What It Is and Why It Matters

Dynamic Delivery What it is and why it matters - White Paper Series - DYNAMIC DELIVERY Why we need it When you search for information, you are seeking insight. You gain insight when your information is contextual and complete. Contextual because it adapts to you and to the device you’re using. Because it focuses on precisely what you want, and understands what you need Content published as static based on the task you are trying to achieve. documents is incompatible with processes and apps Complete because you can see through formats needed to enhance and across silos—content, structured or not, from productivity. any relevant source in an organization, delivered uniformly, consistently, and immediately. You gain insight when you find precisely what you want, avoid the irrelevant, yet are offered intelligent suggestions of alternate pathways of learning. This is the goal of dynamic content delivery, or Customer Field Self-service 360 dynamic publishing. Support Maintenance Portal Product Info In the following pages we explore the principles It is impossible to make it and benefits of dynamic delivery. It will help you feed next gen interaction understand why you must seriously consider it, and channels. how to choose a solution. In-device, Heads-up Chatbot, Inline Help Display, AR Virtual Agent 2 DYNAMIC DELIVERY What it is Static publishing, what we’re Dynamic publishing delivers Content Maturity Level used to since the dawn of printing, personalized, customized, on-the- is pre-generated, one-size-fits- fly documentation. Content stored in static PDF ﬁles, all, always the same. You get one Static only full documents can be viewed It gives results that are more Content manual for a product, regardless or downloaded Findable, more Readable, more of its customizations or the Shareable, and more Interactive. specifics of each installation; it’s Contextual Granular & tagged information, the same manual regardless of What permits this is the marriage Information contextual delivery through web devices whether you read it on a phone, of custom-engineered technology a laptop, or an HDTV. Newer to documentation that is Active Deeply semantic content, consumable and increasingly popular delivery manipulated at a fine-grained nowledge through new devices and interactions vehicles, wearables such as AR level with consistent metadata. goggles or heads-up displays Together, these empower your are not designed to deliver static system of apps and workflows to documentation at all. deliver great user experience and Maturity supported by each delivery channel Static documentation needs to be to create product information evel rinted ortal In-appIoT RAR Chatbot gently laid to rest, with respect for tailored for the moment and heroic service, in that back room adapted to the channel. Static Content where you keep the fax machine. Contextual Information Active nowledge 3 Finailiy Reaailiy FINDABILITY And content diversity Most common tech content formats less structure, less semantics Multimedia, Video 1 Observers report that customers The insight the user seeks might Flat text, RTF, PDF, PPT, prefer to solve problems on their originate in customer support, or Word, Open Document .txt own. You can help them do this it might come from marketing. In HTML, Markdown, Docbook, with a portal that combines an a regulated industry, a technician html Unstructured Framemaker effective search engine with topic- might need information initially oriented documentation. intended for legal or government HTML Based help tools htm relations. Which silo the knowledge Over the last 30 years, enterprises Javadoc, PHP Doc, MathML, comes from is not pertinent, just API Ascii Doc, API docs have used typesetters, word speedy access, compliant with processors, and technical writers DITA, Author-it, S1000D, security constraints. </> with a medley of devices and MIL-STD-40051, XML, SGML software tools. The goal of all this Hence, your search should be Structured records from activity is to package knowledge format agnostic. Your portal databases, JSON from APIs for their customers, technicians, should process, transform, and users… none of whom care and enrich your content to more structure, more semantics whether the knowledge is packed make it consistently readable in a Word document, a spare parts and organized, by accepting catalog, a knowledge base, a wiki, differently formatted and sourced or a database of DITA topics. documents with varying levels of Most common tech content sources metadata. These are factors your customers don’t care about—how you format Formal Tech Doc nowledge Helpdesk Tools your documents, and where in CC ile ystem Bases ickets olutions your org chart they originate. ource Control it 1. Forrester 2016 Community Wikis Learning LM - IM and Forums Management arts Catalog System 5 FINDABILITY From the lab And large manuals We have experimented the impact of PDF publishing on findability: • Our own tech doc, generated as PDF (= 117 documents) and indexed with AFS, our proprietary state of the art semantic search technology. • Let’s run 2 different representative queries: The experience of using your documentation portal has to be at the level your customers are used to—relevant as Google, engaging as Facebook, collaborative as Slack, and as fast as them all. Or filtering replies they will quit you faster than a millennial leaving a Filter: version 7.5 chain restaurant. Results: 41 docs Having your documentation structured and tagged (35% of the corpus) (e.g., product, version number) at a fine-grained level gives your search more handles to find exactly what you want. When you can only search for entire books, the biggest books will always be the top search results. And you always get the same answer: Read the …you know… manual. That’s the very processing filters definition of poor customer support—no help at all. When your content is chunked and finely tagged, Filter: version 7.5 that big book’s boundaries dissolve and your search Results: 39 docs engine can dig inside and deliver dynamically exactly what your user needs, by working and indexing at (33% of the corpus) the component level. Why should a human pore over an index when a great search engine is so much faster at it? Findings: for both queries Analysis: the books are not the issue, Dynamic delivery relies heavily on fine-grained 1) same number of results, but the fact that they are indexed as content, whether it is created that way with a 2) same documents. one single element. modern component-based approach, or started life as large manuals that now have to be automatically Reason: large books tend to contain Action: keep the book as a means analyzed and chunked into small pieces. all keywords and will always be part of to organize content, but change the the result list. indexing strategy. 6 FINDABILITY And metadata Documents ( unstruct content ) are tagged as a whole Metadata play a critical role Metadata how-to in dynamic delivery. It is the The mass of product information mechanism that filters and Author : MIE is growing explosively because of targets content matching a Product : AFS short renewal cycles, customiza- situation. Whether pre-existing or tions, and a multilingual user base. added through AI, it is essential. Version : 7.5 The idea of applying metadata to Metadata is information about Author : MIE every single topic in this gigantic your data, invisible tags describing Product : AFS corpus seems overwhelming. your visible text. Any document can Version : Author MIE Fortunately, metadata can be have metadata, and almost all do: 7.5 Component based documents can be tagged inherited, which makes the task creation date, author, number of manageable. Descriptive tags can pages… An unstructured document be inserted at the document and (such as Word) will have metadata chapter levels; these are inherited at the document level. A structured Author MIE as a whole by the topics below. This way, you document made of components get the advantages of topic-level (such as DITA) can have document- Version tagging without the intimidating level metadata, but also at the topic at a section level prospect of applying metadata to level or below. It simply has many every topic individually. more places for putting tags, so it FT can be described in finer detail. Version Topics can also be tagged auto- at a topic level matically, using machine-learning This ability grants structured docu- individually based algorithms. The AI can mentation superpowers: it permits generate a model from existing tag- content reuse, lowers the cost of FT ged content and apply it to other localization, and supports alternative content, even from different silos. options of information architecture… including dynamic delivery. and tags can be automatically forwarded 7 FINDABILITY And content sources Transparent silos A World of search Your search engine should have Dynamic delivery has the power Not all your users speak English. Structured X-ray vision, to see through silos. In to bypass their rigid constraints documentation provides measurable fact, despite all the bad press silos by reading and combining content advantages in translating and localizing get, source diversity is natural from any source. It generates documentation, enabling your enterprise to and unavoidable. searchable metadata that indicate reach a global audience. the origin of information, without Silos have their uses, stemming But it is not enough to translate your limiting access to it. It should also from the different tools supporting information. What if your delivery platform align metadata, so that this will the processes and operations in is not good enough to cope with these not only yield benefits for the your company. Documentation languages? Your search capabilities in French, end-user’s search, it will benefit can and should reflect the the entire enterprise that can now German, Japanese, or Chinese should be as histories and traditions of different exploit synergies and promote good as in English.

Load more