Bummer! This is just a preview. You need to be signed in with a Pro account to view the entire video.
Content in a Zombie Apocalypse40:50 with Karen McGrane
Friends, a zombie apocalypse is upon us: an onslaught of new mobile devices, platforms, and screen sizes, hordes of them descending every day. We're outmatched. There aren't enough designers and developers to battle every platform. There aren't enough editors and writers to populate every screen size. Defeating the zombies will require flexibility and stamina—in our content. We'll have to separate our content from its form, so it can adapt appropriately to different contexts and constraints. We'll have to change our production workflow so we're not just shoveling content from one output to another. And we'll have to enhance our content management tools and interfaces so they're ready for the future. Surviving the zombie apocalypse is possible. In this talk Karen will explain how: by developing a content strategy for mobile.
[MUSIC] 0:00 Hey. 0:05 You guys wanna know what I think is awesome? 0:05 >> What's awesome? 0:08 [LAUGH] >> New York City. 0:09 Come on guys, give it up. 0:11 My home town crowd. 0:13 Okay, you wanna know what else I think is awesome? 0:14 >> Yeah. 0:18 >> Print. 0:19 My god, print is amazing. 0:19 Like what a fantastic technology. 0:21 I mean you put the words on the paper and they stay there. 0:26 You don't have to worry about keeping it updated all the time. 0:30 So print like, when you really look back at 0:34 the history of human communication, the 500 years we 0:37 have of the printing press and the thousands of 0:41 years that came before that of the written word. 0:44 So much of our culture and values and processes for what it means to 0:47 communicate are all derived from what it took to put ink on paper. 0:52 And then we had to go and invent the web. 1:00 Which, I don't know, I mean on balance, I think it's been totally 1:04 worth it but I mean, man, is it a pain in the ass? 1:07 So the web, the web just turned 25. 1:12 25 years old and I'm like a huge nerd for computing history 1:15 and so when you, when you look back at the very foundations of where 1:20 the Web started, the earliest technologies of HTML, 1:25 HyperText Markup Language, and the URL, the uniform resource locator. 1:30 These tools were designed from the start with the explicit intent that 1:36 anyone, anywhere could create a document 1:41 that could be accessible globally and updated instantaneously. 1:47 And when you, like, when you kind of take a step back 1:55 from like, the work that we do everyday and just kind of 1:58 appreciate just how transformational that is, in the history of human communication, 2:02 I mean just how different the web is, from everything that came before. 2:07 It's like 25 years is barely enough time for human kind to 2:13 have wrapped our head around the way that the Web is different. 2:17 And so we've spent the last quarter 2:23 century just kind of like band-aiding ourselves along. 2:25 You know, just kind of using some training wheels where we told everybody, well, 2:28 I mean, you know, a web page, it's really just like a printed page. 2:33 I mean, you can just imagine that the web page is sort of like a piece of paper. 2:38 And then came the rise of mobile. 2:44 And so now we have people accessing the web through a range of different devices. 2:47 And mobile, when I talk about the, the, the challenge of publishing on 2:53 mobile, what I mean is that mobile has forced us to give up our 2:57 shared hallucination that we have any control over the size screen that somebody 3:02 has or the input mechanism that they're using or how much bandwidth they have. 3:08 And it's that, that challenge with the rise of mobile, and now with tablet 3:14 devices, like it's the fact that we 3:20 no longer have the luxury of making assumptions. 3:23 And that is gone, and it's not coming back. 3:26 And so to me, this is the real transition. 3:30 When I talk about content strategy for mobile I don't mean like what's 3:33 your strategy for getting your stuff onto smart phones or marketing to mobile users. 3:37 What I mean is how do we adapt our publishing processes? 3:42 How do we adapt the very way that we think about how we 3:46 communicate and how we, how we talk to people about the content creation process? 3:50 To understand that now our content has to 3:55 live on a wide range of devices and platforms. 3:58 And to look at it from this vantage point right now it's, it's, it's, 4:02 I think it's still easy to fall into the trap of going, okay well 4:06 you know, but smart phones and tablets and desktop, they're still, I mean maybe 4:09 it's a squishy piece of paper but it's still really just a squishy paper right? 4:12 So I'm not a futurist here okay? 4:17 I'm not here to predict for you what I think the next big thing is gonna be. 4:19 But I am gonna tell you one thing. 4:25 I'll guarantee it. 4:26 There is gonna be a next big thing and whatever that next 4:27 big thing is, we're gonna have to get our stuff onto it. 4:31 Maybe it'll be smart TV, so a lot of people hypothesizing that the next 4:35 big wave is gonna be how we get our stuff into the living room. 4:40 And when you start thinking about all of these different form 4:44 factors, I mean to me this idea really drives home the 4:48 fact that is is a fallacy that we can make any assumptions at all about screen 4:52 size, resolution capabilities, characteristics, input device. 4:57 I mean, the opportunity here isn't to say oh, well how can 5:03 we just treat the smart TV like it's a bigger piece of paper. 5:07 The opportunity here is to say how do 5:10 we think about publishing our content in a way 5:13 that doesn't rely on us having to imagine 5:16 that this is just the same thing as before. 5:20 Or, I don't know. 5:24 Maybe the next big thing is gonna be in car systems. 5:25 So I don't know about you guys, but I live here. 5:28 I don't drive a car. 5:31 The idea of drivers going around with a 5:32 touchscreen on their dashboard is terrifying to me. 5:34 But that's the thing. 5:38 With in-car systems, the opportunity here isn't 5:39 actually having a touchscreen on your dashboard. 5:42 The opportunity here is an interconnected ecosystem of perhaps having 5:45 some kind of controls but also having an audio interface 5:51 that you can use to have things read to you 5:56 and navigate using both the combination of touch and voice. 5:58 And so this, this, you know, this starts to open up all kinds of possibilities. 6:04 Where, you know, you realize that one day, the 6:08 Star Trek computer is going to become a reality. 6:12 And I know it's, it's pretty easy to, to laugh at this right now, right? 6:16 I mean it's like we've all had the experience of using Siri or 6:20 God forbid, one of those like, interactive voice touch, you know, phone things. 6:24 It's terrible, right? 6:28 Audio interfaces right now are terrible. 6:29 You know what else, used to suck? 6:33 Touch screens. 6:36 Touch screens were awful. 6:37 You remember going to the ATM and having to like, 6:38 angle your finger just right to get the button to push? 6:41 For decades, touch screens were seen as this almost 6:45 ran technology that was never quite gonna reach prime time. 6:49 And then one day touch screens worked. 6:54 They worked perfectly and it transformed our entire 6:56 industry in the space of like five years. 7:00 So I don't know when touch screen, or when, 7:03 when audio interfaces are really going to become a reality. 7:05 5 years? 7:07 15 years? 7:07 50 years? 7:08 But I guarantee you when the Star Trek computer is 7:10 a reality, when I go up and ask that computer a 7:13 question, the last thing I want to hear is it 7:15 for to start reading h-t-t-p colon slash slash back at me. 7:18 And so when you, when you realize, like, what this means 7:23 for the actual way that we create content and produce our code. 7:27 Like, for example, I've had developers come to 7:31 me and ask, Karen, why do we have to 7:33 bother parsing out the difference between what is 7:36 rendered as emphasis and what is rendered as italic? 7:39 When every major browser since the dawn of time has rendered emphasis as italics. 7:43 Not on an audio interface they don't. 7:50 One of those things conveys visual styling, the other conveys meaning. 7:53 In an audio interface emphasis conveys tone of voice. 7:58 And so, when you start peeling back what this 8:03 means for how we create content, you start to realize 8:07 that, doing the right thing for creating content for the 8:10 future, actually means doing the right thing for accessibility today. 8:14 And if you don't believe me on this, check out 8:18 what Google is doing with what they call their conversation group. 8:22 So, Google. 8:26 Google is a company that can make cars that drive themselves. 8:27 That this, that's not even their core business. 8:32 That's just some weird side-project that they cooked up. 8:34 But no. 8:37 The conversation group, the idea of how 8:38 human beings communicate with technology through language? 8:40 That is squarely Google's core business, and they are investing really heavily in 8:44 coming up with ways to enable people 8:49 to communicate with computers through their voice. 8:51 So, if that's the, you know, if that's the next big 8:54 thing then maybe the next big thing's gonna be Google Glass. 8:57 >> [LAUGH] 9:01 >> I don't really think Google Glass is gonna be the next big thing. 9:03 I think Google Glass is like the Segway of mobile. 9:06 It's like a Segway, but for your face. 9:11 But, okay. 9:15 It's easy to mock Google Glass, but what I don't mock is 9:17 the very idea that wearables will become a real part of the future. 9:20 I worked with this, this industrial products catalogue company recently. 9:24 So they, they, they produce industrial parts that would be used in factories 9:28 and they put out this huge, I mean just this enormous printed catalog. 9:32 And I went and talked to them and I said, look, 9:36 I'm not gonna tell you when you will stop printing that catalog. 9:38 I mean, I'm not, it might be five years 9:42 from now, it might be fifteen years from now. 9:45 But I guarantee you the day will come when you will make the decision to no 9:47 longer print that catalog, and when that day 9:51 comes the idea that your audience, these, these people 9:54 who work on the factory floor, that they 9:58 might be more likely to have a wearable device 10:01 than they would be to wanna, say, walk 10:04 back to their office and use a desktop computer. 10:07 I think that's a reality. 10:10 And so what do you have to do, starting right 10:12 now, today to get your publishing process in a place, 10:14 so that when you stop printing that printed catalog, and 10:19 you start having to publish to wearable, you can do it. 10:22 Because right now your entire publishing process is 10:25 geared toward how you get that ink on paper. 10:28 And what you're gonna have to do differently for whatever device that 10:31 might come down the pipe, you need to start thinking about that now. 10:34 So maybe not Google Glass, maybe it's watches. 10:39 Everybody is talking about watches. 10:42 Galaxy Gear, the Pebble, the rumored iWatch. 10:44 So, a lot of times when I talk about 10:49 watches, people are like, oh Karen, watches, they're so tiny. 10:51 Can't be the next big thing. 10:55 I mean, who really wants to read a long document on their watch, right? 10:56 Probably not. 10:59 But when you think of the watch as existing in a 11:02 larger ecosystem of devices, that the watch is the controller for your 11:05 smart TV, that the watch enables you to control, you know, navigate 11:09 through information that then is read to you through an audio interface. 11:14 You realize that even if our content does not have to be fully rendered on 11:17 every device type, it still has to 11:23 be accessible and navigable from every device type. 11:25 And, so we need to have processes in place to insure 11:28 that what we're publishing makes sense even on the smallest form factors. 11:31 But, I mean, I get it. 11:37 Watches, so tiny. 11:39 Clearly not the next big thing. 11:41 Maybe the next big thing is gonna be, stadium scoreboards. 11:43 Come on guys, we can get our content where everybody can see it. 11:46 [BLANK_AUDIO] 11:49 This is a joke. 11:52 [LAUGH]. 11:53 >> I put this in here as a joke cuz I was like, 11:54 oh it will be funny cuz watches are tiny and then scoreboards are huge. 11:56 And then I turned around and I talked to a client that 11:59 said that the biggest problem that they are facing today is digital signage. 12:04 It's a university client. 12:07 It's actually SVA right here in, in Manhattan. 12:09 And I talked to them, and they said they'd just purchased all of these 12:12 new screens, these new digital signs that they wanted to put up all over campus. 12:15 And so they wanted a way to manage 12:20 and maintain their content in one place and have 12:21 it automatically publish out to their website, to 12:24 their mobile app, and to these new digital screens. 12:28 And I was like 12:31 hey, have you thought about treating the digital signs as 12:34 a completely separate workflow and managing as their own silo? 12:36 They were like, Karen, ain't nobody got time for that. 12:40 No. 12:46 When we have campus alerts, or events listings, or information for 12:47 students, we want a way for that content to exist, and for 12:52 us to manage it and for it to automatically publish out 12:56 to three very different form factors with three very different design requirements. 12:59 This isn't some crazy, futuristic dystopia that these people are living in. 13:04 This is a real problem that they are trying to solve right now, today. 13:08 So in digital signage is something that many 13:12 organizations in the enterprise are gonna be wrestling with. 13:15 Maybe you too one day, will be trying to solve 13:19 the same problem on the internet refrigerator in your home. 13:21 Come on, guys. 13:25 We can't talk about the future 13:26 without talking about the internet refrigerator, right? 13:27 >> [LAUGH] 13:29 >> I mean, personally I think the biggest problem facing our society today is that 13:30 I can't check Twitter in the 30 seconds it takes me to get a Diet Coke. 13:35 Okay, I don't really think the internet refrigerator is where 13:41 it's at, but obviously the idea that our homes will also 13:43 become a network of interconnected sensors and devices and that there 13:48 maybe screens on which we want to access or navigate information. 13:52 Heck, what if it's not the refrigerator? 13:56 What if instead, it's that your entire cook top, the 13:58 glass cook top of your stove is now, has an 14:02 imbedded iPad interface, and you can navigate through recipes or 14:05 videos or news information right there from your cooking space? 14:10 Or, and I do think that this is the biggest problem 14:14 facing our society today, what happens when toaster printers become a reality? 14:18 >> [LAUGH]. 14:23 >> Is your content going to be ready to be burned onto delicious toast? 14:26 These are the problems that we're here to solve, together. 14:34 So right now today, organizations are already faced with the 14:36 challenge of getting their content onto a variety of different platforms. 14:41 They have the web, and now apps. 14:45 They're wrestling with, you know, mobile apps, smartphone apps versus tablet apps. 14:47 They've gotta get their content onto blogs, or an email, or social channels. 14:51 Heck, some of them are still trying to figure out how to get it on to print. 14:56 And tomorrow, it will be a veritable zombie apocalypse of new 14:59 devices, and platforms, and screen sizes, and resolutions, and input mechanisms. 15:03 And it's just gonna keep coming, and it's not gonna stop. 15:07 And so the real question becomes how do we protect ourselves from the zombies? 15:12 How do we protect ourselves 15:15 knowing that we can't simply battle every zombie as it comes? 15:20 That, that, they're coming too quickly. 15:26 We, we can't wrestle them. 15:27 By insisting that we have to have an 15:29 uniquely siloed workflow for publishing to every single channel. 15:31 And so protecting ourselves from the zombies means that we 15:36 have to think differently about what are publishing process means. 15:39 And for this, I think, I think this really gets to 15:42 the heart of what the challenge of mobile actually means for publishing. 15:45 And that is the true separation of content from form. 15:51 And so, we have to support too many form factors to assume that we can 15:57 tightly couple what it is that we are publishing with what it looks like. 16:03 Do you have any idea how different this is, in the way that people communicate? 16:09 I mean, when you look back across all of human communication, 16:17 and you think about this, this medieval monk, laboriously hand scribing documents. 16:21 For us to go to him and suggest that we have a conversation, about what it 16:27 means to communicate the meaning of that document 16:32 without tying that to any particular form or factor, or visual 16:37 styling information that's embedded in that document, it wouldn't make any sense. 16:43 I mean there's no, there's just no way. 16:47 There's no purpose of even having that conversation because up 16:50 until this point in history, every document that we produce, 16:53 the meaning was so tightly intertwined with the way that 16:56 it looked, with its form, that it wasn't even worth discussing. 17:00 And yet today when I talk about mobile, when 17:05 I talk about having all of these different form factors. 17:07 When I talk about this zombie apocalypse of devices, 17:10 that is precisely the problem we need to solve. 17:13 How can we ensure that the meaning, the essence, the semantic information 17:17 that is embedded in that document is not purely tied to visual styling? 17:22 And so when I say we have to 17:28 separate content from form, it's real easy, especially when 17:32 I talk to an audience of developers, it's real 17:37 easy to think that what I mean is HTML. 17:39 Like get your HTML out of my content. 17:43 And that's part of it but it's really only the beginning. 17:45 When Dan Jacobson, so he's the API guy from Netflix who's formerly at NPR. 17:49 When he says that the future of content management 17:53 systems is in their ability to capture the content 17:55 in a clean, presentation-independent way, you know, it's is 17:58 tempting to reduce this to a problem with mark up. 18:04 To think that we can solve this problem by getting rid of rich text editors in 18:09 our interfaces and I will say I am one of the industry's leading 18:14 advocates for saying we should not be 18:19 giving content creators a tool that encourages 18:22 them to think about publishing to the web as if it's a piece of paper. 18:26 Where they get a giant blob where they can dump whatever they want into it. 18:30 And you know we give them a tool bar at the top that mimics 18:34 the familiar what you see is what you get interface and we tell them oh 18:37 sure, well you can put a table in there and add some custom bullets 18:41 and you can take a box and float it to the right if you want. 18:46 And then when it comes time to take this information and put it on any 18:50 other platform, all of these decisions, all of 18:55 this styling information has to be stripped out. 18:58 The problem is, that styling information isn't just for decoration. 19:01 Like those content creators intended to 19:08 communicate something with their styling choices. 19:11 The person who floated that box to the right, they 19:13 didn't really mean I want a box floated to the right. 19:16 What they meant was I want this box to set off from the content. 19:19 It's just that the tools that we have given 19:24 them, the language that we have to communicate that forces 19:26 them to continue to think of it as a 19:30 visual styling tool rather than as embedding what they mean. 19:32 And so, when I get frustrated by this, when I get frustrated by the 19:37 limitations of blobs, I wail against these kind of wizzy wig interfaces on the web. 19:40 And as a result I've gotten the reputation as 19:46 being, like, the President of the wizzy wig haters club. 19:48 And I want you guys to know, I'm, I'm like really kinds proud of this. 19:52 My graffiti crime spree is going exceptionally well. 19:55 You know, it's nice to get the message 19:59 out in, in you know, conversations like this one. 20:01 But you know, the graffiti, it reaches a much wider audience. 20:03 So, the pro, you know, sometimes when I say this, like, developers will come 20:07 up to me and they'll be like yeah, mark down, mark down's the answer. 20:11 And I'm like, well, the problem's not the toolbar exactly. 20:14 I mean, I don't care if somebody has a toolbar at the top 20:19 of their screen that allows them to push a button to embed some format. 20:23 Now, the problem with WYSIWYG is that it relies on an outdated mental model. 20:28 That is derived from print to communicate to 20:34 people what it means to publish on the web. 20:37 You know where WYSIWYG came from? 20:41 Came from Xerox, came from Xerox PARC. 20:43 Because they invented the laser printer. 20:47 Think about this for a minute, okay. 20:52 Up until this point, Xerox machines could only 20:54 make a copy of a document that already existed. 20:58 Xerox invented a way that you could print out whatever you wanted. 21:04 But they needed a way for you to actually be able to create that document. 21:09 And so Xerox invented the entire concept of 21:15 the what you see is what you get interface. 21:19 They invented the entire concept of the graphical user interface. 21:21 Which drove the rise of adoption of the personal 21:25 computer, which directly led to the rise of the web. 21:28 All of us here today have jobs because people 21:31 bought personal computers because they wanted a laser printer. 21:35 And I feel like I can say with a great deal of respect for what 21:41 this mental model, what this metaphor has given us over the past 40 years or so. 21:45 I can say it with a great deal of respect and 21:50 also with knowledge, that the web is not a laser printer. 21:52 My God, the web has nothing in common with a laser printer. 21:56 There are no reality of what someone is doing or publishing on the web. 21:59 That is anything like printing out a document 22:05 on the laser printer next to your desk. 22:08 And yet, the tools that we give content creators are 22:11 directly derive from that antiquated metaphor from the desktop publishing era. 22:15 That we use to communicate to people that it was like printing out a Word document. 22:21 Ted Nelson says imitating paper on a computer screen is like tearing the 22:28 wings off of a 747 and using it as a bus on the highway. 22:33 The web is better. 22:39 The web can do so much more. 22:41 And yet, we are still grounding creators in this antiquated mental model. 22:44 That doesn't allow them to take advantage of 22:49 all of the ways that the Web is different. 22:52 We are fighting a war here, people. 22:55 We are fighting a war of blobs versus chunks. 22:58 Blobs are these messy, formatting dependent, embedded clumps of stuff. 23:01 That where all of the meaning of that text, the meaning 23:07 of the document, is so tightly intertwined with what it looks like. 23:11 That we cannot reasonably take it to any other platform 23:15 without having to do a significant amount of manual rework. 23:18 In order to communicate again in 23:22 another context what that content creator intended. 23:24 Whereas chunks are clean presentation independent bits of 23:27 content that have been designed from the start. 23:32 That they can and will live on a variety of different platforms. 23:35 Because the semantic meaning and the structure behind them 23:38 is intended to translate wherever it needs to go. 23:42 And so to fight this war, we need to get away 23:47 from what I like to think of as container first thinking. 23:49 And you know, it's so easy for us to fall into this. 23:54 I mean, when you look back at the history of how we communicated in print. 23:56 Why, the simplest possible thing to do would be to figure out what 24:00 the page is, to draw a box around the thing that you were making. 24:04 And so that's why when you open up Photoshop, the first thing 24:08 it asks you is how big do you want the canvas to be. 24:11 That's why when my dear friend Ethan 24:15 Markov describes this process on the Boston Globe. 24:18 He says we started out by identifying common break points. 24:21 And to Ethan's credit here by the way he also will remind you all. 24:25 Don't do this. 24:29 It's why people are now trying to claim that cards are the future of the web. 24:31 Cards. 24:36 They're like pages, but tiny. 24:38 >> [LAUGH]. 24:40 >> And, this is something, you know, when 24:42 it's like when you come from a design perspective. 24:44 You realize that okay, yes, we'd, we'd like to think container first. 24:46 But you know what, this often goes right down into the very bowels of the CMS. 24:50 I'm working with a client right now, where I, you know, 24:56 I get to look at a lot of company's content management underpants. 24:58 And, so I get to go rummaging around in there and, and I was 25:01 like, hey, can you guys send me a list of all your content types? 25:05 And they're like sure. 25:07 And I get this list, and it, it's like, I don't know guys. 25:08 Some of these things are not like the others. 25:13 I mean, at the very foundation of how they 25:16 describe what their content types are, what they are publishing. 25:20 They mix the idea of, well, we publish FAQs or press releases or articles 25:23 with the idea that we publish accordions or carousels or wrappers. 25:30 One of this things are content types, the other are just display types. 25:35 And yet, we honestly, like even in the systems that 25:38 we have, don't have a good way to separate them. 25:42 And the way that those things get separated 25:45 is a process that is call, content modeling, okay. 25:47 Content modeling is what makes possible the much 25:51 heralded NPR Create Once, Publish Everywhere, COPE model. 25:54 Content modelling is what allows my client SVA to have their digital screens. 25:59 And also have their webpage. 26:04 And to have the objects that are attached on 26:06 this web page render appropriately on a different screen. 26:08 And I can totally imagine. 26:12 I can look at a web page like this 26:14 and imagine that everything is managed as one giant blob. 26:15 Which would limit them. 26:19 Because there would be no way for them to pull out the time and date of the event. 26:20 Or the short summary of the event, or the image attached to the event. 26:27 And then have that render appropriately in a different way on that screen, 26:32 if all of that were stored as one giant blob in their CMS. 26:36 It's what makes possible this example from The Guardian. 26:41 So, here, this is The Guardian's beta site. 26:44 They have stories that are shown with an image and a short summary. 26:47 And then at a smaller break point they remove the image. 26:51 Here they've got a row of four stories, images and summary, images and headline. 26:54 And at a smaller break point just for the two most important stories 26:59 they keep the image, they get rid of it for the secondary stories. 27:04 This is a simple example, but you realize that, oh, if we 27:07 wanna have the right hierarchy, the right granularity in our responsive designs. 27:11 So that we can render the right amount and priority of 27:16 information so that it works well across a variety of break points. 27:20 We're gonna need to make sure that that 27:25 granularity is baked into our content creation process. 27:26 And so, the content modeling process involves first, 27:30 figuring out what type of content you have. 27:34 Is it an article, a product spec, a recipe? 27:36 Is it a medical condition, a radio program? 27:39 I worked with client, a hospital client this past year. 27:41 And one of the things we did in a brainstorming session, took me 27:44 half the day, is I asked them to set down and tell me. 27:47 What are all the content types that you think you publish? 27:50 And you know what they were very easily able to make a 27:53 business focused list of all of the things that they thought they did. 27:56 And so given you know I actually think that figuring out content 28:01 types can be a fairly straight forward part of a content process. 28:04 Beyond that though, organizations have to figure out 28:08 then what are the attributes of that content type. 28:11 And I often describe this as what are the fields that 28:14 are attached to that, or what are the objects attached to it. 28:17 I worked with Food and Wine magazine just this past year right here in New York. 28:21 And we worked through a content model for what their recipes might include. 28:26 And so you might think, well, recipe, 28:30 that's a really standard structured content, right? 28:32 You start with, I don't know, the concept of the recipe at the core. 28:35 Recipes have ingredients. 28:39 They have steps attached to them. 28:41 In this content model, they said, well, we're gonna have a category 28:43 for each recipe and we're gonna have chef attached to each recipe. 28:46 One of their business focuses is that they were really focused on the people of food. 28:49 So chef was an important organizing principle. 28:53 And then, we came up with a whole bunch 28:56 of different facets that would attach to each recipe. 28:58 So they could categorize the technique used or 29:01 the cuisine was a German, Greek, or Italian. 29:05 Did it, what require, was it a special diet like gluten free? 29:08 Was it appropriate for holiday like the Fourth of July? 29:11 What methods did it use? 29:14 Did it require a slow cooker or a presser cooker? 29:15 And so, you might look at this and say, oh, okay, Karen, well, great. 29:18 Recipe content model solved, right. 29:21 But the thing is, this process, this is, is actually a strategic process 29:24 that tries to get at the essence of what an organization wants to do. 29:29 So to contrast, the Lark cookbook was a kickstarter funded 29:34 project to create a website, a mobile app, and more than that. 29:38 A printed cookbook that was all derived 29:44 by the same drupal based content management system. 29:46 And so, my friend Jeff Eaton talked to these 29:50 people and talked about what their content model was. 29:52 Now, predictably, they started with recipes as well. 29:55 But below that, even for this incredibly structured content type, what 29:59 they modelled for their business and their goals, was completely different. 30:03 So for them, the concept of the plate was actually the primary organizing principle. 30:07 This chef believed that these recipes should all be cooked 30:13 together in one plate, that would have several different items. 30:16 So the plate was each attached to a particular season. 30:21 And even more interestingly, this chef 30:25 believed that there were only three seasons. 30:27 So, every plate, every recipe had a season and then 30:29 a wine varietal that would be attached for the correct season. 30:32 And then their recipes had steps. 30:36 Those steps had ingredients attached to them. 30:38 But they also had techniques. 30:41 Because one of their goals was to create a media library of the chef himself. 30:43 Showing people how to perform different techniques if they weren't familiar. 30:48 And so this, you know, to me it's like, this is a 30:52 great example of how challenging it can be to communicate a concept model. 30:55 To break down content from its form. 30:59 Even for something inherently structured. 31:02 And when you look. 31:05 Like, I, I don't know about you guys but for me, when 31:06 this New York Times innovation report came out it was like Deep Throat. 31:08 And so one of the, like the big smoking 31:12 gun in the New York Times Innovation Report for me. 31:14 Was when they said that they weren't able to achieve their business goals. 31:16 Because they didn't have the right structured data attached to their content. 31:21 They said the ugly truth about structured 31:25 data is there are substantial costs to waiting. 31:26 They go on to say, for example, because our 31:30 recipes were never properly tagged by ingredients and cooking time. 31:33 We floundered for about 15 years trying to 31:37 figure out how to create a useful recipe database. 31:41 We can do it now, but only after 31:44 spending a huge sum to retroactively structure the data. 31:47 Now to me, this isn't just a problem of what you do your recipes. 31:52 To me this challenge really goes to like, huh, what sorta data? 31:58 What sort of stuff do we have blobbed onto web pages that 32:03 we're not going to be able to get at and re-use more effectively. 32:06 Because we don't have the right structured data attached to it. 32:10 And so to give you an example of this. 32:14 To give you a real world scenario here, let me talk 32:15 about the blobbiest blob that we have ever blobbed onto the web. 32:18 The PDF. 32:24 I mean, I feel like generations past us are 32:24 gonna look back with a sense of shock and horror. 32:28 Where they're gonna be like, you made a piece 32:33 of paper that you could look at on a screen? 32:36 Like, that's how you decided to solve this problem? 32:40 So, I wanna be clear about one very 32:43 important thing before we get started, and it's this. 32:44 Nobody's reading your PDF. 32:48 The World Bank recently went through and did an analysis of 32:50 all of the PDFs that they had published between 2008 and 2012. 32:56 And what they discovered was that a third of 32:59 those PDF's had never been downloaded at all by anybody. 33:03 And 40% of them had been downloaded fewer than 33:06 100 times, many of them presumably by the author's mom. 33:11 And when you start peeling back, it's like okay, so what's going on in human society. 33:15 Where we feel the need to publish a big piece of paper and put it on the web. 33:22 Ethan and I just went and spent some time 33:29 with, working with the people at the Boston FED. 33:31 And the Boston FED like many research organizations like 33:33 the World Bank has multipage PDFs that just go on. 33:35 And they've got charts and tables. 33:40 And they've got equations. 33:43 And they've got graphs embedded. 33:44 And, and you look at this, and you think, what's the 33:47 workflow that forces somebody to have to create a document like this? 33:51 Well, it's not digital, I mean that piece of paper is something that is 33:56 not information that is stored in what I would call an inherently digital format. 34:01 It's not responsive. 34:06 I work with a lot of banks, maybe some of you do to. 34:08 And there's no better story to go tell a bank than, 34:10 hey, all this valuable information you have locked up in a PDF. 34:12 It's really not going to render very well on a smart phone. 34:16 It's not searchable. 34:19 I mean, there's things you can do, but let's face it. 34:20 When people are talking about searching the web, they are 34:23 not talking about find information that is embedded in PDFs. 34:25 And it's not accessible. 34:29 I mean, the, the tools that we have to create 34:31 truly accessible content online are stymied by having to use PDFs. 34:34 And again, there are things that can be done. 34:39 But a better solution would be to fix the PDF problem. 34:41 But how do we do that when the root of the 34:45 PDF problem is not that somebody wants a piece of paper. 34:48 It's that they have a workflow that allows them to use a familiar tool. 34:51 They have a very simple process like the researchers for work 34:56 at one of, at the Federal Reserve Bank or the World Bank. 35:00 These researchers can use Microsoft Word. 35:02 Save it to PDF. 35:05 Send it to somebody else and have it instantly appear online. 35:07 The work flow for creating one of these documents in 35:11 what is true digital text would be so much complicated. 35:13 They can embed the kinda charts and tables in way 35:19 that, yes, you know what, that's just a big graphic. 35:22 For them to actually embed those charts 35:24 and tables in a truly digital responsive format. 35:26 Like those are complex content management 35:30 challenges that simply haven't been solved yet. 35:33 And I think probably the root of this is that the PDF supports 35:36 something, that all of the content 35:40 modeling that I've talked about simply doesn't. 35:42 Which is that it supports an unstructured flow of a 35:45 longer narrative document that has other content types embedded in it. 35:48 At random and unstructured places. 35:53 My friend Jeff Eaton published something on elist part not that 35:57 long ago where he called it the battle for the body field. 35:59 And describes the fact that while many content modeling processes work 36:02 well for information that is or can be inherently structured like recipes. 36:06 They don't provide the tools that are meaningful enough 36:12 for content creators who need to create a narrative flow. 36:15 And he said that while standard HTML is 36:18 rich enough for a designer to represent complex content. 36:21 It isn't precise enough to describe and 36:25 store that content in a presentation-independent fashion. 36:28 What we have here is a mismatch of language. 36:33 We have a mismatch of vocabulary. 36:36 And so, it requires us to continue relying on 36:38 tools like the PDF rather than having truly structured content. 36:41 Having workflows and content management technology that 36:47 will support having digital accessible responsive information. 36:49 So let me talk to you a little bit, about 36:55 what the cost is, for our organizations in doing this. 36:56 I spoke recently to somebody who just completed a redesign, 37:00 for a large, professional services company, here in New York. 37:02 And when they did, as they went in and they said, we've got a big pile of PDFs. 37:06 And they said, they, and we're searching just as I've said, 37:10 we've got 6,000 dish PDFs that have valuable information in them. 37:13 And we wanna make sure we can get that content out 37:18 so we can use it in a digital format on our website. 37:20 And so they went through a process where they started with modeling. 37:23 So they started by doing exactly what I would describe, where 37:26 they said, let's try to identify what these content types are. 37:29 What the attributes of them are and see if we can get 37:32 that in a format where we can manage it in our CMS. 37:35 Them, once they done the model. 37:38 They went back and they looked at some new PDFs in a pilot program. 37:40 So they said, they were essentially the model against untested model 37:44 to make sure that the model stood up and made sense. 37:48 This process just to evaluate and create the model cost them more than $100,000. 37:51 And this process took about a month. 37:59 Once they were done with this then they did 38:02 the work that I like to call copy and paste. 38:04 Where they went back through, they dug through those 6,000 PDFs. 38:06 They had a team of people, three people working for five months to pull 38:10 the text out of these PDFs and represent it in a meaningful, structured format. 38:14 That work cost them more than $200,000. 38:19 And so, by the end of it, they'd spent six months with a total of 38:22 eight people and more than $300,000 to unlock their valuable content. 38:27 That was trapped in these 6,000 PDFs. 38:33 In comparison to that, they had another 25,000 38:37 PDFs that they thought might be interesting information. 38:40 Possibly somebody might want it sometime. 38:44 It wasn't worth the effort of going through and truly 38:46 modeling and pulling that content back out of the PDF. 38:49 So instead they went through a process that I like to call shoveling. 38:53 And 38:57 so they had one developer work for one week and to get 39:00 those 25,000 PDFs on the internet cost them a bargain of $2,800. 39:03 And the end result of that is most likely that they 39:09 have yet they have continued the giant dumpster full of documents. 39:13 That no one is ever going to access. 39:18 I think this is a great case study because it 39:20 shows the value of creating this kinda structured content, starts now. 39:21 Like the challenge that we face. 39:27 Our future depends on our ability to escape from the blobs. 39:29 Our future depends on not having content locked up in PDF's or worse. 39:33 Having, not having content locked up on unstructured web pages that were not 39:38 going to be able to get at in a meaningful way in the future. 39:42 Our future depends then on having this kinda structured content. 39:46 And the time to start doing that is now. 39:50 The New York Times, if they had started 15 years ago 39:53 instead of continuing to do things the way they've always done it. 39:56 They would be 15 years ahead in their plan to have a recipe database. 39:59 If these banks or this professional services firms stopped making PDFs now. 40:03 Ten years from now, when they need to get 40:08 their pl, their content onto audio interfaces or wearable devices. 40:10 They'll be in a better place. 40:14 Our future guys, depends on fighting off the zombies. 40:16 I guarantee you, if there's one thing that I can promise you. 40:20 Is that humanity's desire for new technologies, and 40:24 devices, and platforms has not yet been sated. 40:27 There will always be another device, 40:31 another platform, another screen size, another resolution. 40:32 Our ability to survive the zombie 40:36 apocalypse means investing in structured content today. 40:38 Thank you. 40:43 [SOUND] 40:44
You need to sign up for Treehouse in order to download course files.Sign up