 Everyone, let's start with some good news. I'm not the speaker on the session. We have a bit of a better speaker for you, but I just quickly wanted to make a small introduction. I'm Kun Plateo, I'm the track chair for the Conductor Agit Track, first one ever at Drupalcon Europe. So how are we finding Barcelona so far? Everything okay, good place for an event, isn't it? It's actually such a good place that last year we had CONFAP, one of the biggest events when it comes to CONFAP strategy, also hosted in Barcelona. And I'm a Barcelona local, but I missed it. I missed it because I was at Drupalcon Amsterdam, which was happening in the same week, which is kind of a pity, but hey, I wanted to be with the Drupal community, you know? There was one CONFAP strategy talk at Drupalcon Amsterdam, one out of 90-something sessions. And the session got canceled because I think the speaker lost his flight to something, was replaced by something about accessibility. So I went back to the office, wrote this giant rant on a corporate blog, saying how can the Drupal community forget about common strategy altogether? It's one of the biggest trends when it comes to web communication and we're just doing as if it's not happening or if it's not important. So a couple of months ago, I got a phone call from a Drupal association saying, hey, if you really think that we're not doing good job, why don't you do it yourself? So I got commissioned by finding some excellent speakers to speak at a dedicated common strategy track, which is, hooray, but then at the same time, okay, but now where can I find these people? And we all know people that are quite vocal in that sphere in the Drupal community, people like Jeff Eaton, everybody has seen or heard speaking. But I thought, let's try to find somebody with a bit more of a local thought leader, if you want. So I started looking around and didn't have to look very far because one of the people, one of the thought leaders that we know quite well in Europe is Nonsurbina. He's the author of the book, a common strategy connected in dots between, and I need to quickly speak, between business brands and benefits. Probably many people have read or seen the book. So he's also a quite well-known speaker when it comes to common strategy events. So he's quite an authority out there. Being, as we are at a DrupalCon event, it might be that not for everybody in the room, he's that well-known, but believe me, it's a really opportunity to see and hear him speak. So apart from being an author of a great book, which I actually use quite often in my work, Nonsurbina also has his own consultancy, so sure if people have business inquiry, he will not be angry if you contact him. So yeah, I'll dispute it very quickly from the stage so you can listen and learn from him, like many of us common strategies have already done so in the past. All yours? Thank you very much, Kun. Thank you very much everybody for coming. You heard a little bit of introduction there, I will do a brief little bit on my background as well. My name is Nonsurbina and I'm doing the non-terrifying intro to semantic content. So you got the key stats from Kun. I'm a content strategist and modeler. I have spent about 14 years in the content world, specifically in the semantic and structured content world, which is kind of what we're gonna be talking about here. The role of semantic and structured content in the market has changed significantly since I started. It's kind of been an uphill battle for a decade and a half and then about 24 months ago, suddenly life became much, much, much easier. And we're gonna talk about what's happening in the market and why that is. I like to say that I'm an H to H, that is a human to human because most of my projects have been in business to business, microchips, medical devices, technology products. I like to say that I'm a masochist, so I take the big, painful problems that no one else wants to kind of deal with. So, but I do come to those clients, even if you're selling brain scanners or microchips. And I tell them, you still have a lot to learn from business to consumer. And business to consumer has a lot to learn from business to business and multinationals because even small business to consumer businesses are going global and they're trying to sell at scale and the internet is a scale operation. And simultaneously, in business to business, we're realizing that relationship building, branding, personalized communications are all just as important when you're selling to a procurement team as it is when you're selling to an individual. Okay, so today we're gonna go through trends in the market, fixing our content with semantic models, and then what to do next. So, some tips and some take-homes for you to go and apply this in your work. Before I get any further, I will point out that I'm used to working in conferences that are more of a mix between the actual brands and the agencies and service providers, whereas Drupalcon is much more kind of the industry and the service providers talking amongst themselves. So, you may see me kind of slip on that a little bit, but I am gonna talk to you as my colleagues as service providers, but this is for you to kind of take these messages home to your clients so you can get bigger, more interesting projects and they can get more benefit out of what you provide. So, the big picture, I like to start with a kind of a big context thing here. I like to talk a lot about the idea of content karma. When we're talking about content, the more real value we give to consumers, the more value that will come back our way. The internet and communication and sales and marketing have evolved significantly in the last 20 years. And what we're moving to is from a world of publishers putting out their message and trying to persuade consumers to the engagement model, where we're trying to build ongoing relationships with people, accelerated by social media, to be able to have a back and forth with our consumer. And as we progress into platforms that can provide us more and more communications capability, our audience becomes more and more resistant to those communications. So we're seeing things like the discussion of the death of the banner ad. We're seeing discussion of the death of traditional advertising period. All because people are becoming more and more resistant to any sort of commercial or self-interested message. So organizations are realizing they have to be more customer centric and focus more on how they can deliver appropriate and relevant value to consumers rather than just kind of showing off with celebrity endorsements and pretty branding and awesome logo design. We have to really deliver substance and value if we wanna get value back. But we have to do that on a huge number of channels. So the more we can make our content adaptive, the more we can realistically deliver tailored, high value content without running out of budget, resources or time. So this talk at the highest level is we've got a huge diversity of audiences with a big diversity of interests and needs. So how do we give them all a personalized, relevant experience on the device that they're choosing without completely going out of business? So I'm gonna look at the difference between a search result today versus even five years ago. So there's a supermarket called ASDA. It's a supermarket in the UK where I do a lot of my work. And Sutton is a city in the UK. So if we come up to the top here, if I have a mouse, we've got what time does ASDA Sutton close? So what time does the local supermarket close? Five years ago, if I put that into Google, I would have gotten a list of links. It would have given me different pages based on their keyword ranking for the terms that I'd put in. And then I could go on those pages and go get my answer. Now I get that in the browser. That's not a webpage. That's an answer to my question. That's what I actually wanted as a user. We all make something web related, web pages, or some part of the web page machinery if we're at this conference. But funny enough, people don't actually want web pages. Websites and web pages are not interesting. People don't go online for the purposes of finding web pages. They go online to have experiences, to connect with other people, to satisfy some need or answer some kind of question. So the more we focus on what they're actually going on for, the more successful we'll have, the more valuable we'll bring to them. So here, the web is not a series of index pages which people can go and read. They're not passive. They are proactive systems that deliver the right dose of information to the right person. And they do that by leveraging metadata. So the rest of this talk is essentially about metadata and how you can do it too. People are asking all sorts of natural language, non keyword term based questions. How to get one of X? Where should I go on vacation? In the business to business world. I'm on a factory shop floor and there's a machine that's not working. So what troubleshooting is needed for this machine that has these modules installed that's at this point in its maintenance cycle? I'm evaluating a new circuit board to integrate into my product. So does this have the specs that I want? Does it have the operating parameters? How is it better than this other one that I might use? And they're entering more natural language questions and much more complex questions than they used to. As the web gets more sophisticated, we actually get simpler in how we use it. We expect more from it so we just talk to it like a person. And we wanna give very targeted, very specific answers to all these kinds of questions. But we wanna do that targeting on this endless plethora of devices. So we've lived through the movement from the desktop web to the mobile web. We have the Internet of Things coming out. We have one of my personal favorites which is wearable technologies and the ability to augment the real world around us with digital overlays of content. So the role of the web page is changing. We're moving from managing web pages to managing digital content that may or may not exist in pages. So how do we manage the same answer to a question that might on one case be delivered out as a web page, but in another case be read into someone's ears through text to speech? Brands are starting to ask this question because they want to be able to deliver answers on all these different device formats and channels. 94% of businesses also not only wanna do all these channels, but they're saying they want to personalize the response. So where are you? What time is it? Do you have a purchase history with me? Have I never seen you before? Is it raining where you are? All these things are actual questions that have come up with me with clients. So they wanna do promotions or communications that are based on ambient geographic or weather or situational data for the people in that location. Or they wanna personalize based on what they've seen you click on where you came in from in terms of URLs, what other queries you've done on their site or off site. So personalization is big business. 48% of shoppers would like to use a phone to shop while in stores. So I use this as an example to remind us that we can kind of create these user experiences on a website, but what is the context in which that website is being used? So the consumers are telling the retailers, we wanna be able to use the web as a physical guide while we're walking through a physical environment. But not enough designers are thinking about those two scenarios. How do I get answers while sitting at home? How do I get more relevant answers when I'm standing there right in front of the product and I have a sales agent five meters away from me? 74% of shoppers who show room are older than 29, 48% are older than 40. So when we talk about this combined digital, physical purchasing process, we kind of think automatically of hip and trendy millennial shoppers on smartphones. That's not actually accurate. You don't have to be 21 to realize that if you walk into a store, get to actually touch the product and go home and buy it on Amazon. That's just a smart way to shop. You don't have to be a young, hip and savvy person because you'll quickly figure out that I can actually see what I want and get the benefits of the physical experience then get the benefits of the digital experience back home. So everybody is doing this regardless of age. So this isn't up and coming. This is very present for any generation. How do we optimize for crossover contextual mixed environment, mixed device experiences? And then it's not a business consumer thing either. I use lots of consumer examples because we can all relate to them. But 66% of B2B suppliers say customer expectations are driving them towards omnichannel. So customers don't expect great website, crappy print materials. They don't expect great pre-sales materials, crappy post-sales materials. If you communicate to me as a brand, then your website, your print materials, your pre-sales, your post-sales, your training, all of that has gotta be working together as one face of the company. So although your businesses will be in the support of the web and digital channel, ask the questions. How does this integrate with all of these other touchpoints? Have you thought about that? It's a way to differentiate yourself as a supplier. And if you're a brand, they know that it's a way to differentiate themselves to their customers. If they can tie all of these different channels together and deliver a unified experience, then they are gonna be more successful than their competitors. All these different pressures, all these different trends, I think are leading us to what I call the smartening. A new era where we are really looking at not only what data we're creating and what content we're creating, but metadata. This is the first great age of metadata. We have to fix the content that we're creating. And what that means is making it format agnostic. There was a lot of work done in the last five to 10 years about writing for the web. Now there's a new trend of writing for everywhere. This is, Kuhn mentioned that I was at CONFAB. Christina Halverson, who's heard of Christina Halverson? Okay, so Christina Halverson is one of the bigger thought leaders in content. And she kind of got the term content strategy and content strategy for the web popularized. It was around for probably a decade before, but she was really kind of a watershed author. And she wrote that book, Content Strategy for the Web. Now, I think about seven years later, she just released Content Strategy for Everything. Because we don't have to write for the web anymore, we have to write for all of the different ways that our message could appear. And as digital content managers and digital content solution providers, we have to be designing for all the ways that that content could appear, not just for the web page, not just for the web experience. So we have to be ignoring the format of output and worrying about the message itself. To be able to deliver that personalization, we need to be able to take all of our little pieces of content and break them up into components, which we can then put together automatically or dynamically to give the impression that we're giving a fresh personalized answer to every individual. And we're gonna need a lot of semantic metadata. Semantic metadata that defines the content and describes the content. So what is it and also who's it for? When should it be shown? What context is it relevant for? And as well as audience, applicability and context metadata to decide when and where to route it. So when should we filter out this content? When should we show that content? And I am really talking at a much more detailed level than I think some of you will have worked before. So I'll show you some examples. So what does fixing the content look like? We're gonna go through some key concepts. Component content versus traditional content, metadata, content models, semantic versus structured, this should be controversial, taxonomy and link data. So I mentioned we want to break up the content. So when we break up into format free components, we go from managing our PDF downloads and our weighted pages to smaller modules, each with their own structure. What that does is it changes in a content management system our unit of use from documents or pages to components and then things that make up components, which we would call fragments or even variables. So if we're talking about a blog post, was our unit of use before, what is the structure of a blog post? What are the things that I can reuse from other deliverables and put into the middle of a blog post and just plunk them in there? Kind of like a content image in the sense that we take an image and we take it from our image library, we slap it in. Why can't we do that with text? Why can't we have an anchor link that points to a table or a bullet list? This concept of we're gonna take these little pieces of text and bring them together in more complex ways if we model them instruction properly. So we're gonna go deep, deep, deep into the content and start breaking it into smaller and smaller components. And that's what gives us the term component content management, which some of you will probably heard. There is a big movement towards this space right now. A lot of the big content management providers are either becoming component content management systems or saying that they're becoming component content management systems. This componentization gives us what I call modularity. Well, I shouldn't say I call. What is called modularity? The point of this is that we are moving to having a bunch of web pages to having a pool of modular components where we can pull out a certain set of components and deliver them for scenario A or for scenario B or scenario C from the same common pool. The different scenarios could be different devices. So you could be talking about I'm gonna pull this out for delivery on the desktop web or I'm gonna pull it out for delivery in an app or there could be different people. This is a new customer. This is an existing customer. This is an existing customer who has registered to our premium service and deliver them the right package of components regardless of device or location they've asked. Also, if we do things like define a number of components as an assembly, we can then reuse whole assemblies. So if I'll give you an example of this, I've got clients who want to be able to do a product overview brochure and they wanna be able to take then seven product overview brochures, wrap them up and deliver that as an offline ebook catalog. So instead of delivering to the web, we've actually packaged that up and put it through InDesign to make a nice proper print quality PDF with print quality graphics and so on and so on. But they only define the set of modules that makes up a brochure once. And then regardless of channel or output, they can just take that and say, okay, that was a brochure that was making a little microsite and now it's a PDF for an ebook reader. Once you've got all these little bits, once you've got all these components, you need metadata. So you all know what metadata is, I'm assuming, at DrupalCon, yeah. So you've all heard data about data, but I think there's more to it now and I'm gonna show you some of the different uses of metadata. So metadata is everything about the content in it and around it that describes the content. So what I like to say about metadata is it's what you use to stitch back together all the little components once you've broken them apart. Because you're not gonna manually reassemble all these things and manually define all these different relationships again once you've broken all the content into little pieces, you have to use metadata to enable automation. The question is where to put it, at what level, how much of it? So we started with documents. Back in the day, we had our document or our PDF or deliverable and we could put different metadata on that. So if we had different audience tags, we could tag it as being for her or for him or for mobile display or for USB gadget buyers or whatever our metadata flags were. And we could put that at the info product level. But once we've done our breaking up into components, we've got all these little bits of content running around and so now they all need metadata too. So we can have a whole module like an entire blog post or a whole product overview. We can have assemblies like an entire brochure and the content for a brochure might form part of the navigation of a site but it also might break off on its own and form the navigation of a single brochure output. But then we've also got smaller bits of content which we want to manage. So a single paragraph, an obvious example would be a note or a legal disclaimer but it can also be a reusable paragraph from a product line that actually appears on several product's pages. Or inline, individual tags. I've had projects where authors wanted to be able to write content and then they were talking about boats and they were writing about these boats but the boats are in motion. So they wanted to be able to talk about the boat and then inline in the sentence when they were talking about the boat, its current location, when they mentioned the location would be up to date with its current actual location. So as they wrote and they would mention the location of the boat or the amount of cargo it was carrying or how fast it was going or all sorts of things about the boat and every time they said something like that, they would drop in a reference to a web service that was providing that data in real time. So every time you looked at the page, all the inline references would be talking about that boat at that moment in time when you loaded the page. So we can be managing various tiny little bits of information. Another obvious one is product names or model numbers. As you probably have experienced, product names are kind of a random, mushy, squishy thing that marketing may decide to change at a whim. So I've had clients who, in one case, they were running around and manually finding every place they mentioned that product name or that model number. Other clients who had already gone through the migration to this kind of content, they were able to change a couple of variables on the whole site and the print documentation just refreshed itself. We can do this with tagging. We can do this with short codes. I'm sorry, did I just use a wordpress term? Sorry. We can do this with kind of hacks or workarounds or little techniques like that. But if we're just kind of giving all this power to the users, as you will well know, with great power comes great responsibility. We don't want to enable people to just tag and mark up as they want. We have to have an underlying logic that's guiding this and controlling when and where and how metadata is applied. All that guidance is what we call the content model. So if we come back to our example here, how did we do that? And this example really works, by the way. For some of you in the audience, it may actually be quite obvious. But take this example, show it to clients. I've had this work as an aha moment. You can't believe how many times. So we've got my answer in the webpage and clients go, I want to do that for my customers. So I say, fine, you want to do that for your customers. The implication then is on the back end, somewhere there's a structural content model. Someone has defined everything that makes up a store. And then once they've made those rules and said that a store, is that flashing? That a store has not only the content that we write in paragraphs and headings, but it also has a specific location, which is defined in a specific way with city and town and hours and a schedule. We can then tell everyone, go create the store content. And they can create well tagged machine readable descriptions of a store. And then map that out to, for example, schema.org metadata. Who knows schema.org already? Okay, so that should have been 100% of the room. Schema.org is an international standard for shipping metadata out with HTML. So HTML used to be the place that metadata went to die. Once you're in HTML, all the good meaningful data smarts was gone and you just had titles and headings and P's and ULs and all this stuff, which is useless to machines. So with schema.org, you can have persistent metadata, which leaves on out into the public web, where your tools or third party tools, or for example Google, can reach in and pull out that specific answer. Because it can identify the part of the content, which is the store name, its location and its hours and construct an answer. So that is semantic metadata in its most common popular sense. People then ask me, why did you define any kind of model internally? Why do we have to have a semantic model? Why do we have structures? Why can't we just write in schema.org directly? And the point of this is not so that we can satisfy Google and schema.org. We wanna be able to publish to something that Google and search engines will be able to scrape, but our metadata is about our products. It's according to our business processes and the way that our customers wanna do things. Some of that will overlap with schema.org. Some of it won't. So you are defining the model and the metadata structures for your business and what your customers wanna experience, not what schema.org wants and what not what schema.org supports. So we don't wanna write specifically for them, we wanna write for us. For me, I make comparison, compare it to when we started writing for search engines. We were trying to game them and write search engine friendly text instead of user friendly text. Same thing. We're not designing schema.org metadata, we're designing our business's metadata and then mapping the schema.org when necessary. So you put all those together and you have what we call intelligent content. There's an intelligent content conference that I speak out at a regular basis. And the reason, I like the term intelligent content. Schema.org is not a very friendly name to anybody outside this room. Tim Berners-Lee says linked data. I don't think that content people like the term data. I don't think they like having that content referred to as data. So intelligent content is a much more friendly name and it's actually a much more human name. This is again an example that I use and I use it very graphical examples and I invite you to do this with your clients. Take their content and show them the intelligence behind it. So reverse engineer it. What we do is we take all these different bits and we ask them, what do you think those are? And in this case based on layout and formatting, people can usually guess, like within a few minutes I can get any group of over five people to predict what's gonna be on my slide. They're gonna say product name, tag line, they're gonna say point out the feature list. They're gonna be some discrepancies but we all kinda get this and we all kind of get this because we're familiar with it. We've kinda grown up now or we've used the web enough to figure out whether these things are even if we can't read the text because we understand the intelligence behind it, the logic behind it and it's very simple. We've got our product overview, our product name, model number and so on and so on. And so this is a way for you to bring the idea of structure and models back to clients and say your content already has structures and models. What we're trying to do here is document that and implement that in the content management system so the content management system understands your content structure and models because once the content management system understands it, then we can manipulate it and reuse it and componentize it and do all this fancy stuff but we want to get that model established and agreed so everybody can work against it. Not only will your content management system be more consistent in itself, then authors become more consistent. Has anyone read Jeff Eaton's Battle for the Body field? No, good article. Jeff Eaton, really great guy, one of the biggest advocates of this in the Drupal community and he wrote this article called Battle for the Body field where he's talking about semantics inside the body because as we know, you've got your nice field control data where you can control the metadata and then you get to this body field and then users go nuts and they put whatever kind of junk in there. So the body field is somewhere where we can now start to go in more detail and put in more guidance and more control not to create a kind of like super form hideous experience but something that has this metadata around single words, around single bullet items. I like to say there's no reason that when we write a feature in a list of features, it becomes a list item. It's a feature in a feature list. So we should tag it as a feature in a feature list, not tag it as an LI in a UL. We wanna keep that intelligence and keep the human intelligence evenly righted down. Once we do that, all our content becomes more consistent because our authors are more constrained and they're better guided and then your consumers get benefits because with more consistent content going out, they can read more easily. They can scan more easily. They understand your content better because it's more consistent within itself and that makes them happier. As beings, we like consistency. You saw the keynote hopefully this morning. I really enjoyed that. I'm also a big cognitive science and psychology person when it comes to web content and I say we are sense making machines. I give an example. Oh, yeah, okay. So I've got a diagram here. Can anyone tell me what my diagram is of? Okay, now? Anybody? Yeah, it's a face, obviously it's a face. So duh, yeah. But it's really, really not a face. Well, there's a hell of a lot more information on the right than there is on the left. But you would have to actually have a serious brain injury not to recognize that as a face. As beings, we use data compression. Our minds compress and take the fundamentals from the right and match it to the model on the left. All the content models are, all the semantic data is, is taking a simplified model of a complex reality. This isn't about Drupal. It's not about data science. It's about making machines work in a more natural, more human way. This is how we already operate. I've got a whole talk specifically on that called the Biological Imperative for Intelligent Content, which you can check out. And I'm gonna go over here now. Okay, once you've got this, once you've got a nice structured metadata model, you can then use that to interconnect different teams. So we have this rift between UX designers and developers and content people. UX designers are creating these beautiful wireframe designs and then we have to squish this awful, inconsistent, mushy content into it. What we're doing here is we're creating a wireframe for content and we can map from wireframe to wireframe and we say product overview. When we're talking about the product family page, that's where product names go. When we're talking about the desktop product layout, that's where they go. On mobile, that's where they go. And on print. Once we've tagged everything up, there's no reason that a print composition engine can't take that same metadata and flow out on a static page which is going to a printer. Same thing with model numbers, same thing with taglines. Taglines don't go on to the product family page. And again, this, like literally get screenshots, draw lines, your customers will get it. Like I started doing this and it seems ridiculous, but clients really connect with this and they go, oh, I understand the connection now between what I wanted and this model in the background and I can work with it better. What's a bit annoying though is once you show them one, they want you to do them for every bloody page and model you have. The content model is the backbone of cross media adaptive omni-channel strategies. It future-proofs you so that we've got all these channels handled and then someone comes out with Microsoft HoloLens which is what I call the product that Google Glass should have been. And then we say, okay, on Microsoft HoloLens, that's where taglines go. On a lot of these augmented reality or card-based displays or social media, you don't have any visual control. You just ship them structured content. You don't get to say what size it is or where it is on the page. The rendering system does that for you. You have to ship them semantically marked up content. So if you want to be future-proofed against all the kinds of delivery that are coming, this is not optional anymore because there are systems that are going to ingest your content whether you like it or not. Or they won't ingest your content, which is even worse. When your semantics are explicit in your content, machines can help you reuse, transform, translate and format. Okay, I'm gonna go through taxonomy very quickly. So taxonomy is probably one of the easiest ones. In this room, most people are already familiar. So we have our taxonomy, our grouping and categorization system. Each value in here is a taxonomy facet. The reason that I'm showing you this is because we wanna be able to take our product overview model and then map that to taxonomies. And say we've got a taxonomy of values, evaluator, advanced user, sales user. And we wanna be able to say for sales people, they don't need the feature details on initial load. That's a level two priority for them. Everybody needs the top stuff, but for advanced users, they only want the feature name. And for sales people, they don't want anything except for the key features. So we've written one block of text and then with the taxonomy, we've said certain parts of our model should go to these people. We could do the same thing with weather. This paragraph was written for to display when it's raining. So we're adding again, not only a structure, but then categorization to the bits of that structure so we can selectively show these bits or those bits to the different people. Once we've got all that together, we've got content that's marked up in a very granular way and you get situations where we've got our model and our outputs and we've got two different short descriptions. In one point of the interface, when the audience is a USB gadget buyer, we're gonna show that one short description, but if someone else logs in and the audience is him, then in the exact same place, we will substitute a different piece of content. So cement to content models make your content programmable for maximum reuse and relevance. So you're not loading a page as much as you're running a query with a set of parameters and that delivers a page according to who that person is and what they want. So I'm coming into the last little bit here, link data. I mentioned that's Tim Berners-Lee's term for schema.org and these public data standards. I like link data a lot because it's old. I mentioned at the beginning of this talk, I've been doing this for 15 years and no one cared, no one cared. I was such a niche player 12 years ago because people created print or they created web. That was it. You need to do websites, you got a web team and copy and pasted all the print stuff onto the website as fast as they could and it was only once mobile hit, people went, oh crap, it's happening again. What if it happens again? And so now that we have the Internet of Things augmented reality, data feeds, syndication, microsites, apps, people are going, you know what, this isn't stopping. Suddenly I become a really popular guy because people needed a future proof methodology for doing this and so when we talked about OWL even like 10 years ago, I thought OWL was nerd stuff. Now I'm realizing actually we do need a standard methodology for describing our stuff that we all agree on and that's essentially what linked data is. So it's now linked and endorsed, backed and endorsed by people like Google, Facebook, Bing, BBC and so on and so on. And that's what enables this. So we've got our store. If you go on to schema.org you can find out how they did that. There's an opening hour specification for the metadata and that allows you to do the kind of intelligent display that we talked about. And that's what drives all this fancy stuff that we're seeing all over the place now where search results are becoming more and more specific, more and more intelligent because they've got this metadata inside them that allows you to render it. So putting all together, you've got your content, you've got your models, you've got your users, you've got your scenarios, location, a certain event, are they at a wine tasting, a certain time of day and you've got your outputs. We wanna be taking all of these together. So when we plan a website, have we looked at all of these? Have we ticked all the boxes and said, we're considering content in the context of output and scenarios and different users and are we modeling appropriately for those end delivery experiences that we want? You don't have to do all this tomorrow. I've got a lot of clients who are banking metadata. They are starting to think in a more structured methodological way because they know that later the system will be capable of leveraging it. This is part of the strategy bit of content strategy. We have to do things as we're able to execute on today. I'll give you an example. So there's a podcast by the Content Marketing Institute. I do a lot of stuff with them and I did this podcast with them. So now your average podcast, you get on there and you kind of shoot the shit with an interviewer for a while. 45 minutes goes by, load the file up to iTunes, done. The CMI podcast has a structure to it. They have predictable sections of the podcast. So they have obvious stuff like the audio file, who it was, the series theme, and these all exist on the page there, a short description and the contributor. Then they have a section called New Now and Next, a section called Blast the Buzzword, a Q&A section, book recommendations, and final advice. And most of those sections will appear. The New Now Next are not optional, they're in every interview, and then the others are optional. They may or may not be there. So they have that model and if they enrich that model, I'll just almost skip this for time, but the idea is you've got this model and we put keywords on each bit of it. You could automatically assemble from the audio files a podcast which is all about the buzzwords. So what do all the interviewees say about buzzwords? Or about recommended books. We could do a recommended books podcast from all of our existing podcasts. Or what's up and coming in the market? Or let the user define something. Let the user say, this is what I'm interested in and have them return all content strategist podcasts from Europe, why not? The system, once it has that metadata, can pull that all together. And there's no reason that we couldn't reuse some of that information out to a print brochure or catalog when we're speaking at a conference and pull that same data out. So what you're essentially doing is you're writing this love note to the future where you have metadata which you can use some time when you're ready. They're not ready to do this yet but they know that someday the system will be able to do it. So they are structuring their content like that now so that they're able to use all that good metadata when their systems are caught up. So all doable because of consistency and not so scary semantics. So what you should do now. Go home and do an omnichannel readiness review or tell your clients that they should do one or they should pay you one to do one for them. Look at the people, platform positioning. How is content used? How does that client understand the role of content in their business? How is that working across channels and platforms? And do they have the people they need to do something more intelligent? Audit that state. Build some detailed multi-context journeys and stories. Show how it could be. Consider how you'll measure the success of that. And then do a small conservative scope where you can do a minimum viable prototype. And say we wanna be able to demonstrate how much value this could add if we do some key strategic changes. The impact can be quite substantial. If you're interested in knowing more, I do some workshops about this. I got two coming up, one in New Orleans and one in London. I put this up here not just as a plug but to invite you and say what we often do is we run them in your offices. So if you've got offices where we could run a workshop or you have clients who might wanna have workshop, we can run it there and invite other people and then yourselves or the client can invite free people. So if you wanna maybe host us doing a workshop, then let me know. Thank you very much. We do have time for questions. Also, Noss probably won't dare to say it himself but I know he's having some of his copies of his book with him so if anybody's interested. And I actually saw today because I had to quickly peek to that exact title. I actually saw that it's temporarily sold out on Amazon so that's your chance. Anyway, opening up for questions. So the question is, have I ever worked at schema.org directly to influence having model data on the site? Not as of yet. It's on my to-do list. I'm working already with a smaller standards body in a particular niche called DDA. And they're doing something called DDA for marketing and e-commerce right now. And that's very exciting for me because of my background. So that's kinda all my committee work used up. I can barely keep up with one committee. Committees are as much fun as you think. So the real difference. Schema.org is a general metadata standard whereas DDA is actually a modular metadata standard where it defines particular types. So you could define a blog post and define all the bits of a blog post and structure your inline validation rules in DDA. Schema.org tells you if you do a blog post, write this on it so we know it's a blog post. That's it. It doesn't give you the ability to do all that rich modeling and tell you users what they can do and what they can't do and where to do it and et cetera. Oh, sorry, can I just, I wanna put a very, so my summary of that is, Schema.org is a publishing standard. DDA is a authoring and management standard. Yes. Yeah, there is a bit of that. So for example, ASDA, they blew hundreds of thousands of not millions of dollars on visual design, CSS, UX, logo. I never even saw the bloody logo. Like, yeah, why I started this with a content karma message was in the same way that Tim Berners-Lee 15 years ago had to be going around saying, put your content out on the web. People will just find it and people said, well, I don't want people to just read my content. I want them to talk to my sales people. I want them to talk to me. I don't wanna just have them self educate and buy online. And he goes, tough. That's where the whole world's gonna go. It's kind of the same answer, is that people are gonna want those answers. And if you're the one who's not putting your data out there, then you'll be in the same situation as the people who didn't put their content out on the web. And then the only thing I would add to that is, is there some value add that your site can do, which is Google is not? Is there something unique and particular, I don't know what your site does, but there's probably some user experience value add that you can still add, which Google is not gonna bother investing in. Different content for different profiles, yeah. With technological changes, you said? Challenges, yes. So technological challenges of doing things like caching for very dynamic websites. Yes, that is a challenge. I am not a developer. So, for example, Jeff Eaton, that might be a good person to ask about that, because he will have been more tied into the technical side of enablement. I'm working on the strategy side. I'm saying, as an organization, this is what we can do. This is what the benefit is. This is how we get our people to do it. From a technical perspective, that's not really my area to say, how is it gonna be load balanced and efficient and all that stuff? I have a question more like a technical question. There's like two ways to integrate schema.org in your webpage. You can do it in line, like within the HTML. Also, you can use JSON-LD for it and to add it to the head of your body, of your document page. Is there like a preferred way? What's better? I don't think there is a preferred way. There might be an SEO preferred way, then you'd probably wanna ask an SEO person. From a creation point of view, provided you have a structural model on your storage and input side, it makes no difference, because you're gonna write it all down to the JSON, or you are going to write it in line. But it's an automatic process, so it doesn't really matter from a management point of view. So I think the only answer would be from an SEO perspective. I like in line just because my instinct says it makes more sense, especially if we're compiling that page from various modules, you want each module to have its metadata with it, as opposed to treating it as one block. We're getting away from treating anything as one block these days, so I would say that for my instinct, in line is better. Cool, thank you. For, sorry. Right, so prioritize the user rather than search engine. So what's the question exactly? Oh, right, so when I was showing the taxonomy, I was prioritizing for the user which part of the content is more important to this kind of user. Yes, yes, you treat the search engine as one more audience. So I have in some of my models social, and so the authors have to fill out the social content because they write titles, for example, which are not social friendly. So they have to write the title in the main part of the body, and then they have to write the social title so that when it's shared, it's a social friendly title, and then those things are mapped to that persona effectively. Do you have a tool that you use for modeling content? Like, you know, do you use like Vizio or do you use some kind of, how do you actually model your content? It's an up and coming tool. It's called Excel, I come in called Microsoft. No, I'm serious. Really, I do a whole another presentation on why this is actually so difficult, and it has to do with the fact that we're giving our content dimensions effectively. There's ways that we could look at it, ways that we could think about it that we couldn't before. So when I'm modeling, I end up using Excel because there's not really a tool out there which is kind of designed to allow me to look at this model in all the different ways I wanna do it. So if I'm exchanging and reviewing with clients, if I'm working with other content strategists, Excel just ends up being the best of the available options, the least of all evils, I'll put it that way for modeling content. And I've got some templates in it which I've been working on for the past couple of years and so moving to another tool now would be kind of a pain in the butt too. Do I create more than one model? Oh, absolutely, sorry. So the question is, are there more than one entity, a more than one model in a website? Yes, absolutely. So there's kind of this content types and then there's content model and I didn't go into all the subtleties here but actually the model is several types that have been put together. So we might have a simple type like a feature list or a promo box. And then a model might say, okay, if you have a short description paragraph and then a feature list and a promo box, that makes up a landing page. So you have your various different types which you have their own and you have rules for how your different types come together and that overall thing is your content model. So you may have dozens of types. I would try to avoid actually having dozens of types. If you can get away with seven, then great. And you don't have to model every little thing that is on the site. You can have, this is a page and it's about us. We only have one that we don't need a model. We just write it. So you don't have to model everything. You want to model everything that you're gonna process in some fancy way or everything you're gonna do a lot of so that they're consistent with each other. Anybody else? Going once? Going twice? Sold. Thank you very much.