“I love the idea in theory, but I still don’t know how to put it into practice.”
If you listen even briefly to what people are saying about intelligent content, online and off, you’ll hear lots of variations on that comment (which was posted March 4, 2015, by Erica Holthausen, J.D., on the LinkedIn Content Strategy group’s discussion entitled What is your experience with intelligent content?). Many people want to know this very thing: How do we get started?
The folks speaking at this year’s Intelligent Content Conference (ICC) have stepped up to share their suggestions. Read on to see how some of them answer this question …
Lots of people are giving good advice on intelligent content. So I’ll concentrate not on the basics, but on the things that I find (with due respect to my peers, of course) get left out of the conversation too often.
First, think about intelligent content not as a technology but as the formalization of your content’s natural essence. Seriously. All of your content already has structure, labels, and all that good stuff because we naturally think about content types, taxonomies, and models, we just don’t call them that. We know what a brochure is and what goes in it, and we know about grouping documents by audience, by relevant products, or by life cycle stages, but all this is rather informal or in our minds.
Making your old content intelligent is about finding those consistent structures and category labels that resonate with the largest numbers of your audience members and content creators, then modeling them into a consistent contract agreed amongst the people and systems that need to work with content. This contract states explicitly, instead of implicitly, your editorial and business rules, like which components make up each of your content types, which are optional, or how many. It should also take into account future needs for delivery and transformation of content.
Once you’ve defined your content formally, you’ve got to implement a way for a computer to validate that everyone is working off your new model. Don’t overengineer. Build models that make sense for users, are feasible and maintainable, and have clear business benefits for the future.
Don’t try to do everything, and definitely don’t try to do everything all at once. Chop the problem up and it’ll be easier to address.
Do you deposit checks with mobile banking? Arrange rides with Uber? Use Nest to automatically regulate the temperature of your house? Stream your favorite show through Netflix? To succeed in this new world, delivering superior experiences that engage your customers is critical to success. And it’s all done with software.
We live in the application economy, a place in which everything is driven by a connected, application-based world where customers experience and interact with your brand through a software application rather than a live person.
To survive and thrive in the application economy, companies are turning to high-velocity development practices such as DevOps that leverage automation, continuous testing, and increased collaboration. Yet, despite the focus and continued strides forward in software development, the content supply chain lags far behind or is totally ignored.
Your organization falls into this category if the bulk of your technical content is written by a siloed team of traditional writers. Do you publish PDF/HTML files at the time of general availability? Do your support teams write hundreds of knowledge articles that are posted after general availability to address documentation “bugs”? If your technical content is not helping to generate leads, then your enterprise is not yet ready for the application economy.
DocOps (the content cousin of DevOps) is about creating a content supply chain that is, at its core, collaborative, agile, and continuous. With DocOps, technical writers curate content from a variety of authors collaborating across the enterprise (and even external customers). Content is continuously updated throughout the life cycle of the product based on customer feedback and sophisticated analytical tools including Google Analytics and even machine learning. Both the content updates and translation occur in a near real-time fashion due to the agility of the platform.
And, as if delivering best of breed intelligent content were not enough, DocOps also aids in the sales cycle of the product by helping to attract and identify lead prospects.
Creating intelligent content in the application economy has just taken on a whole new meaning with DocOps.
At ICC, James is participating in a panel discussion, How to Increase Content Velocity Through Agile Practices, March 25. He’ll be joined by fellow panelists Andrew Bredenkamp , Laurel Counts, and Margo Stern.
Start with the business case. Understand why you are moving to intelligent content. It’s not for everyone; just because other people are doing it doesn’t mean you should. Have a clear understanding of the goals you need to achieve to succeed.
Then, move on to the content life cycle. Be sure you understand how content is created and moved through the life cycle today. What is the workflow? What works? What doesn’t?
Evaluate the content itself. Are there opportunities for reuse? How will you go about modeling the content? What are the exceptions and what are the norms?
Only then would I consider tools. Far too often (really, really too often) I find that companies choose a tool without evaluating the content and the workflow. Often, this results in chaos. Often, we have to fit the content to the capabilities of the tool rather than the other way around.
This is equally true for large and small companies. For large companies, you have far more constituents to deal with, usually more content, and often more types of content. For small companies, the budget might be the driver. Regardless, understanding the goals, the way things are done today, and the content itself allows you to analyze the gap between where you are and where you want to be. THEN you can look for technologies and workflows that will best help you bridge that gap.
Val’s ICC presentation is What’s In a Word: How to Manage Terminology March 24.
Read Ann Rockley’s book Managing Enterprise Content. Read Scott Abel and Rahel Bailie’s book The Language of Content Strategy. Read Mark Lewis’ DITA Metrics 101. Then go to a conference where you can see these authors speak. Take a ton of notes, and follow up with them. They’re all nice people who will reply to your questions about their books.
You can probably figure out everything they have to say on your own, but there are so many things you need to know about working with intelligent content that you need this kind of active curation. Buy extra copies of these books and hand them out to people on your team. This kind of thing works best when everyone is aligned in the same direction for the same goal.
Buddy’s ICC presentation is The Long and Short of Content: Strategies for Intelligent Content Planning March 24. He’ll be joined by co-presenter Michelle Killebrew, Program Director of Social Business at IBM.
If you think of intelligent content as a tree, start with the core concepts “trunk,” and then move to the “branches” in the field. The trunk includes a basic understanding of what content structure is, why it’s important, and which disciplines are involved.
In large versus small organizations, the main difference is in granularity and application of intelligent content. All content should have a basic structure, but not all content needs detailed taxonomy or relationships applied, and not all kinds of markup are needed for all content providers.
Marketers with smaller budgets can start with the basics of content structure and metadata, decide whether they need markup like Schema.org, and start with a longer-term content strategy to understand where content fits in to the outcomes for the business and its customers. Then they can apply defined content types and structure to the existing content to enable content strategy and goals, build out essential taxonomy, and identify which markup tasks are necessary to meet goals and content strategy.
Cruce’s ICC presentation is Orchestrating Intelligent Content: The Role of the Content Engineer March 25.
Start with the content; don’t worry about technology in the beginning. Review your pain points, and determine which ones can be addressed by intelligent content. Even if you have a small budget or no budget, doing the following significantly improves your content and processes and positions you to move forward with intelligent content.
If you have a budget, take the next step and develop these things:
Then, and only then, pick tools. Never pick tools until you have completed an analysis of your organizational needs, customer needs, and content strategy. Tools do the heavy lifting in intelligent content, but if you pick them first, there is a good chance your project will fail. Analysis and an awareness of your content strategy helps you to identify what you really need to deliver intelligent content, and it ensures that you evaluate the tools based on those needs.
Ann’s ICC presentation is How to Create an Agile Content Factory March 25. She also gives a full-day preconference workshop, Building a FrameWork for Intelligent Content, March 23. And don’t miss out on the Healthcare Roundtable Ann is leading March 24, along with Buddy Scalera and Ahava Leibtag.
The first step is to understand the business problem you are trying to solve. Then, attack the technical problem. Start seeing content not as a static thing, but as a living entity that you can manipulate into various containers and shapes.
Sarah’s ICC presentation is Crown Equipment Case Study: Getting Started With Intelligent Content March 25. She’ll be joined by co-presenter Jodi Shimp, Global Content Manager at Crown Equipment.
Learn about your audiences and listen to them. That’s job #1. Easier said than done, I know. But as hard as it is, it is possible only now because we can gather data about our audiences and feed it into our content publishing systems.
Here is a rough process for getting started:
Listen for keywords. Search engines are the best sources of data on what your target audiences need. The practice starts with Google AdWords. But AdWords does not provide the intelligence needed to sort and filter keywords by their relevance to your target audience. You need a data analyst or ontologist to do this – someone who can categorize the keywords into meaningful segments. One place to start is your corporate taxonomy. This is a decent representation of the way your company names things. If you can associate these values with the keywords you discover in your research, you can find out how your audience describes the categories of products or services you sell. Then you can turn your taxonomies outside-in, and start using client language in your user experiences.
Use your audience’s language. Perform an audit of your current publishing systems to determine who needs the keyword data in what phases of the process and in what form. Then build APIs into those tools to give the authors and editors access to the data as they are writing.
Measure your content effectiveness. As close as your authors can get with keyword data, it will not meet audience needs perfectly. They will miss the mark, and need to adjust content after it is published. Several metrics tools can measure to what extent your content is used. Build these tools into your authoring environment so that authors and editors can continue to learn audience preferences and adjust content accordingly.
Practice agile iteration. Authors and editors need to adjust the expectation that they will publish perfection. They need to embrace a culture of rapidly publishing their best efforts, and adjusting live content as the audience votes for its quality with clicks. This kind of culture change only happens if you give them the incentives to work with agility, and reward them when they do.
James’ ICC presentation is IBM Case Study: Making Marketing Content Intelligent March 24. He’ll be joined by co-presenter Michael Priestley, Enterprise Content Technology Strategist at IBM.
Intelligent content begins with understanding the implicit structure and semantic nature of your content. Regardless of the size of your organization, the process is more of an approach to design than a series of steps.
Imagine a conversation between your business, your customer, and your content. As a content strategist stepping up to this new role of content design, your aim is to facilitate that conversation so that your content can interact with both the customer and your business, deftly offering the right information in the right amount of detail when needed.
Start with a bottom-up analysis (that is, looking for small, common characteristics in your content). Identify some common terms used in your industry or that describe parts of a product or service. The vocabulary you or your client both talk and write about may represent bits of meaning or knowledge that, when captured, add to the intelligence of your content (by representing implicit links to related information, for example).
Then, do a top-down analysis (that is, looking for major differences across your content). First look for suggestions of organization in the existing content (inherent templates, as it were). Then imagine how you would change things, if you could, to make the existing content better suit the needs of your reader and your business.
I recommend repeating this process several times because on each new pass you will identify additional or better associations between your content and the needs of your business and your customers. The eventual model that you come up with, once documented, will be your ongoing road map for both developing and using that information in smart new ways.
At Moody’s Analytics, we haven’t yet gotten deep into intelligent content. The team has just started analyzing a wide selection of documents to identify our high-level content chunks. I am as eager as anybody to hear other responses on this question!
Even so, I already have two pieces of advice:
Get the needed experience and skill on your team
We have two key players on our CMS project team: one writer who previously worked at a company that developed an in-house CMS, and another who started as a programmer. These two worked together to develop our prototype. If you don’t have this experience and skill on your team now, be looking to bring it in with every open position. If you remain dependent on the engineers to provide technical expertise, your project will go to the back burner when product release schedules start to slip. Send other team members to training so they understand the concepts. We are using this option for those who are joining the original two on our CMS team.
Take a phased approach
Our top priority is to continue delivering documentation for product releases, and our CMS has to be developed around this. With many products, multiple documents, thousands of pages, ongoing releases – and no extra staff – we saw too much risk in attempting to convert all documents within a short period. We developed a phased approach to reduce that risk.
We broke the project into manageable pieces. Our first phase includes the development of transformation sheets that will allow us to generate PDF files and help files that are similar to the PDFs and help files we currently produce. With these, we can begin using our new system (XML) and continue using our old system (FrameMaker and RoboHelp) simultaneously without the customer noticing a difference in the final documents. Now, we can migrate our documents into the new system gradually, according to priorities that we define.
At ICC, Laurel is participating in a panel discussion, How to Increase Content Velocity Through Agile Practices, March 25. She’ll be joined by fellow panelists Andrew Bredenkamp , Margo Stern, and James Turcotte.
People seem to have little interest in figuring out intelligent content until the pain of creating and maintaining content becomes so great that it creates a willingness to look at alternative solutions.
Intelligent content solutions require two basic things:
The most counterintuitive request I hear is whether it’s possible to “dumb down” the system so that writers don’t have to think about it or whether the writers can “just write” and then have a more technical person go and do the “techie stuff.” To me, that’s like putting someone in the driver’s seat of a fancy car without expecting them to know how to use the gas pedal, the brakes, or the clutch.
Now, I’m not advocating that every driver needs to know how to put in a new transmission, change the oil, or even understand everything going on under the hood (the way content engineers work with the mechanics behind intelligent content). But if you want to be a professional communicator, particularly one who can create intelligent content, you’ve got to learn at least the basics. You must know how to adjust your mirrors, shift gears, steer, etc. (structure your content properly, create content for reuse, apply metadata tags appropriately, etc.) so that you can get where you want to go.
Rahel’s ICC presentation is Stories That Sell: What Content Marketing Can Learn From Content Engineering March 25. She also gives a full-day preconference workshop, The Nuts and Bolts of Intelligent Content, March 23.
In my opinion, modularity is key when setting up your content. What does that mean? It means that the structure of your data allows pieces of information to be reused – used unchanged – in as many places and contexts as possible.
Take a company that sells motorcycles. This company must provide an owner’s manual for every motorcycle. Some details will vary across the manuals, but other information appears in every manual in the same way, for example, warning labels. Every motorcycle has brakes, and the underlying principle is always the same: Don’t lock up the wheels under braking. Finish braking before going into a corner. Don’t overheat the brakes.
With modular content, these warning labels can be saved in one location. Instead of copying and pasting them into separate owner’s manuals, writers can link to them so that every document refers to the same set of warning labels. This way, every label only needs to be translated once, and changes are made in one place only instead of manually changing every occurrence.
Small companies may reap the benefits of modular content faster than large companies because the ratio of reusable to variable content is more favorable if only a small amount of product variation is documented.
—Michael Arnold, Technical Editor, KTM Motorrad AG
Michael’s ICC presentation is KTM Motorcycle Case Study: Lessons Learned in Developing Intelligent Content March 25. He’ll be joined by co-presenter Alan Horvath, Managing Director of STAR Group America, who will share additional insights into the benefits of creating and maintaining technical information in intelligent content structures applicable to any industry sector.
If you’re among the many who are drawn to the idea of intelligent content but aren’t sure where to begin, I hope you’ve found a few pointers here that make sense for your situation. Maybe you found it reassuring to hear that you don’t have to do it all at once, and, in fact, you couldn’t if you tried. You may have found it useful to hear that you can – and should – get started with a common-sense, business-minded evaluation of your content before you give a thought to expensive, complex tools. Isn’t it nice to know that you can start today without buying a thing?
Still, you have a challenge ahead. Many of us are looking for examples to follow. In the LinkedIn discussion I referred to previously, Joe Pack speaks for many of us, I’m sure, when he says, “I wish someone could predict the future and give us a little insight into how we should structure content marketing strategies over the long-haul. I guess what we’re all going to have to do is set ourselves to continually adapt and change as trends come and go.”
You don’t have to be a marketer to share Joe’s sentiment. People in various roles across the organization are looking for stories that can “give us a little insight” into ways intelligent content might work for them and their departments.
Do you have a story or insight to share? Got an example that others might be grateful to hear about? Please leave a comment below.
Want to hear more from these speakers? Want to learn more about getting started with intelligent content? Here are three things you can do right now:
Marcia Riefer Johnston (@marciarjohnston), managing editor of Intelligent Content for the Content Marketing Institute, has attended several Intelligent Content Conferences and looks forward to the next one. She is the author of Word Up! How to Write Powerful Sentences and Paragraphs (And Everything You Build From Them). See Writing.Rocks.
The answers provided by the ICC speakers above were submitted via email and edited according to Content Marketing Institute style guidelines.
Title image courtesy of Joseph Kalinowski, Content Marketing Institute