Category: Big Content

Seamless: Structural Metadata for Multimodal Content

Post author By Michael Andrews
Post date October 25, 2017

Chatbots and voice interaction are hot topics right now. New services such as Facebook Messenger and Amazon Alexa have become popular quickly. Publishers are exploring how to make their content multimodal, so that users can access content in varied ways on different devices. User interactions may be either screen-based or audio-based, and will sometimes be hands-free.

Multimodal content could change how content is planned and delivered. Numerous discussions have looked at one aspect of conversational interaction: planning and writing sentence-level scripts. Content structure is another dimension relevant to voice interaction, chatbots and other forms of multimodal content. Structural metadata can support the reuse of existing web content to support multimodal interaction. Structural metadata can help publishers escape the tyranny of having to write special content for each distinct platform.

Seamless Integration: The Challenge for Multimodal Content

In-Vehicle Infotainment (IVI) systems such as Apple’s CarPlay illustrate some of challenges of multimodal content experiences. Apple’s Human Interface Guidelines state: “On-screen information is minimal, relevant, and requires little decision making. Voice interaction using Siri enables drivers to control many apps without taking their hands off the steering wheel or eyes off the road.” People will interact with content hands-free, and without looking. CarPlay includes six distinct inputs and outputs:

Audio
Car Data
iPhone
Knobs and Controls
Touchscreen
Voice (Siri)

The CarPlay UIKit even includes “Drag and Drop Customization”. When I review these details, much seems as if it could be distracting to drivers. Apple states with CarPlay “iPhone apps that appear on the car’s built-in display are optimized for the driving environment.” What that iPhone app optimization means in practice could determine whether the driver gets in an accident.

CarPlay screenshot — CarPlay: if it looks like an iPhone, does it act like an iPhone? (screenshot via Apple)

Multimodal content promises seamless integration between different modes of interaction, for example, reading and listening. But multimodal projects carry a risk as well if they try to port smartphone or web paradigms into contexts that don’t support them. Publishers want to reuse content they’ve already created. But they can’t expect their current content to suffice as it is.

In a previous post, I noted that structural metadata indicates how content fits together. Structural metadata is a foundation of a seamless content experience. That is especially true when working with multimodal scenarios. Structural metadata will need to support a growing range of content interactions, involving distinct modes. A mode is form of engaging with content, both in terms of requesting and receiving information. A quick survey of these modes suggests many aspects of content will require structural metadata.

Platform Example	Input Mode	Output Mode
Chatbots	Typing	Text
Devices with Mic & Display	Speaking	Visual (Video, Text, Images, Tables) or Audio
Smart Speakers	Speaking	Audio
Camera/IoT	Showing or Pointing	Visual or Audio

Multimodal content will force content creators to think more about content structure. Multimodal content encompasses all forms of media, from audio to short text messages to animated graphics. All these forms present content in short bursts. When focused on other tasks, users aren’t able to read much, or listen very long. Steven Pinker, the eminent cognitive psychologist, notes that humans can only retain three or four items in short term memory (contrary to the popular belief that people can hold 7 items). When exploring options by voice interaction, for example, users can’t scan headings or links to locate what they want. Instead of the user navigating to the content, the content needs to navigate to the user.

Structural metadata provides information to machines to choose appropriate content components. Structural metadata will generally be invisible to users — especially when working with screen-free content. Behind the scenes, the metadata indicates hidden structures that are important to retrieving content in various scenarios.

Metadata is meant to be experienced, not seen. A photo of an Amazon customer’s Echo Show, revealing code (via Amazon)

Optimizing Content With Structural Metadata

When interacting with multimodal content, users have limited attention, and a limited capacity to make choices. This places a premium on optimizing content so that the right content is delivered, and so that users don’t need to restate or reframe their requests.

Existing web content is generally not optimized for multimodal interaction — unless the user is happy listening to a long article being read aloud, or seeing a headline cropped in mid-sentence. Most published web content today has limited structure. Even if the content was structured during planning and creation, once delivered, the content lacks structural metadata that allows it to adapt to different circumstances. That makes it less useful for multimodal scenarios.

In the GUI paradigm of the web, users are expected to continually make choices by clicking or tapping. They see endless opportunities to “vote” with their fingers, and this data is enthusiastically collected and analyzed for insights. Publishers create lots of content, waiting to see what gets noticed. Publishers don’t expect users to view all their content, but they expect users to glance at their content, and scroll through it until users have spotted something enticing enough to view.

Multimodal content shifts the emphasis away from planning delivery of complete articles, and toward delivering content components on-demand, which are described by structural metadata. Although screens remain one facet of multimodal content, some content will be screen-free. And even content presented on screens may not involve a GUI: it might be plain text, such as with a chatbot. Multimodal content is post-GUI content. There are no buttons, no links, no scrolling. In many cases, it is “zero tap” content — the hands will be otherwise occupied driving, cooking, or minding children. Few users want to smudge a screen with cookie dough on their hands. Designers will need to unlearn their reflexive habit of adding buttons to every screen.

Users will express what they want, by speaking, gesturing, and if convenient, tapping. To support zero-tap scenarios successfully, content will need to get smarter, suggesting the right content, in the right amount. Publishers can no longer present an endless salad bar of options, and expect users to choose what they want. The content needs to anticipate user needs, and reduce demands on the user to make choices.

Users will aways want to choose what topics they are interested in. They may be less keen on actively choosing the kind of content to use. Visiting a website today, you find articles, audio interviews, videos, and other content types to choose from. Unlike the scroll-and-scan paradigm of the GUI web, multimodal content interaction involves an iterative dialog. If the dialog lasts too long, it gets tedious. Users expect the publisher to choose the most useful content about a topic that supports their context.

Pattern: after saying what you want information about, now tell us how you’d like it (screenshot via Google News)

In the current use pattern, the user finds content about a topic of interest (topic criteria), then filters that content according to format preferences. In future, publishers will be more proactive deciding what format to deliver, based on user circumstances.

Structural metadata can help optimize content, so that users don’t have to choose how they get information. Suppose the publisher wants to show something to the user. They have a range of images available. Would a photo be best, or a line drawing? Without structural metadata, both are just images portraying something. But if structural metadata indicates the type of image (photo or line diagram), then deeper insights can be derived. Images can be A/B tested to see which type is most effective.

A/B testing of content according to its structural properties can yield insights into user preferences. For example, a major issue will be learning how much to chunk content. Is it better to offer larger size chunks, or smaller ones? This issue involves the tradeoffs for the user between the costs of interaction, memory, and attention. By wrapping content within structural metadata, publishers can monitor how content performs when it is structured in alternative ways.

Component Sequencing and Structural Metadata

Multimodal content is not delivered all at once, as is the case with an article. Multimodal content relies on small chunks of information, which act as components. How to sequence these components is important.

photo of Echo Show — Alexa showing some cards on an Echo Show device (via Amazon)

Screen-based cards are a tangible manifestation of content components. A card could show the current weather, or a basketball score. Cards, ideally, are “low touch.” A user wants to see everything they need on a single card, so they don’t need to interact with buttons or icons on the card to retrieve the content they want. Cards are post-GUI, because they don’t rely heavily on forms, search, links and other GUI affordances. Many multimodal devices have small screens that can display a card-full of content. They aren’t like a smartphone, cradled in your hand, with a screen that is scrolled. An embedded screen’s purpose is primarily to display information rather than for interaction. All information is visible on the card [screen], so that users don’t need to swipe or tap. Because most of us are accustomed to using screen-based cards already, but may be less familiar with screen-free content, cards provide a good starting point for considering content interaction.

Cards let us consider components both as units (providing an amount of content) and as plans (representing a purpose for the content). User experiences are structured from smaller units of content, but these units need have a cohesive purpose. Content structure is more than breaking content into smaller pieces. It is about indicating how those pieces can fit together. In the case of multimodal content, components need to fit together as an interaction unfolds.

Each card represents a specific type of content (recipe, fact box, news headline, etc.), which is indicated with structural metadata. The cards also present information in a sequence of some sort.¹ Publishers need to know how various types of components can be mixed, and matched. Some component structures are intended to complement each other, while other structures work independently.

Content components can be sequenced in three ways. They can be:

Modular
Fixed
Adaptive

Truly modular components can be sequenced in any order; they have no intrinsic sequence. They provide information in response to a specific task. Each task is assumed to be unrelated. A card providing an answer to the question of “What is the height of Mount Everest?” will be unrelated to a card answering the question “What is the price of Facebook stock?”

The technical documentation community uses an approach known as topic-based writing that attempts to answer specific questions modularly, so that every item of content can be viewed independently, without need to consult other content. In principle, this is a desirable goal: questions get answered quickly, and users retrieve the exact information they need without wading through material they don’t need. But in practice, modularity is hard to achieve. Only trivial questions can be answered on a card. If publishers break a topic into several cards, they should indicate the relations between the information on each card. Users get lost when information is fragmented into many small chunks, and they are forced to find their way through those chunks.

Modular content structures work well for discrete topics, but are cumbersome for richer topics. Because each module is independent of others, users, after viewing the content, need to specify what they want next. The downside of modular multimodal content is that users must continually specify what they want in order to get it.

Components can sequenced in a fixed order. An ordered list is a familiar example of structural metadata indicating a fixed order. Narratives are made from sequential components, each representing an event that happens over time. The narrative could be a news story, or a set of instructions. When considered as a flow, a narrative involves two kinds of choices: whether to get details about an event in the narrative, or whether to get to the next event in the narrative. Compared with modular content, fixed sequence content requires less interaction from the user, but longer attention.

Adaptive sequencing manages components that are related, but can be approached in different orders. For example, content about an upcoming marathon might include registration instructions, sponsorship info, a map, and event timing details, each as a separate component/card. After viewing each card, users need options that make sense, based on content they’ve already consumed, and any contextual data that’s available. They don’t want too many options, and they don’t want to be asked too many questions. Machines need to figure out what the user is likely to need next, without being intrusive. Does the user need all the components now, or only some now?

Adaptive sequencing is used in learning applications; learners are presented with a progression of content matching their needs. It can utilize recommendation engines, suggesting related components based on choices favored by others in a similar situation. An important application of adaptive sequencing is deciding when to ask a detailed question. Is the question going to be valuable for providing needed information, or is the question gratuitous? A goal of adaptive sequencing is to reduce the number of questions that must be asked.

Structural metadata generally does not explicitly address temporal sequencing, because (until now) publishers have assumed all content would be delivered at once on a single web page. For fixed sequences, attributes are needed to indicate order and dependencies, to allow software agents to follow the correct procedure when displaying content. Fixed sequences can be expressed by properties indicating step order, rank order, or event timing. Adaptive sequencing is more programmatic. Publishers need to indicate the relation of components to parent content type. Until standards catch up, publishers may need to indicate some of these details in the data-* attribute.

The sequencing of cards illustrates how new patterns of content interaction may necessitate new forms of structural metadata.

Composition and the Structure of Images

One challenge in multimodal interaction is how users and systems talk about images, as either an input (via a camera), or as an output. We are accustomed to reacting to images by tapping or clicking. We now have the chance to show things to systems, waving an object in front of a camera. Amazon has even introduced a hands-free voice activated IoT camera that has no screen. And when systems show us things, we may need to talk about the image using words.

Machine learning is rapidly improving, allowing systems to recognize objects. That will help machines understand what an item is. But machines still need to understand the structural relationship of items that are in view. They need to understand ordinary concepts such as near, far, next to, close to, background, group of, and other relational terms. Structural metadata could make images more conversational.

Vector graphics are composed of components that can represent distinct ideas, much like articles that are composed of structural components. That means vector images can be unbundled and assembled differently. The WAI-ARIA standard for web accessibility has an SVG Graphics Module that covers how to markup vector images. It includes properties to add structural metadata to images, such as group (a role indicating similar items in the image) and background (a label for elements in the image in the background). Such structural metadata could be useful for users interacting with images using voice commands. For example, the user might want to say, “Show me the image without a background” or “with a different background”.

Photos do not have interchangeable components the way that vector graphics do. But photos can present a structural perspective of a subject, revealing part of a larger whole. Photos can benefit from structural metadata that indicates the type of photo. For example, if a user wants a photo of a specific person, they might have a preference for a full-length photo or for a headshot. As digital photography has become ubiquitous, many photos are available of the same subject that present different dimensions of the subject. All these dimensions form a collection, where the compositions of individual photos reveal different parts of the subject. The IPTC photo metadata schema includes a controlled vocabulary for “scenes” that covers common photo compositions: profile, rear view, group, panoramic view, aerial view, and so on. As photography embraces more kinds of perspectives, such as aerial drone shots and omnidirectional 360 degree photographs, the value of perspective and scene metadata will increase.

For voice interaction with photo images to become seamless, machines will need to connect conversational statements with image representations. Machines may hear a command such as “show me the damage to the back bumper,” and must know to show a photo of the rear view of a car that’s been in an accident. Sometimes users will get a visual answer to a question that’s not inherently visual. A user might ask: “Who will be playing in Saturday’s soccer game?”, and the display will show headshots of all the players at once. To provide that answer, the platform will need structural metadata indicating how to present an answer in images, and how to retrieve player’s images appropriately.

Structural metadata for images lags behind structural metadata for text. Working with images has been labor intensive, but structural metadata can help with the automated processing of image content. Like text, images are composed of different elements that have structural relationships. Structural metadata can help users interact with images more fluidly.

Reusing Text Content in Voice Interaction

Voice interaction can be delivered in various ways: through natural language generation, through dedicated scripting, and through the reuse of existing text content. Natural language generation and scripting are especially effective in short answer scenarios — for example, “What is today’s 30 year mortgage rate? ” Reusing text content is potentially more flexible, because it lets publishers address a wide scope of topics in depth.

While reusing written text in voice interactions can be efficient, it can potentially be clumsy as well. The written text was created to be delivered and consumed all at once. It needs some curation to select which bits work most effectively in a voice interaction.

The WAI-ARIA standards for web accessibility offer lessons on the difficulties and possibilities of reusing written content to support audio interaction. By becoming familiar with what ARIA standards offer, we can better understand how structural metadata can support voice interactions.

ARIA standards seek to reduce the burdens of written content for people who can’t scan or click through it easily. Much web content contains unnecessary interaction: lists of links, buttons, forms and other widgets demanding attention. ARIA encourages publishers to prioritize these interactive features with the TAB index. It offers a way to help users fill out forms they must submit to get to content they want. But given a choice, users don’t want to fill out forms by voice. Voice interaction is meant to dispense with these interactive elements. Voice interaction promises conversational dialog.

Talking to a GUI is awkward. Listening to written web content can also be taxing. The ARIA standards enhance the structure of written content, so that content is more usable when read aloud. ARIA guidelines can help inform how to indicate structural metadata to support voice interaction.

The ARIA encourages publishers to curate their content: to highlight the most important parts that can be read aloud, and to hide parts that aren’t needed. ARIA designates content with landmarks. Publishers can indicate what content has role=“main”, or they can designate parts of content by region. The ARIA standard states: “A region landmark is a perceivable section containing content that is relevant to a specific, author-specified purpose and sufficiently important that users will likely want to be able to navigate to the section easily and to have it listed in a summary of the page.” ARIA also provides a pattern for disclosure, so that not all text is presented at once. All of these features allow publishers to indicate more precisely the priority of different components within the overall content.

ARIA supports screen-free content, but it is designed primarily for keyboard/text-to-speech interaction. Its markup is not designed to support conversational interaction — schema.org’s pending speakable specification, mentioned in my previous post, may be a better fit. But some ARIA concepts suggest the kinds of structures that written text need to work effectively as speech. When content conveys a series of ideas, users need to know what are major and minor aspects of text they will be hearing. They need the spoken text to match the time that’s available to listen. Just like some word processors can provide an “auto summary” of a document by picking out the most important sentences, voice-enabled text will need to identify what to include in a short version of the content. The content might be structured in an inverted pyramid, so that only the heading and first paragraph are read in the short version. Users may even want the option of hearing a short version or a long version of a story or explanation.

Structural metadata and User Intent in Voice Interaction

Structural metadata will help conversational interactions deliver appropriate answers. On the input side, when users are speaking, the role of structural metadata is indirect. People will state questions or commands in natural language, which will be processed to identify synonyms, referents, and identifiable entities, in order to determine the topic of the statement. Machines will also look at the construction of the statement to determine the intent, or the kind of content sought about the topic. Once the intent is known — what kind of information the user is seeking — it can be matched with the most useful kind of content. It is on the output side, when users view or hear an answer, that structural metadata plays an active role selecting what content to deliver.

Already, search engines such as Google rely on structural metadata to deliver specific answers to speech queries. A user can ask Google the meaning of a word or phrase (What does ‘APR’ mean?) and Google locates a term that’s been tagged with structural metadata indicating a definition, such as with the HTML element <dfn>.

When a machine understands the intent of a question, it can present content that matches the intent. If a user asks a question starting with the phrase Show me… the machine can select a clip or photograph about the object, instead of presenting or reading text. Structural metadata about the characteristics of components makes that matching possible.

Voice interaction supplies answers to questions, but not all answers will be complete in a single response. Users may want to hear alternative answers, or get more detailed answers. Structural metadata can support multi-answer questions.

Schema.org metadata indicates content that answers questions using the Answer type, which is used by many forums and Q&A pages. Schema.org distinguishes between two kinds of answers. The first, acceptedAnswer, indicates the best or most popular answer, often the answer that received most votes. But other answers can be indicated with a property called suggestedAnswer. Alternative answers can be ranked according to popularity as well. When sources have multiple answers, users can get alternative perspectives on a question. After listening to the first “accepted” answer, the user might ask “tell me another opinion” and a popular “suggested” answer could be read to them.

Another kind of multi-part answer involves “How To” instructions. The HowTo type indicates “instructions that explain how to achieve a result by performing a sequence of steps.” The example the schema.org website provides to illustrate the use of this type involves instructions on how to change a tire on a car. Imagine car changing instructions being read aloud on a smartphone or by an in-vehicle infotainment system as the driver tries to change his flat tire along a desolate roadway. This is a multi-step process, so the content needs to be retrievable in discrete chunks.

Schema.org includes several additional types related to HowTo that structure the steps into chunks, including preconditions such as tools and supplies required. These are:

HowToSection : “A sub-grouping of steps in the instructions for how to achieve a result (e.g. steps for making a pie crust within a pie recipe).”
HowToDirection : “A direction indicating a single action to do in the instructions for how to achieve a result.”
HowToSupply : “A supply consumed when performing the instructions for how to achieve a result.”
HowToTool : “A tool used (but not consumed) when performing instructions for how to achieve a result.”

These structures can help the content match the intent of users as they work through a multi-step process. The different chunks are structurally connected through the step property. Only the HowTo type ( and its more specialized subtype, the Recipe) currently accepts the step property and thus can address temporal sequencing.

Content Agility Through Structural Metadata

Chatbots, voice interaction and other forms of multimodal content promise a different experience than is offered by screen-centric GUI content. While it is important to appreciate these differences, publishers should also consider the continuities between traditional and emerging paradigms of content interaction. They should be cautious before rushing to create new content. They should start with the content they have, and see how it can be adapted before making content they don’t have.

A decade ago, the emergence of smartphones and tablets triggered an app development land rush. Publishers obsessed over the discontinuity these new devices presented, rather than recognizing their continuity with existing web browser experiences. Publishers created multiple versions of content for different platforms. Responsive web design emerged to remedy the siloing of development. The app bust shows that parallel, duplicative, incompatible development is unsustainable.

Existing content is rarely fully ready for an unpredictable future. The idealistic vision of single source, format free content collides with the reality of new requirements that are fitfully evolving. Publishers need an option between the extremes of creating many versions of content for different platforms, and hoping one version can serve all platforms. Structural metadata provides that bridge.

Publishers can use structural metadata to leverage content they have already that could be used to support additional forms of interaction. They can’t assume they will directly orchestrate the interaction with the content. Other platforms such as Google, Facebook or Amazon may deliver the content to users through their services or devices. Such platforms will expect content that is structured using standards, not custom code.

Sometimes publishers will need to enhance existing content to address the unique requirements of voice interaction, or differences in how third party platforms expect content. The prospect of enhancing existing content is preferable to creating new content to address isolated use case scenarios. Structural metadata by itself won’t make content ready for every platform or form of interaction. But it can accelerate its readiness for such situations.

— Michael Andrews

Dialogs in chatbots and voice interfaces also involve sequences of information. But how to sequence a series of cards may be easier to think about than a series of sentences, since viewing cards doesn’t necessarily involve a series of back and forth questions. ↩︎

Tags metadata, structural, voice interfaces

Big Content

Content Velocity, Scope, and Strategy

Post author By Michael Andrews
Post date September 12, 2017

I want to discuss three related concepts.

First, the velocity of the content, or how quickly the essence of content changes.

Second, the scope that the content addresses, that is, whether it is meant for one person or many people, and whether it is intended to have a short or a long life span.

Lastly, how those factors affect publishing strategy, and the various capabilities publishers need.

These concepts — velocity, scope and strategy — can help publishers diagnose common problems in content operations. Many organizations produce too much content, and find they have repetitive or dated content. Others struggle to implement approaches that were developed and successfully used in one context, such as technical documentation, and apply them to another, such as marketing communication. Some publishers don’t have clear criteria addressing the utility of content, such as when new content is needed, or how long published content should be kept online. Instead of relying on hunches to deal with these issues, publishers should structure their operations to reflect the ultimate purpose of their content.

Content should have a core rationale for why it exists. Does the organization publish content to change minds — to get people to perceive topics differently, learn about new ideas, or to take an action they might not otherwise take without the content? Or does it publish content to mind the change — to keep audiences informed with the most current information, including updates and corrections, on topics they already know they need to consult?

When making content, publishers should be able to answer: In what way is the content new? Is it novel to the audience, or just an update on something they already know about? The concept of content velocity can help us understand how quickly the information associated with content changes, and the extent to which newly created content provides new information.

Content Velocity: Assessing Newness and Novelty

All content is created based on the implicit assumption that it says something new, or better, than existing content that’s available. Unfortunately, much new content gets created without ever questioning whether it is completely necessary. True, people need information about a topic to support a task or goal they have. But is new content really necessary? Or could existing content be revised to address these needs?

Let’s walk through the process by which new content gets created.

Is new content necessary, or should existing content be revised?

The first decision is whether the topic or idea warrants the creation of new content, or whether existing content covers much of the same material. If the topic or idea is genuinely new and has not been published previously, then new content is needed. If the publisher has only a minor update to material they’ve previously published, they should update the existing content, and not create new content. They may optionally issue an alert indicating that a change has been made, but such a notification won’t be part of the permanent record. Too often, publishers decide to write new articles about minor changes that get added to the permanent stock of content. Since the changes were minor, most such articles repeat information already published elsewhere, resulting in duplication and confusion for all concerned.

The next issue is to decide if the new content is likely to be viewed by an individual more than once. This is the shelf life of the content, considered from the audience’s perspective. Some content is disposable: its value is negligible after being viewed, or if never viewed by a certain date. Content strategists seldom discuss short-lived, disposable content, except to criticize it as intrinsically wasteful. Yet some content, owing to its nature, is short lived. Like worn razor blades or leftover milk, it won’t be valuable forever. It needs to disappear from the individual’s field of vision when it is no longer useful. If the audience considers the content disposable, then the publisher needs to treat it that way as well, and have a process for getting the content off the shelf. Other content is permanent: it always needs to be available, because people may need to consult it more than once.

Publishers must also decide whether the content is either custom (intended for a specific individual), or generic (intended for many people). We will return to custom and generic content shortly.

If the publisher already has content covering the topic, it needs to ask whether new information has emerged that requires existing content to be updated. We’d also like to know if some people may have seen this existing content previously, and will be interested in knowing what’s changed. For example, I routinely consult W3C standards drafts. I may want to know what’s different between one revision compared with the prior one, and appreciate when that information is called out. For content I don’t routinely consult, I am happy to simply know that all details are current as of a certain date when the content was last revised.

One final case exists, which is far too common. The publisher has already covered the topic or idea, and has no new information to offer. Instead, they simply repackage existing content, giving it a veneer of looking new. While repackaged content is sometimes okay if it involves a genuinely different approach to presenting the information, it is generally not advisable. Repackaged content results from the misuse of the concept of content reuse. Many marketing departments have embraced content reuse as a way to produce ever more content, saying the same thing, in the hopes that some of this content will be viewed. The misuse of content reuse, particularly the automated creation of permanent content, is fueling an ever growing content bubble. Strategic content reuse, in contrast, involves the coordination of different content elements into unique bundles of information, especially customized packages of information that support targeted needs or interests.

Once publishers decide to create new content, they need to decide content scope, the content’s expected audience and expected use.

Content Scope: Assessing Uniqueness and Specificity

Content scope refers to how unique or specific newly created content is. We can consider uniqueness in terms of audiences (whether the content is for a specific individual, or a collective group), and in terms of time (is the content meant to be used at a specific moment only, or will be viewed again). Content that is intended for a specific individual, or for viewing at a specific time, is more unique, and has narrower range of uses, than content that’s been created for many people, or for viewing multiple times by the same person. How and when the audience uses the content will influence how the publisher will need to create and manage that content.

Scope can vary according to four dimensions:

The expected frequency of use
The expected audience size
The archival and management approach (which will mirror the expected frequency of use)
The content production approach (which will mirror the expected audience size)

The expected frequency of use looks at whether someone is likely to want to view content again after seeing it once. This looks at relevance from an individual’s perspective, rather than a publisher’s perspective. Publishers may like to think they are creating so-called evergreen content that people will always find relevant, but from an audience perspective, most content, once viewed, will never be looked at again. When audiences encounter content they’ve previously viewed, they are likely to consider it as clutter, unless they’ve a specific reason to view it again. Audiences are most likely to consider longer, more substantive content on topics of enduring interest as permanent content. They are likely to consider most other content as disposable.

Disposable content is often event driven. Audiences need content that addresses a specific situation, and what is most relevant to them is content that addresses their needs at that specific moment. Typically this content is either time sensitive, or customized to a specific scenario. Most news has little value unless seen shortly after it is created. Customized information can deliver only the most essential details that are relevant to that moment. Customers may not want to know everything about their car lease — they only want to know about the payment for this month. Once this month’s payment question has been answered, they no longer need that information. This scenario shows how disposable content can be a subset of permanent content. Audiences may not want to view all the permanent content, and only want to view a subset of it. Alerts are one way to deliver disposable content that highlights information that is relevant, but only for a short time.

The expected audience refers to whether the content is intended for an individual, or addresses the interests of a group of individuals. Historically, nearly all online content addressed a group of people, frequently everyone. More recently, digital content has become more customized to address individual situational needs and interests, where the content one person views will not be the same as the content another views, even if the content covers the same broad topic. The content delivered can consider factors such as geolocation, viewing history, purchase history, and device settings to provide content that is more relevant to a specific individual. By extension, the more that content is adjusted to be relevant to a specific individual, the less that same content will be relevant to other individuals.

A tradeoff exists, between how widely viewed content is, and how helpful it might be to a specific individual. Generic reference content may generate many views, and be helpful to many people, but it might not provide exactly what any one of those people want. Single use content created for an individual may provide exactly what that person needed, at the specific time they viewed the content. But that content will be helpful to an single person only, unless such customization is scalable across time and different individuals.

Disposable content is moment-rich, but duration-poor. Marketing emails highlight the essential features of disposable content. People never save marketing emails, and they rarely forward them to family and friends. They rarely even open and read them, unless they are checking their email at a moment of boredom and want a distraction — fantasizing about some purchase they may not need, or wanting to feel virtuous for reading a tip they may never actually use. Disposable content sometimes generates zero views by an individual, and almost never will generate more than one view. If there’s ever a doubt about whether someone might really need the information later, publishers can add a “save for later” feature — but only when there’s a strong reason to believe a identifiable minority has a critical need to access the content again.

Publishers face two hurdles with disposable content: being able to quickly produce new content, and being able to deliver time-sensitive or urgent content to the right person when it is needed. They don’t need to worry about archiving the content, since it is no longer valuable. Disposable content is always changing, so that different people on different days will receive different content.

With permanent content, publishers need to worry about managing existing content, and having a process for updating it. Publishers become concerned with consistency, tracking changes, and versioning. These tasks are less frenetic than those for disposable content, but they can be more difficult to execute well. It is easy to keep adding layers of new material on top of old material, while failing to indicate what’s now important, and for whom.

Content that’s used repeatedly, but is customized to specific individual needs, can present tricky information architecture challenges. These can be addressed by having a user login to a personal account, where their specific content is stored and accessible.

Strategies for Fast and Slow Content: Operations Fit to Purpose

All publishers operate somewhere along a spectrum. One end emphasizes quick turn-around, short-lived content (such as news organizations), and the other end emphasizes slowly evolving, long-lived content (such as healthcare advice for people with chronic conditions.) Many organizations will publish a mix of fast and slow content. But it’s important for organizations to understand whether they are primarily a fast or slow content publisher, so that they can decide on the best strategy to support their publishing goals.

Most organizations will be biased toward either fast or slow content. Fast moving consumer goods, unsurprisingly, tend to create fast content. In contrast, heavy equipment manufacturers, whose products may last for decades, tend to generate slow content that’s revised and used over a long period.

Different roles in organizations gravitate toward either fast or slow content. Consider a software company. Marketers will blitz customers with new articles talking about how revolutionary the latest release of their software is. Customer support may be focused on revising existing content about the product, and reassuring customers that the changes aren’t frightening, but easy to learn and not disruptive. Even if the new release generates a new instance of technical documentation devoted to that release, the documentation will reuse much of the content from previous releases, and will essentially be a revision to existing content, rather than fundamentally new content.

Fast content is different from slow content

Some marketers want their copywriters to become more like journalists, and have set up “newsrooms” to churn out new content. When emulating journalists, marketers are sticking with the familiar fast content paradigm, where content is meant to be read once only, preferably soon after it’s been created. Most news gets old quickly, unless it is long form journalism that addresses long-term developments. Marketing content frequently has a lifespan of a mosquito.

Marketing content tends to focus on:

Creating new content, or
Repackaging existing content, and
Making stuff sound new (and therefore attention worthy)

For fast content, production agility is essential.

Non-marketing content has a different profile. Non-marketing content includes advisory information from government or health organizations, and product content, such as technical documentation, product training, online support content, and other forms of UX content such as on-screen instructions. Such content is created for the long term, and tends to emphasize that it is solid, reliable and up-to-date. Rather than creating lots of new content, existing content tends to evolve. It gets updated, and expands as products gain features or knowledge grows. It may lead with what’s new, but will build on what’s been created already.

Much non-marketing content is permanent content about a fixed set of topics. The key task is not brainstorming new topics to talk about, but keeping published information up-to-date. New permanent topics are rare. When new topics are necessary, it’s common for new topics to emerge as branches of an existing topic.

Fast and slow content are fundamentally different in orientation. Organizations are experimenting with ways to bridge these differences. Organizations may try to make their marketing content more like product content, or conversely, make their product content more like marketing content.

Some marketing organizations are adopting technical communications methods, for example, content management practices developed for technical documentation such as DITA. Marketing communications are seeking to leverage lessons from slow content practices, and apply them to fast content, so that they can produce more content at a larger scale.

Marketers want their content to become more targeted. They want to componentize content so they can reuse content elements in endless combinations. They embrace reuse, not as a path to revise existing content, but as a mechanism to push out new content quickly, using automation. At its best, such automation can address the interests of audiences more precisely. At its worst, content automation becomes a fatigue-inducing, attention-fragmenting experience for audiences, who are constantly goaded to view messages without ever developing an understanding . Content reuse is a poor strategy for getting attention from audiences. New content, when generated from the reuse of existing content components, never really expresses new ideas. It just recombines existing information.

Some technical communicators, who develop slow content, are implementing practices associated with marketing communications. Rather than only producing permanent documents to read, technical communication teams are seeking to push specific disposable messages to resolve issues. Technical communication teams are embracing more push tactics, such as developing editorial calendars, to highlight topics to send to audiences, instead of waiting for audiences to come to them. These teams are seeking to become more agile, and targeted, in the content they produce.

As the boundaries between the practices of fast and slow content begin to overlap, delivery becomes more important. Publishers need to choose between targeted verses non-targeted delivery. They must decide of their content will be customized and dynamically created according to user variables, or pre-made to anticipate user needs.

The value of fast content depends above all on the accuracy of its targeting. There is no point creating disposable content if it doesn’t resolve a problem for users. If publishers rely on fast content, but can’t deliver it to the right users at the right time, the user may never find out the answer to their question, especially if permanent content gets neglected in the push for instant content delivery.

Generic fast content is becoming ever more difficult to manage. Individuals don’t want to see content they’ve viewed already, or decided they weren’t interested in viewing to begin with. But because generic content is meant for everyone, it is difficult to know who has seen or not seen content items. Fast generic content still has a role. Targeting has its limits. Publishers are far from being able to produce personalized content for everyone that is useful and efficient. Much content will inevitably have no repeat use. Yet fast generic content can easily become a liability that is difficult to manage. Recommendation engines based on user viewing behaviors and known preferences can help prioritize this content so that more relevant content surfaces. But publishers should be judicious when creating fast generic content, and should enforce strict rules on how long such content stays available online.

Automation is making new content easier to create, which is increasing the temptation to create more new content. Unfortunately, digital content can resemble plastic shopping bags, which are useful when first needed, but which generally never get used again, becoming waste. Publishers need to consider content reuse not just from their own parochial perspective, but from the perspective of their audiences. Do their audiences want to view their content more than once? Marketing content is the source of most fast content. Most marketing content is never read more than once. Can that ever change? Are marketers capable of producing content that has long term value to their audiences? Or will they insist on controlling the conversation, directing their customers on what content to view, and when to view it?

Creating new content is not always the right approach. Automation can make it more convenient for publishers to pursue the wrong strategy, without scrutinizing the value of such content to the organization, and its customers. Content production agility is valuable, but having robust content management is an even more strategic capability.

— Michael Andrews

Tags scope, velocity

Big Content

Content Maintenance: A Framework

Post author By Michael Andrews
Post date June 15, 2016

What happens to content after its publication seems to vary widely. Content maintenance is not considered exciting — and is often overlooked. Even the term content maintenance has no commonly accepted definition. Despite its lowly status and fuzzy profile, content maintenance is a many sided and fundamental activity. I want to explore what content maintenance can involve, and how to prioritize its different aspects.

An Inconspicuous Activity

Many diagrams showing the content lifecycle have a placeholder for content maintenance. After creating content and delivering content, content maintenance is required. But what that means in practice is often not well-defined. This shouldn’t be surprising: content maintenance somehow lacks the urgency that content creation and content delivery have. After the adrenaline rush of creating new content, and watching the initial audience response to it, the content is no longer top of mind.

Content maintenance is often a lower priority. The UK's Government Digital Service awaits guidance on the topic. — Content maintenance is often a lower priority activity. The UK’s Government Digital Service awaits guidance on the topic. [screenshot]

In many organizations, content maintenance isn’t planned at all. Some content gets updated because it is rewritten periodically, while other content is never touched after publication. The organization may do a cleanup every few years in conjunction with a website redesign or IT upgrade, relying on a tedious content inventory and audit to evaluate how messed up the situation has become — content strategy’s equivalent to doing a root canal.

Kristina Halvorson and Melissa Rich, in their popular book Content Strategy for the Web, offer one of the most detailed discussions of content maintenance. They call for a content maintenance plan that reflects content objectives such as assuring accuracy and consistency, archiving or reordering older content, confirming links are working and metadata is current, and removing redundant content. This is a good list, but in itself doesn’t suggest how to implement a repeatable process for maintenance. These authors further recommend establishing rules to govern content. Another sound idea.

But critical questions remain: On what basis does the organization prioritize its maintenance plan? What criteria govern maintenance decisions?

A pivotal issue with content maintenance is defining its scope. What activities are necessary, desirable but optional, or unnecessary? Does a maintenance task apply to all content, or are there any distinctions in types of content that effect how tasks are allocated? If there are differences in how various content is treated in maintenance, on what basis are priorities made? Are they arbitrary decisions made by a committee crafting a plan, or are decisions guided by a stable set of rules that are always dependable? How do we know we are doing content maintenance effectively, given finite resources?

Decisions about content maintenance can reflect deep convictions about the core value of different kinds of content. The content maintenance approach of an organization can express unconscious attitudes about content. One common approach is to make sure that popular content is kept up-to-date. An exclusive focus on popular content is not a comprehensive approach to content maintenance.

Many people consider content maintenance as a simple housekeeping task. But it can play a bigger role. The ultimate purpose of content maintenance is to help organizations use content to grow their engagement with audiences in a sustainable way. Content maintenance deserves to be reappraised as a foundation of sustainable growth, instead of a zero-sum exercise of pruning and fixing last year’s publications. When organizations understand how content maintenance is essential to making content reach and connect with audiences more effectively, they will place more emphasis on what to do after content’s been published.

Content maintenance serves two functions:

Countering entropy, where published content starts to decay due to various factors
Improving relevance through optimization

Entropy-fighting Maintenance

Keeping content up-to-date often seems like an impossible task because content managers don’t have a good understanding of why content gets out-of-date. Admittedly, the reasons are manifold. But a clearer understanding of why content becomes dated will help with the maintenance process.

Recently, some content managers have begun to speak about a concept called content debt. The concept refers to when there is no plan for the content once it has been published, so that a debt is incurred because the content’s creators have deferred decisions about what should happen to the content in the future. Many organizations practice bad habits when publishing content. With so many ways things can be done badly, establishing a robust process to keep content accurate is not an easy task.

We need to distinguish two related concepts: content relevance, and content validity. Content relevance refers to whether audiences care about the content. Content validity refers to whether the content is accurate, a subset of content relevance. When people speak of updating content, they often do not make clear whether they are concerned with fixing inaccuracies in the content, or whether they are assessing how to make it more relevant. A lack of precision when speaking of updating content can become confusing for those responsible for the task.

Generally, when people speak of updating content, they are presuming the content is intrinsically relevant, but that some aspect of it needs correction. Two types of updating are common:

Content accuracy updating
Technical updating

Content accuracy updating addresses what has changed factually in the content since it was published? Facts have a tendency to go out of date. In his book, The Half Life of Facts: Why Everything has an Expiration Date, Samuel Arbesman notes that even scientific facts are not stable, and are subject to revision.

Web content often created around events, past and future. Some content refers to forthcoming events on specific dates, which may indicate a need to update the content after that date, depending on what the content says. Other content may have a hidden dependency on events. A change in executive leadership, product names, or business location may impact many items of content that make reference to these now-changed facts. To some extent, content referring to internally determined facts can be managed utilizing content structure and metadata.

When content refers to facts that change due to external factors, the maintenance process can be messier. Content may refer to whether products comply with certain regulations that may be subject to change but that could not have been fully anticipated when the content was created. Service providers such as airlines may need to update content to indicate changes in factual information reflecting service disruptions due to external factors such as strikes or volcanic eruptions. Organizations should map how content accuracy may be dependent on external events, and plan for how content will be updated in such scenarios.

Technical updating can be challenging for organizations that have a lot of legacy content. The web is over 20 years old, and has become our external memory of what’s happened in our world. Many people expect to find something they saw in the past, but how that content was constructed from a technical perspective may not be compatible with current web standards.

External developments are often responsible for technical decay. Broken links referring to third party content are the most obvious example. The more pervasive problem arises when technical standards shift, and the content is not renewed to conform to the new standards. Two examples of standards shifts are changes in markup standards, such as the growth of semantic search markup, and changes in media format standards, such as the recent demise of the Flash format for video.

Updating content requires content professionals to think about the content lifecycle more globally. The content lifecycle depends on the event lifecycles to which the content refers and relies on. The content is part of a larger story: the story of how a product is managed, how an organization evolves, or how an enabling technology is used. The specifics and timing of these events may not be known, but the broad patterns of their occurrences can be anticipated and planned for.

Popularity as a Factor When Updating Content

Deciding what content to update is an essential part of content maintenance. It often involves judgment calls, which take time and do not always result in consistency. In the search for rules that govern what content to update, some content professionals have been advocating content popularity as the criterion for deciding what content to keep, and hence keep current.

Paul Boag has suggested a traffic-based approach to managing legacy content. Gerry McGovern has written extensively about the value of using traffic to guide content decisions. McGovern’s writings on the importance of content popularity have become popular themselves, so I want to examine the reasoning behind this approach. He uses the “long tail” metaphor of content demand first popularized by former Wired columnist Chris Anderson, but draws different conclusions. McGovern differentiates the “long tail (low demand or no demand stuff) and the long neck (high demand stuff). The long tail has been seen as a major opportunity, but it can become a major threat.”

The Long Tail concept of content popularity, where many items are of interest to few people.

McGovern considers the popularity of content in terms of the distribution of page views. He contrasts the head of the distribution, where the top ranked pages get a high volume of views, with the many low ranking pages in the tail that get few views. McGovern says: “Much of the long tail is a dead zone. It’s a dead and useless tail full of dead and useless content.” In contrast, content in the head of the distribution (the top-ranked pages) get the lion’s share of views. How much of total demand is attributable to the head is not always clear. McGovern has cited the figure that the top 5% of content generates 20% of demand; the next 35% of content (what he calls the body, though this is not a standard term used in head/tail distributions) accounts for 55% of demand; the remaining 60% is basically useless, accounting for only 20% of demand. More recently, McGovern has featured the success stories of various organizations when they removed 90% of their content, implying that only the top 10% of content viewed is worth keeping and maintaining.

McGovern’s argument is that low traffic content gets in the way of people accomplishing their “top tasks,” which are represented by the highest volume content. Moreover, long tail content is difficult to maintain.

McGovern makes some excellent points about the costs of low usage content, and the importance of making sure that frequently accessed content is easily available. Unfortunately, his formulation is not a reliable guide to deciding what content to maintain because it doesn’t distinguish different content purposes. McGovern considers all content as task-oriented content. But content can have other roles, such as being news, or providing background information for educational and entertainment purposes.

Let’s assume, like McGovern, that all content follows a head/tail distribution. With task-based content, the rank ordering of popularity is stable over time. If all the content is task-focused, then the ranking of pages in terms of their popularity won’t change. We can argue how many pages deserve to be included —that is, how many tasks should be supported —but the customer’s prioritization of what’s important is not impacted by when the content was created.

In other cases, when the content was created can have an impact on how important consumers consider the content, even when the age of the content has no bearing on the accuracy of the content. In these cases, relevance is a function of time.

Comparison of how content popularity might change over time, according to the content's purpose. — Comparison of how content popularity might change over time, according to the content’s purpose.

The most obvious case is when popularity of content decays over time because it becomes less topical. Any informative content that might be considered news, such as high profile product announcements or announcements of changes to membership services, might become less popular over time. News oriented content becomes less relevant over time, but old news content is not intrinsically irrelevant. The need to keep legacy content will depend on the organization’s mission, and the utility of the information in the future.

A different example of content is so-called evergreen content — content that will have a long shelf-life (even if it isn’t truly maintenance-free). Often such content is created to build awareness and interest in a topic, and isn’t tied to a specific event. The content may debut with a low ranking, but gain views as awareness builds. It may gain exposure through a promotional tie-in, a third party’s endorsement, or relevance to some event that wasn’t foreseen when the content was created. After gaining in popularity, its popularity may settle back down into the long tail. The popularity of such content can yo-yo over time.

Without considering the purpose of the content, it is difficult to know what content to maintain. Relying solely on content age or content popularity can result in a meat cleaver approach to content maintenance. While content that is jettisoned does not need to be maintained, the content that is kept does not all need to be treated the same way. Sometimes content is intentionally kept in an archive, with a minimum of maintenance promised or provided.

Content Maintenance as Quality Improvement

A very different view of content maintenance considers maintenance as continuous improvement. Instead of just keeping content up to date, content is actively managed to become better.

Gerry McGovern touches on this approach when discussing what he calls “top tasks”. He says: ”Continuously improve your top tasks. The Long Neck [top-ranked pages in the head of the distribution] is made up of a small set of top tasks and it’s important to manage them through a process of continuous improvement.“

Continuous improvement can be applied to any content, not just content supporting common tasks. The purpose of such activity is to make the content more relevant to audiences. It is a form of content optimization, a term most often used to refer to far narrower dimensions of content, such as search engine positioning or call-to-action behavior.

Optimizing content — in the broader sense of making it more relevant for audiences — involves two sets of decisions:

Deciding what content to optimize as part of an ongoing evaluation program
Making choices about evaluation methods, and assigning resources to support such activities

Optimizing content is often associated with A/B testing. But the range of approaches available to improve content relevance are diverse, and include:

Analyzing content usage and actions
Surveying attitudes and preferences relating to content
Testing content comprehension and receptivity

Behavioral outcomes can certainly be tested, but content testing and analysis can probe more upstream factors such as how audiences evaluate the content prior to considering taking action, and suggest more fundamental changes that could improve content relevance to audiences.

Some techniques available to improve content relevance can be done prior to publication. The question arises of what content should continue to be evaluated after publication with the aim of improving its performance? If the content was well thought out before its publication, why does it need optimization after publication? Any evaluation and enhancement activities that are done multiple times will require additional commitments of resources. What content justifies the extra attention?

Business critical content will be the best candidates to optimize through continuous improvement. Business criticality will reflect the reach of the content (amount of views), its importance to audiences, and its influence on business outcomes. It may be content that comprises part of a funnel, but not necessarily. In some cases, it might be key content that provides the first impression of an organization, product or topic that will have long term impacts in terms of future behavior or social influence.

Strategic Content Maintenance

Content maintenance involves many aspects, and content with different characteristics and purposes must be prioritized differently. Maintenance-free content is a myth, but not all content requires the same level of maintenance.

Many times organizations jettison content for the wrong reasons. Content should be retired because it is no longer relevant, not simply because something about it, factually or technically, has become dated through neglect, and it is now too much effort to update it given the volume of content that’s in a similar state. Manage content based on current and potential relevance, and plan to maintain the relevant content. Jettison content that’s no longer relevant, and unlikely to be relevant again.

While having less content can be easier to maintain than having large volumes of content, one shouldn’t base one’s content strategy on operational convenience. Small, tightly focused websites can be the right choice for smaller, tightly focused organizations, who can devote their energies toward optimizing the entirety of their content. But for larger organizations with diverse missions, imposing a rigid content diet — say restricting the website to 50 pages — may help the web operations run smoothly, but at the cost of preventing the organization from fully serving their audiences, and allowing the organization to meet their wider digital objectives. If the breadth of content offered is insufficient to satisfy the needs of audiences, the content published will sustain neither the audience’s interests, nor the mission of the organization. The organization needs to understand what effort is required to maintain what it decides is relevant content, and then plan how to maintain that content appropriately.

No one-size strategy for content maintenance is right for all organizations. But organizations can use a framework to help them prioritize their content maintenance, in terms of identifying differences in content, and different approaches to maintaining such content. The end goal is to find the right balance. As alluded to earlier, content maintenance is about delivering sustainable growth. It is about fostering a pool of content that can establish, maintain, and even expand its relevance to audiences without being a drag on operations.

The matrix considers content according to two properties. First, it considers how popular the content is over a defined time period, say six months. Second, it considers two broad approaches to maintaining content: through continuous optimization, and basic updating. The goal of the matrix is to encourage content managers and owners to characterize the kinds of content that belong in each quadrant, and decide how they will manage such content.

The filled out framework is illustrative. This strawman framework is intended to spark discussion rather than dictate action.

A matrix that can be used as a framework for considering different dimensions of content maintenance.

It is important to note that what’s low traffic content today may not be so tomorrow. Organizations also need to consider the diversity of their mission or corporate strategy when making decisions about how to maintain content that gets less traffic.

What Happens After Publication?

Ideally, every item of content published should belong to a specific content maintenance category. At the time of publication, it could be slugged with the category that best represents how it will be maintained in the future. Doing so can help support of the operational processes associated with maintenance: for example, determining how long it had been since the last update for items associated with a certain category.

Content requires maintenance, and maintenance requires process, not just willpower. Rules, criteria and plans are helpful to ensuring that content maintenance happens. The challenge is connecting these rules and plans to the larger purpose of the organization’s content strategy. Content maintenance needs to be upgraded from its status as a boring chore, to being seen as a contributor to sustainable growth in audience engagement.

— Michael Andrews

Tags content_maintenance, updating