 --> 00:03:56.669 is now crossing oceans in minutes. Consequently, 00:03:56.669 --> 00:03:58.810 newspapers are just booming. Right. There is 00:03:58.810 --> 00:04:01.750 so much being printed that no single person or 00:04:01.750 --> 00:04:04.490 even a dedicated team at a company could possibly 00:04:04.490 --> 00:04:06.930 read it all. It's just an overwhelming flood 00:04:06.930 --> 00:04:10.400 of paper. Exactly. And that overwhelming flood 00:04:10.400 --> 00:04:13.620 is what prompted a man named Henry Romike and 00:04:13.620 --> 00:04:16.500 a news dealer named Curtis to establish the very 00:04:16.500 --> 00:04:19.740 first press clipping agency in London in 1852. 00:04:20.300 --> 00:04:22.220 So they essentially looked at this mountain of 00:04:22.220 --> 00:04:25.279 broadsheets and realized people would pay to 00:04:25.279 --> 00:04:27.560 have someone else do the reading. Precisely. 00:04:27.560 --> 00:04:29.699 But the story that I find absolutely hilarious 00:04:29.699 --> 00:04:32.720 happens a little later over in France. 1879, 00:04:32.860 --> 00:04:36.060 Paris. A guy named Alfred Cherie starts an agency 00:04:36.060 --> 00:04:38.620 called Largousse de la Presse. And his target 00:04:38.620 --> 00:04:41.360 demographic wasn't big business. No, his business 00:04:41.360 --> 00:04:44.300 model was brilliant and it leaned entirely into 00:04:44.300 --> 00:04:47.220 human ego. He offered a press clipping service 00:04:47.220 --> 00:04:49.860 specifically targeted at Parisian actors. Which 00:04:49.860 --> 00:04:52.459 makes perfect sense. Think about it. You are 00:04:52.459 --> 00:04:55.180 the star of a play in Paris. You want to read 00:04:55.180 --> 00:04:57.220 your reviews, but you don't want to buy 10 different 00:04:57.220 --> 00:04:59.360 bulky newspapers every morning and comb through 00:04:59.360 --> 00:05:02.120 pages of politics and grain prices just to find 00:05:02.120 --> 00:05:04.740 the one paragraph about your performance. Cherie's 00:05:04.740 --> 00:05:07.319 pitch was essentially, pay me and I will hand 00:05:07.319 --> 00:05:09.620 you a neat little stack of paper that is entirely 00:05:09.620 --> 00:05:12.509 about you. It's the 19th century equivalent of 00:05:12.509 --> 00:05:14.550 doom -scrolling your own Twitter mentions. That 00:05:14.550 --> 00:05:17.610 is exactly what it was, a pure vanity service. 00:05:18.189 --> 00:05:21.329 And that vanity aspect was the engine that kick 00:05:21.329 --> 00:05:24.290 -started the early industry. Initially, these 00:05:24.290 --> 00:05:26.250 clipping services catered almost exclusively 00:05:26.250 --> 00:05:29.949 to actors, wealthy tycoons, and socialites. It 00:05:29.949 --> 00:05:33.290 was entirely about personal ego, public image, 00:05:33.430 --> 00:05:36.269 and high society gossip. But obviously a business 00:05:36.269 --> 00:05:39.470 model that good doesn't stay confined to Parisian 00:05:39.470 --> 00:05:42.350 theaters. How fast did this spread? Very quickly. 00:05:42.730 --> 00:05:45.449 By 1885, the National Press Intelligence Company 00:05:45.449 --> 00:05:47.550 launched in New York. Just a few years later. 00:05:47.670 --> 00:05:50.189 Yeah. And by 1899, you have over a dozen of these 00:05:50.189 --> 00:05:51.870 clipping services operating across the United 00:05:51.870 --> 00:05:53.910 States alone. What's really interesting is that 00:05:53.910 --> 00:05:55.569 they quickly realized they couldn't possibly 00:05:55.569 --> 00:05:57.970 buy every local paper in the country. So they 00:05:57.970 --> 00:05:59.970 actually formed a cooperative network. Oh, that's 00:05:59.970 --> 00:06:02.430 smart. A bureau in New York would trade clippings 00:06:02.430 --> 00:06:04.709 with a bureau in Chicago to increase their geographic 00:06:04.709 --> 00:06:07.579 range. But the clientele must have evolved right. 00:06:07.800 --> 00:06:10.360 You can't build a massive global industry just 00:06:10.360 --> 00:06:13.360 by feeding the egos of theater actors and tycoons. 00:06:13.560 --> 00:06:16.540 When do the real heavy hitters enter the picture? 00:06:16.740 --> 00:06:19.660 The shift happens as the 20th century progresses. 00:06:20.199 --> 00:06:23.480 By the 1930s, the primary users of these services 00:06:23.480 --> 00:06:26.629 underwent a dramatic change. The vanity crowd 00:06:26.629 --> 00:06:28.889 was still there, but the bulk of the subscriptions 00:06:28.889 --> 00:06:31.350 were now coming from big businesses. Makes sense. 00:06:31.529 --> 00:06:33.970 Government agencies started subscribing to track 00:06:33.970 --> 00:06:37.279 public sentiment on policies. Even other newspapers 00:06:37.279 --> 00:06:39.379 subscribed to these bureaus just to track what 00:06:39.379 --> 00:06:41.360 their rival papers in different cities were reporting. 00:06:41.560 --> 00:06:43.879 It became foundational corporate intelligence. 00:06:44.360 --> 00:06:46.699 And the consolidation of that power was staggering. 00:06:46.920 --> 00:06:50.480 By 1932, just two companies, the Romike Company, 00:06:50.660 --> 00:06:53.360 which had expanded from London, and Luce's Press 00:06:53.360 --> 00:06:57.100 Clipping Bureau, they monopolized 80 % of the 00:06:57.100 --> 00:07:00.399 entire U .S. market. 80%. That is a massive amount 00:07:00.399 --> 00:07:02.360 of intelligence flowing through just two corporate 00:07:02.360 --> 00:07:05.000 bottlenecks. It is. Which brings us to the actual 00:07:05.000 --> 00:07:07.220 physical logistics of this operation. Because 00:07:07.220 --> 00:07:09.399 when we say clipping business, we mean that completely 00:07:09.399 --> 00:07:11.500 literally. Right. I want you to really visualize 00:07:11.500 --> 00:07:13.699 this for a second. Imagine you are running a 00:07:13.699 --> 00:07:16.699 business in the 1930s. You want to know if your 00:07:16.699 --> 00:07:19.740 new product is landing well in the Midwest. Today, 00:07:19.860 --> 00:07:22.660 you would perform a treater plus F keyword search 00:07:22.660 --> 00:07:25.199 on your computer and software would aggregate 00:07:25.199 --> 00:07:28.480 thousands of articles in a millisecond. In the 00:07:28.480 --> 00:07:31.120 1930s, you were relying on a room full of human 00:07:31.120 --> 00:07:33.720 beings armed with literal scissors and paste. 00:07:34.319 --> 00:07:36.899 Walk us through how this massive human search 00:07:36.899 --> 00:07:40.540 engine actually functioned. It was a highly regimented, 00:07:40.560 --> 00:07:44.220 factory -like operation with a very strict division 00:07:44.220 --> 00:07:47.160 of labor. If you walked into one of these bureaus, 00:07:47.180 --> 00:07:48.620 the first thing you'd see is the reading department. 00:07:49.439 --> 00:07:52.019 Women were primarily employed to sit at large 00:07:52.019 --> 00:07:54.519 desks, meticulously scanning through mountains 00:07:54.519 --> 00:07:57.480 of daily periodicals, newspapers, and trade journals. 00:07:57.680 --> 00:08:00.019 They had to memorize what exactly? Hundreds of 00:08:00.019 --> 00:08:02.920 client name specific brands and keywords. When 00:08:02.920 --> 00:08:04.920 they spotted a mention, they would mark it with 00:08:04.920 --> 00:08:07.240 a colored pencil. And that's just step one. What 00:08:07.240 --> 00:08:10.040 happens to the marked paper? The marked periodicals 00:08:10.040 --> 00:08:11.699 were then passed to the next department, which 00:08:11.699 --> 00:08:14.480 was typically staffed by men. Their entire job 00:08:14.480 --> 00:08:17.379 was to physically take shears, cut out the marked 00:08:17.379 --> 00:08:19.480 articles, and paste them onto specific slips 00:08:19.480 --> 00:08:22.360 of paper. And these slips had to be labeled so 00:08:22.360 --> 00:08:24.860 they didn't lose the context, right? Yes. These 00:08:24.860 --> 00:08:27.040 slips were pre -stamped with the date and the 00:08:27.040 --> 00:08:29.220 publication's name so the context wasn't lost. 00:08:29.600 --> 00:08:31.879 Imagine the sensory experience of that room. 00:08:32.340 --> 00:08:35.240 The constant rustling of giant broadsheet pages, 00:08:35.399 --> 00:08:38.259 the nonstop snipping of hundreds of scissors, 00:08:38.559 --> 00:08:41.620 the heavy smell of the glue. It was industrial 00:08:41.620 --> 00:08:44.360 -scale reading. And after the pasting, the slips 00:08:44.360 --> 00:08:46.740 went back to another department of women who 00:08:46.740 --> 00:08:49.139 would sort these individual clippings into specific 00:08:49.139 --> 00:08:52.500 client batches, package them into manila envelopes, 00:08:52.500 --> 00:08:54.460 and physically mail them out. It's just wild 00:08:54.460 --> 00:08:56.740 to think about. You are waiting days, sometimes 00:08:56.740 --> 00:08:59.440 weeks, for a physical envelope to arrive in the 00:08:59.440 --> 00:09:01.559 mail just to know what happened in the world 00:09:01.559 --> 00:09:03.919 last Tuesday. And that highly manual physical 00:09:03.919 --> 00:09:07.019 process remained the standard for decades. It 00:09:07.019 --> 00:09:09.019 was entirely constrained by the physical nature 00:09:09.019 --> 00:09:11.779 of print media. Until it wasn't. Exactly. Yeah. 00:09:11.919 --> 00:09:14.039 As we know, the media landscape was about to 00:09:14.039 --> 00:09:16.179 undergo a seismic shift. Here's where it gets 00:09:16.179 --> 00:09:18.340 really interesting. You move further into the 00:09:18.340 --> 00:09:21.539 20th century and suddenly news isn't just inked 00:09:21.539 --> 00:09:25.120 on a page. We enter the broadcast era. Information 00:09:25.120 --> 00:09:27.720 is coming through the radio and shortly after 00:09:27.720 --> 00:09:30.340 beaming directly into living rooms via television. 00:09:30.620 --> 00:09:33.340 Right. How on earth does a clipping bureau adapt 00:09:33.340 --> 00:09:36.139 to that? You can't take scissors to a radio wave. 00:09:36.340 --> 00:09:38.480 That was the existential crisis for the industry. 00:09:39.559 --> 00:09:42.860 Broadcast media is ephemeral. Once the words 00:09:42.860 --> 00:09:46.580 are spoken on the radio, they vanish. Press clipping 00:09:46.580 --> 00:09:48.740 agencies had to figure out how to capture lightning 00:09:48.740 --> 00:09:51.320 in a bottle. So what did they do? This transition 00:09:51.320 --> 00:09:53.559 was only made possible by the invention and widespread 00:09:53.559 --> 00:09:55.779 commercialization of audio and videotape recording 00:09:55.779 --> 00:09:59.159 systems in the 1950s and 1960s. So instead of 00:09:59.159 --> 00:10:00.799 a room full of people reading papers, you have 00:10:00.799 --> 00:10:02.299 a room full of people listening to tape decks. 00:10:02.720 --> 00:10:05.259 Precisely. They would record the broadcasts, 00:10:05.259 --> 00:10:07.740 have human monitors sit with headphones and listen 00:10:07.740 --> 00:10:10.299 for client mentions, and then physically type 00:10:10.299 --> 00:10:12.139 out transcripts or summaries of the relevant 00:10:12.139 --> 00:10:15.500 sections. That sounds incredibly tedious, but 00:10:15.500 --> 00:10:17.740 I imagine it fundamentally changed the way people 00:10:17.740 --> 00:10:20.000 viewed information. Once you start isolating 00:10:20.000 --> 00:10:22.299 a spoken quote from a 30 -minute news broadcast, 00:10:22.700 --> 00:10:25.220 you are treating information differently. If 00:10:25.220 --> 00:10:27.559 we connect this to the bigger picture, this is 00:10:27.559 --> 00:10:29.539 actually one of the most crucial turning points 00:10:29.539 --> 00:10:33.200 in the history of information technology. The 00:10:33.200 --> 00:10:35.940 very concept of clipping this fundamental idea 00:10:35.940 --> 00:10:39.220 that a specific piece of text or data could be 00:10:39.220 --> 00:10:41.899 isolated, extracted from its original publication 00:10:41.899 --> 00:10:45.320 context, and evaluated entirely on its own was 00:10:45.320 --> 00:10:48.360 revolutionary. This conceptual leap directly 00:10:48.360 --> 00:10:51.379 influenced the design of early digital news databases. 00:10:51.820 --> 00:10:53.700 Wait, really? So the physical act of clipping 00:10:53.700 --> 00:10:56.620 a newspaper paved the way for things like LexisNexis? 00:10:56.759 --> 00:10:59.580 Exactly. The mindset of a press clipping bureau 00:10:59.580 --> 00:11:01.620 introduced the corporate world to the concept 00:11:01.620 --> 00:11:04.019 of keyword searching long before computers were 00:11:04.019 --> 00:11:06.360 doing it. Before you could search a digital database 00:11:06.360 --> 00:11:09.059 for a term, these bureaus were manually performing 00:11:09.059 --> 00:11:12.019 keyword searches across physical media and compiling 00:11:12.019 --> 00:11:14.519 the results. That makes total sense. They essentially 00:11:14.519 --> 00:11:17.399 trained the business world to expect information 00:11:17.399 --> 00:11:20.679 to be searchable and extractable. So when digital 00:11:20.679 --> 00:11:23.320 architecture was being built, it was designed 00:11:23.320 --> 00:11:25.580 to mimic the output of a press clipping bureau. 00:11:25.740 --> 00:11:28.419 That is fascinating. The analog scissors literally 00:11:28.419 --> 00:11:31.480 shaped the digital search bar, which brings us 00:11:31.480 --> 00:11:33.600 right to the doorstep of the internet boom in 00:11:33.600 --> 00:11:36.460 the 1990s. This had to be a do -or -die moment 00:11:36.460 --> 00:11:39.460 for those legacy clipping companies. If you were 00:11:39.460 --> 00:11:42.360 still mailing out pasted slips of paper in 1996, 00:11:42.940 --> 00:11:45.080 you were in trouble. It was an adaptor parish 00:11:45.080 --> 00:11:48.539 decade. A great example of this evolution is 00:11:48.539 --> 00:11:51.000 the Universal Press Clipping Bureau out in Omaha, 00:11:51.139 --> 00:11:53.820 Nebraska. They had been a staple of the industry 00:11:53.820 --> 00:11:57.700 operating since 1908. But in the 1990s, to survive 00:11:57.700 --> 00:12:00.279 and reflect the new digital reality, they had 00:12:00.279 --> 00:12:02.580 to entirely rebrand. What did they change it 00:12:02.580 --> 00:12:04.840 to? They changed their name to Universal Information 00:12:04.840 --> 00:12:07.480 Services. It was a signal to the market that 00:12:07.480 --> 00:12:09.159 they weren't just clipping anymore. They were 00:12:09.159 --> 00:12:12.360 managing digital data streams. And the pace of 00:12:12.360 --> 00:12:15.919 that digital shift is staggering. By 1998, websites 00:12:15.919 --> 00:12:18.240 like Web Clipping Launch becoming some of the 00:12:18.240 --> 00:12:21.240 very first to monitor Internet -based news. And 00:12:21.240 --> 00:12:23.539 from there, it's an absolute gold rush. It really 00:12:23.539 --> 00:12:26.519 was. To give you an idea of how fast this exploded 00:12:26.519 --> 00:12:29.440 by 2012, researchers at Gartner estimated there 00:12:29.440 --> 00:12:32.399 were more than 250 different vendors doing just 00:12:32.399 --> 00:12:35.559 social media monitoring. It went from a few legacy 00:12:35.559 --> 00:12:38.580 companies with scissors to hundreds of tech startups 00:12:38.580 --> 00:12:41.450 overnight. The barriers to entry dropped significantly 00:12:41.450 --> 00:12:44.029 once the physical element was removed, and the 00:12:44.029 --> 00:12:46.190 sheer volume of data on the Internet demanded 00:12:46.190 --> 00:12:49.169 automated solutions. So what does this all mean 00:12:49.169 --> 00:12:51.929 for how the machine actually works today? We've 00:12:51.929 --> 00:12:54.649 gone from cut -and -clip sweatshops to massive 00:12:54.649 --> 00:12:57.669 server farms. How does a modern media monitoring 00:12:57.669 --> 00:13:00.450 service operate? Well, first, the industry has 00:13:00.450 --> 00:13:02.870 entirely rebranded itself to match its technological 00:13:02.870 --> 00:13:05.649 capabilities. You will rarely hear the term clipping 00:13:05.649 --> 00:13:08.019 bureau anymore. Today, it's referred to as information 00:13:08.019 --> 00:13:11.019 logistics or media intelligence. Media intelligence. 00:13:11.080 --> 00:13:14.580 It sounds so covert. It does. And the delivery 00:13:14.580 --> 00:13:17.360 methods are completely transformed. Clients aren't 00:13:17.360 --> 00:13:19.379 waiting for the mailman. We are talking about 00:13:19.379 --> 00:13:21.659 customized dashboards, instant email alerts, 00:13:21.779 --> 00:13:24.120 keeping clients updated minute by minute as news 00:13:24.120 --> 00:13:26.600 breaks and auto analysis formats, where vast 00:13:26.600 --> 00:13:28.940 amounts of data are visually graphed to show 00:13:28.940 --> 00:13:31.740 sentiment trends over time. And the tools driving 00:13:31.740 --> 00:13:33.940 this are names a lot of people might recognize. 00:13:34.259 --> 00:13:36.519 On the consumer side, you have Google Alerts. 00:13:36.519 --> 00:13:38.500 But on the massive enterprise side, you have 00:13:38.500 --> 00:13:41.480 platforms like Scission Meltwater Medinet and 00:13:41.480 --> 00:13:44.580 MuckRack. Yes. And the mechanics of how these 00:13:44.580 --> 00:13:47.299 systems capture the mentions are incredibly sophisticated, 00:13:47.580 --> 00:13:50.960 though they vary by medium. Let's look at television 00:13:50.960 --> 00:13:53.590 today. In the United States, television monitoring 00:13:53.590 --> 00:13:56.610 companies use a really clever workaround. They 00:13:56.610 --> 00:13:59.029 aren't just having people watch TV all day. No. 00:13:59.129 --> 00:14:01.470 Instead of having humans watch hundreds of channels, 00:14:01.669 --> 00:14:04.289 they capture and index the closed captioning 00:14:04.289 --> 00:14:07.370 text of the broadcasts. Once you turn the audio 00:14:07.370 --> 00:14:09.970 into a text file, it's easily searchable for 00:14:09.970 --> 00:14:12.250 client references. But surely they don't rely 00:14:12.250 --> 00:14:15.409 completely on the automated text. Closed captioning 00:14:15.409 --> 00:14:17.830 can be notoriously messy. That's a great point, 00:14:17.909 --> 00:14:19.789 and it's why the industry is actually split on 00:14:19.789 --> 00:14:22.350 this. Some TV monitoring companies still employ 00:14:22.350 --> 00:14:25.009 human monitors who review the video and write 00:14:25.009 --> 00:14:28.029 abstracts of the program content to ensure context 00:14:28.029 --> 00:14:30.389 is captured. Okay, that makes sense. But others 00:14:30.389 --> 00:14:32.870 rely entirely on those automated search programs 00:14:32.870 --> 00:14:35.870 to index the stories. It often comes down to 00:14:35.870 --> 00:14:38.190 what the client is willing to pay for speed versus 00:14:38.190 --> 00:14:40.649 nuanced accuracy. And what about the Internet? 00:14:41.129 --> 00:14:44.309 The sheer volume of online media is mind -boggling. 00:14:44.370 --> 00:14:46.350 You can't have humans reading every blog and 00:14:46.350 --> 00:14:48.649 news site. For online monitoring, it is entirely 00:14:48.649 --> 00:14:51.350 automated. These services utilize software known 00:14:51.350 --> 00:14:54.470 as spiders or bots. These bots are programmed 00:14:54.470 --> 00:14:57.370 to constantly crawl the web, monitoring the content 00:14:57.370 --> 00:14:59.990 of free online news sources. Newspapers, magazines, 00:15:00.269 --> 00:15:03.230 trade journals, TV station websites, news syndication 00:15:03.230 --> 00:15:05.769 feeds the bots, grab the links, and often pull 00:15:05.769 --> 00:15:08.309 the text versions of the articles directly into 00:15:08.309 --> 00:15:10.899 the client's dashboard in real time. That sounds 00:15:10.899 --> 00:15:13.120 incredibly efficient, but there's no way a bot 00:15:13.120 --> 00:15:15.700 understands context right. I mean, if I'm tracking 00:15:15.700 --> 00:15:17.899 the tech company Apple, how does the bot know 00:15:17.899 --> 00:15:19.980 an article isn't just about a local farmer's 00:15:19.980 --> 00:15:21.940 market having a great apple harvest? That is 00:15:21.940 --> 00:15:25.039 the exact flaw in the modern system. The automated 00:15:25.039 --> 00:15:27.100 bot results are pulled in at lightning speed, 00:15:27.259 --> 00:15:29.519 but they aren't always verified for accuracy 00:15:29.519 --> 00:15:32.980 by the monitoring service. A bot lacks human 00:15:32.980 --> 00:15:35.960 discernment. Furthermore, there is a massive 00:15:35.960 --> 00:15:39.240 structural blind spot. What's that? Not all print 00:15:39.240 --> 00:15:42.460 content makes it online. And conversely, a lot 00:15:42.460 --> 00:15:44.759 of web content never appears in physical print. 00:15:45.299 --> 00:15:48.620 If you rely solely on a digital bot, scraping 00:15:48.620 --> 00:15:52.440 the web in, you might miss a crucial piece of 00:15:52.440 --> 00:15:54.639 physical journalism and vice versa. Which is 00:15:54.639 --> 00:15:57.500 a huge deal when you consider who is relying 00:15:57.500 --> 00:15:59.940 on this data. If you are a listener thinking, 00:16:00.100 --> 00:16:02.620 why does this matter to me? Just look at the 00:16:02.620 --> 00:16:05.580 use cases today. Every single public relations 00:16:05.580 --> 00:16:08.639 team on earth uses this. They use it to track 00:16:08.639 --> 00:16:11.279 their own publicity, to monitor what their competitors 00:16:11.279 --> 00:16:13.620 are doing, and to prove to their bosses that 00:16:13.620 --> 00:16:15.600 their marketing campaigns are actually reaching 00:16:15.600 --> 00:16:18.299 the target audience. And it goes way beyond corporate 00:16:18.299 --> 00:16:21.019 PR. Imagine you're a local mayor or a federal 00:16:21.019 --> 00:16:23.559 agency director. City, state, and federal governments 00:16:23.559 --> 00:16:26.059 are massive clients of these services. Wow, I 00:16:26.059 --> 00:16:28.230 didn't even think about that. They rely on media 00:16:28.230 --> 00:16:30.230 intelligence to stay informed about regional 00:16:30.230 --> 00:16:32.450 issues they couldn't possibly monitor manually. 00:16:32.669 --> 00:16:35.149 They use the data to verify that the public safety 00:16:35.149 --> 00:16:37.169 information they are putting out is accurate, 00:16:37.269 --> 00:16:39.470 accessible in multiple formats, and actually 00:16:39.470 --> 00:16:41.580 reaching the communities that need it. It's such 00:16:41.580 --> 00:16:44.379 a vital infrastructure that the industry actually 00:16:44.379 --> 00:16:47.500 has its own global trade associations to establish 00:16:47.500 --> 00:16:50.539 best practices. You have the North American Conference 00:16:50.539 --> 00:16:52.519 of Press Clipping Services, the International 00:16:52.519 --> 00:16:55.960 Association of Broadcast Monitors. This is a 00:16:55.960 --> 00:16:58.840 highly formalized, heavily entrenched global 00:16:58.840 --> 00:17:01.340 enterprise. But as with any industry that scales 00:17:01.340 --> 00:17:03.940 this massively by utilizing automated technology, 00:17:04.319 --> 00:17:07.059 it eventually runs headfirst into a brick wall. 00:17:07.259 --> 00:17:10.190 In this case, that wall was the law. Specifically, 00:17:10.210 --> 00:17:12.970 the intense friction between automated web scraping 00:17:12.970 --> 00:17:15.670 technology and existing copyright legislation. 00:17:15.970 --> 00:17:18.869 The legal minefield. Because this raises a fascinating 00:17:18.869 --> 00:17:21.130 question. Who actually owns a snippet of news? 00:17:21.369 --> 00:17:23.529 If a bot reads an article and shows me a two 00:17:23.529 --> 00:17:26.009 -sentence summary, is that theft? In 2012, this 00:17:26.009 --> 00:17:28.569 tension exploded into two parallel lawsuits that 00:17:28.569 --> 00:17:30.789 hit the industry simultaneously. One in the United 00:17:30.789 --> 00:17:32.789 States and one in the United Kingdom. Right. 00:17:32.890 --> 00:17:35.150 And what's wild is that both cases involved the 00:17:35.150 --> 00:17:37.589 exact same defendant, the media monitoring giant 00:17:37.589 --> 00:17:40.480 Meltwater Group. This raises an important question 00:17:40.480 --> 00:17:42.900 about how intellectual property is handled in 00:17:42.900 --> 00:17:46.220 the digital age, where copying data is effortless. 00:17:46.619 --> 00:17:50.200 The core legal battle in both jurisdictions revolved 00:17:50.200 --> 00:17:53.299 around two things, the concept of temporary copies 00:17:53.299 --> 00:17:56.420 and the actual business model of generating snippets 00:17:56.420 --> 00:17:58.450 for clients. Let's break that down, starting 00:17:58.450 --> 00:18:01.029 with the U .S. case. This was Associated Press 00:18:01.029 --> 00:18:04.089 v. Meltwater. The Associated Press essentially 00:18:04.089 --> 00:18:06.849 argued you can't just have your bots scrape our 00:18:06.849 --> 00:18:09.190 journalists' hard work, pull out the most important 00:18:09.190 --> 00:18:11.910 sentences, and sell that summary to your clients. 00:18:12.089 --> 00:18:14.910 And the court agreed right. They did. The U .S. 00:18:14.930 --> 00:18:16.970 court ruled that the activity of scraping and 00:18:16.970 --> 00:18:19.130 showing those specific snippets to clients was 00:18:19.130 --> 00:18:22.269 unlawful. Crucially, they determined that Meltwater's 00:18:22.269 --> 00:18:24.990 automated service did not fall under the protection 00:18:24.990 --> 00:18:27.670 of the fair use doctrine. it was deemed a substitute 00:18:27.670 --> 00:18:29.650 for the original work. Okay, so that's the U 00:18:29.650 --> 00:18:32.289 .S. But the U .K. case, Public Relations Consultants 00:18:32.289 --> 00:18:34.569 Association v. the Newspaper Licensing Agency, 00:18:34.890 --> 00:18:37.950 dealt with this idea of temporary copies. Can 00:18:37.950 --> 00:18:39.730 you explain that for me? Because on the surface, 00:18:39.750 --> 00:18:42.230 it sounds like the court was regulating how a 00:18:42.230 --> 00:18:44.609 computer cache works. It does sound incredibly 00:18:44.609 --> 00:18:48.509 technical, but it's vital. Under U .K. and E 00:18:48.509 --> 00:18:51.930 .U. copyright law, the nuance is very specific 00:18:51.930 --> 00:18:55.799 regarding what constitutes a copy. Okay. If you, 00:18:55.920 --> 00:18:59.039 as a user, go to a news website and simply view 00:18:59.039 --> 00:19:01.519 the original source, meaning you just look at 00:19:01.519 --> 00:19:04.539 the article online without printing it or downloading 00:19:04.539 --> 00:19:07.460 a permanent file, that is not copyright infringement. 00:19:07.539 --> 00:19:10.920 So my cash saving it is fine? Yes. The temporary 00:19:10.920 --> 00:19:12.859 transient copies your computer automatically 00:19:12.859 --> 00:19:15.700 makes in its cache just to display the web page 00:19:15.700 --> 00:19:18.279 on your screen are considered lawful. Okay, that 00:19:18.279 --> 00:19:20.220 makes sense. Otherwise, browsing the internet 00:19:20.220 --> 00:19:22.619 at all would be illegal. So why were the media 00:19:22.619 --> 00:19:24.539 monitoring companies in trouble? Because the 00:19:24.539 --> 00:19:26.299 media monitoring business model doesn't work 00:19:26.299 --> 00:19:28.609 that way. Clients aren't paying Meltwater or 00:19:28.609 --> 00:19:30.990 Cision to just go read original websites one 00:19:30.990 --> 00:19:33.269 by one. They're paying the service to actively 00:19:33.269 --> 00:19:36.170 generate, extract, and share customized headlines, 00:19:36.390 --> 00:19:38.730 summaries, and snippets in an aggregated permit 00:19:38.730 --> 00:19:41.509 dashboard. Ah, I see. Because the service inherently 00:19:41.509 --> 00:19:44.650 relies on generating and distributing those extracted 00:19:44.650 --> 00:19:46.890 pieces of the original work, not just temporarily 00:19:46.890 --> 00:19:49.609 viewing them, the courts determined that both 00:19:49.609 --> 00:19:51.930 the service providers and the end users of those 00:19:51.930 --> 00:19:55.069 services now require explicit licenses from the 00:19:55.069 --> 00:19:57.990 publishers. We are stating this purely factually 00:19:57.990 --> 00:20:00.130 based on the outcomes provided in the history, 00:20:00.309 --> 00:20:02.930 but it completely forced the industry to adapt. 00:20:03.210 --> 00:20:05.490 They had to restructure their licensing agreements 00:20:05.490 --> 00:20:08.569 globally to comply with these rulings. It perfectly 00:20:08.569 --> 00:20:11.549 highlights a persistent truth technology almost 00:20:11.549 --> 00:20:14.529 always outpaces legislation. The media intelligence 00:20:14.529 --> 00:20:16.849 firms built the sweeping spiders and bots first 00:20:16.849 --> 00:20:19.880 because the technology allowed it. But the legal 00:20:19.880 --> 00:20:22.240 framework regarding digital fair use and temporary 00:20:22.240 --> 00:20:24.920 copies had to painfully catch up to the reality 00:20:24.920 --> 00:20:26.720 of how business was actually being conducted 00:20:26.720 --> 00:20:29.319 on the ground. It has been a wild journey. We 00:20:29.319 --> 00:20:32.059 started with Alfred Cherie in 1879, smelling 00:20:32.059 --> 00:20:34.480 like glue, cutting out physical newspaper clippings 00:20:34.480 --> 00:20:36.579 so Parisian stage actors could read their own 00:20:36.579 --> 00:20:38.480 reviews without having to buy the whole paper. 00:20:38.720 --> 00:20:41.480 A pure, simple vanity project. A long way from 00:20:41.480 --> 00:20:43.599 where we are now. And today we are looking at 00:20:43.599 --> 00:20:46.500 a global spider bot driven intelligence network. 00:20:46.920 --> 00:20:49.599 It is an invisible infrastructure of information 00:20:49.599 --> 00:20:52.539 logistics that local governments, federal agencies 00:20:52.539 --> 00:20:55.680 and massive corporations rely on every single 00:20:55.680 --> 00:20:58.720 day just to function and understand the world 00:20:58.720 --> 00:21:01.180 around them. And yet it all stems from that same 00:21:01.180 --> 00:21:04.799 fundamental human impulse. The tools change from 00:21:04.799 --> 00:21:08.059 steel scissors to scraping algorithms. But the 00:21:08.059 --> 00:21:10.980 desire to isolate, track and analyze what the 00:21:10.980 --> 00:21:13.420 world is saying about us remains exactly the 00:21:13.420 --> 00:21:16.190 same. It is incredible to think about. As we 00:21:16.190 --> 00:21:17.690 wrap up this deep dive, we want to leave you 00:21:17.690 --> 00:21:20.170 with a final thought to mull over. Consider this. 00:21:20.490 --> 00:21:23.589 If media monitoring spiders and bots are constantly 00:21:23.589 --> 00:21:25.529 scraping the internet for brand mentions and 00:21:25.529 --> 00:21:28.470 PR departments, use that automated bot -generated 00:21:28.470 --> 00:21:30.769 data to write and distribute new press releases. 00:21:31.069 --> 00:21:33.049 Are we approaching a closed loop? Exactly. A 00:21:33.049 --> 00:21:35.289 world where bots are simply writing media for 00:21:35.289 --> 00:21:37.710 other bots to monitor. Where does actual nuanced 00:21:37.710 --> 00:21:40.089 human understanding fit into that automated feedback 00:21:40.089 --> 00:21:42.559 loop? It is certainly a profound question to 00:21:42.559 --> 00:21:44.220 ponder the next time you read a piece of corporate 00:21:44.220 --> 00:21:46.480 news. Thank you so much for taking this deep 00:21:46.480 --> 00:21:49.259 dive with us. Stay curious and always look for 00:21:49.259 --> 00:21:51.920 the hidden systems and stories behind the information 00:21:51.920 --> 00:21:53.619 you consume every day.