By Scott Abel | Senior Member
In the digital age, change happens quickly. This column features interviews with the movers and shakers—the folks behind new ideas, standards, methods, products, and amazing technologies that are changing the way we live and interact in our modern world. Got questions, suggestions or feedback? Email them to scottabel@mac.com.
The Content Wrangler: Michael, tell us about yourself and what you do for a living.
Michael A. Rosinski: My background centers on sales, marketing, and finance. Over the course of three decades in the dynamic world of high technology and application software, I’ve reinvented myself, as many technical communicators have done, moving from mainframes to midrange to the Internet to mobile to social media.
I have a deep background with Wall Street, including two initial public offerings in the ERP space and two mergers-and-acquisitions transactions to my credit. I was a founding executive with American Software and trademarked them as the “Supply Chain Management Company.” Being the first to market with supply-chain software ranks up there with what Facebook, LinkedIn, and Twitter have accomplished in social media.
On the international landscape, I spent three years in Paris as president of Augeo Software in professional-services automation, helping to develop a solution that automates project-management functions.
In 2006, I was hired as president and CEO of Astoria by Goldman Sachs, the premier Wall Street investment firm and an investor in Astoria, to turn the company around. I have made the company profitable, introduced the first SaaS (software as a service) platforms and mobile platforms to component content management, and sold the company to TransPerfect, the largest private translation company in the world.
TCW: What is a CCMS, exactly? And what makes it different from other types of content management systems? Tell us about your flagship product, Astoria CCMS.
MAR: All content management systems (CMSs) are not created equal. Some are document-oriented, managing files—whole documents—and the attributes attached to those files. These are systems like Microsoft SharePoint and EMC Documentum. Other CMSs are Web-oriented, giving IT control over the display of HTML pages. Oracle Stellent and Adobe CQ5 are two notable products of this type.
Separate from these two types of CMS is the component CMS, or CCMS. A CCMS manages XML-tagged content components. Each component has its own history, permissions, and life cycle. Since each XML component is independent, a CCMS promotes content reuse, content collaboration, and translation efficiency more than document-centric or Web-centric CMSs do.
Even within the CCMS world, all systems are not created equal. For example, some CCMSs, including Astoria CCMS, take a NoSQL approach to storage and retrieval, which results in more consistent performance as the number of XML components increases. To achieve the same consistency, a SQL-based CCMS would require substantial (and expensive) hardware upgrades.
To show the kind of problems a CCMS can address, here’s a scenario from one of our customers. A team of over 60 writers were building and maintaining 4,000 technical manuals—18 million words—in Adobe FrameMaker and using a document-centric CMS to manage that content. They were spending $7 million annually to translate these manuals into 20 languages. When they converted their content to XML, they found themselves in charge of millions of XML components. Millions of pieces of content. How were they going to manage them all?
They came to us for help. The problem was architectural. Their CMS managed 4,000 documents as documents. It wasn’t designed to manage millions of XML components. That’s what CCMSs do. And CCMSs can enable access to that content by even hundreds of writers around the world. We set up a system for them, imported all their XML content, and turned their writers loose to reuse that content, prepare it for translation, and produce output for delivery to their customers.
Over the next couple years, this company saved over $6 million in translation costs alone. They also cut the effort required to create and maintain content in half. That’s cutting over 60 writers’ work hours by 50 percent. This company also discovered that broad changes, such as rebranding, could be effected in minutes instead of weeks. Finally, the writers on this team say that their ability to find reusable content and the system’s ability to track content versions make their jobs easier.
TCW: What type of content is best to manage in a CCMS like Astoria?
MAR: To answer this question, I need to convey the environment for which systems like ours were designed. We serve companies that have what we call the 3 V‘s: high volume, high variability, and high velocity of data for technical product documentation. These kinds of companies are under tremendous pressure to release new products in competitive markets and they face myriad obstacles, such as regulatory scrutiny, product differentiation for niche markets, acquired companies whose content they need to integrate, and the list goes on.
Technical documentation is the intellectual property underpinning these companies’ products. So for any company that produces a complex product requiring technical documentation (for example, verticals such as high technology—hardware and software—medical devices, discrete manufacturing, telecommunications), a CCMS is a must-have, not a nice-to-have.
What writers store in a CCMS are reusable components of XML-tagged content. A content component might be a topic for an online help system, a section for a book-oriented user guide, or some other core idea, concept, task, or reference component that exists as an indivisible unit of information. Each customer defines that level of indivisibility and builds a library in the CCMS from those components. Each customer cross-connects those content components in various ways, mixing and matching them as needed to build various documentation sets.
For example, in a regulated industry, such as medical-device manufacturing, all products must be accompanied by the correct notes, cautions, and warnings. Writers in regulated industries assemble a corpus of approved material that is then cross-linked into “Instructions For Use” documentation. Cross-linking ensures that there is no variation in the notes, cautions, and warnings, ensuring compliance with legal strictures and regulations. Companies producing documentation from a corpus of reusable knowledge components are less prone to mistakes in product labeling, and they thereby reduce their exposure to litigation arising from mislabeled products.
This is just one example. I could cite dozens of others in both regulated and nonregulated markets.
TCW: What are typical uses for a component content management system?
MAR: One use case for a component content management system is unifying the teams across technical communication, marketing, customer service, and other departments. In the absence of a CCMS, these departments often duplicate the effort of creating and maintaining information. As the amount of content grows, this duplication of effort results in exponential growth of redundancy, inconsistency, and error.
A CCMS, when implemented thoughtfully to support integrated processes across departments, enables companies to present consistent information to the customer while supporting cross-department efficiencies.
Breaking down silos and speaking with one voice: these two reasons are precisely what drove one of our customers to adopt a CCMS. The marketing department spearheaded the effort to unify all customer-facing content so that customers would have a common and consistent experience whether reading brochures, instructional material, or repair guides. The marketing team, led by a vice president who understood the value of correct and consistent content, convinced each of the company’s line-of-business leaders to subscribe to what they called their single-voice project. A global documentation team reporting to the marketing VP provided guidance and training.
That company chose a CCMS as the global repository for all business units, making common content available to each writing group. Working together, these groups dissolved content silos and they systematically identified and promoted content components that had potential for reuse. The CCMS easily accommodated more than 400 GB of data and graphics, giving users a reliable platform on which to achieve the marketing department’s vision of a single-voice experience for its customers.
TCW: Why would a technical communication team want to consider using a CCMS like Astoria instead of a CMS like SharePoint?
MAR: A CCMS lends itself to managing content—specifically XML-based content and especially DITA content—better than a system like SharePoint for several reasons. Only a CCMS does these things:
- Flags content for translation
- Maintains cross-reference links
- Supports single sourcing
- Supports advanced release management
First, let’s look at the translation advantage. When content changes, the CCMS keeps track of which affected components have already been translated into the target languages. The CCMS then helps localization coordinators build translation packages that contain only those files that need to be retranslated.
For some companies, this capability alone provides enough ROI to justify the CCMS investment.
For example, let’s say your company has 250 documents for a certain product, and those documents have all been translated into 10 languages. Along comes a product change requiring revisions to several safety warnings across those documents.
In this scenario, if you were using SharePoint—which “knows” nothing about XML and, consequently, has no idea which content components change between revisions of a file—you would have to send all 250 documents to your language service provider. Even if only a tiny percentage of the content within those documents is changing, you would incur charges for evaluating the entire document set against translation memory for all languages on top of paying for the new translation. Furthermore, SharePoint doesn’t support segmentation rules or cross-references. Without getting too technical, this means that the material sent to the service provider may be far more or far less than what’s needed, complicating the localization coordinator’s life.
In plain English, if you have a lot of translated content and you don’t have a CCMS, updates are an expensive pain. A document management system like SharePoint leaves companies no choice but to pay translators far more than is merited by “tiny” content updates—and to pay their own employees to find and prepare all those documents up front and then process all the updated translations afterward.
In our example, we’re talking about 2,500 translated documents.
Now if you were using a CCMS in this scenario, the system would know which components within those 250 source-language documents had been altered and needed retranslating. A CCMS eliminates the need for comparison with translation memory. It enables people to single out the components affected by any given change and to flag only those components for the translators to update.
After those components are updated in all languages and reimported into the CCMS, all linkages that existed in the source-language material are preserved in the target-language material. Document management systems like SharePoint can’t do this.
Let’s examine that last point more closely. If you’ve ever moved content around in SharePoint, you know that links break. They break because SharePoint is a hierarchical file-management system; links are based on file names and folder structures. In a CCMS, content components are not stored in a hierarchical system. Instead, each content component has a unique ID that never changes. Other components link to that ID. So, regardless of how many times a component moves around in the repository, the links never break.
Back to the main question—what can CCMSs do that other CMSs can’t—another huge differentiator is single sourcing. Have you ever tried to repurpose content in SharePoint, such as converting HTML content for display on mobile devices? It’s somewhere between cumbersome and impossible because SharePoint has no publication pipeline capability, and it provides no interface for producing differentiated output from a common set of XML files.
With a CCMS, it’s another story. CCMSs are designed to single source. The Astoria CCMS, for example, has a publication interface that allows you to specify filtering criteria and output types in an single operation. You can produce HTML, PDF, Word, and Eclipse Help outputs in a single operation from a single set of source files. Imagine doing this for one set of products and then for another set of products seconds later. With SharePoint, you couldn’t dream of doing this. This is what CCMSs are built for.
Finally, SharePoint does not support advanced release management. Users must invent file-naming conventions and other procedures built around SharePoint’s file-compare capability. SharePoint implementation teams have to invent their own version-management and release-management mechanisms.
Not so with a CCMS. Release management is a core capability. CCMS users—at least, I can say with confidence, Astoria users—can define any number of releases, branch and merge content across releases, add or remove content from specific releases, and publish any release, even historical content that has been updated or overwritten.
This capability is particularly helpful when you need to freeze content versions for regulatory purposes. Freezing or archiving content requires you to preserve snapshots of your document set with each change. To do this without a CCMS, you have to create what amounts to a content silo, which organizations want to avoid. A CCMS can reconstruct any document set from any point in its life cycle. No need to copy content to an archive location. No content silos.
TCW: You mention DITA. What more might you want to say?
MAR: Astoria is agnostic on the question of DITA. Our CCMS supports any DTD.
TCW: In your experience talking with leaders in the C-suite, what are some of the big challenges facing enterprises today?
MAR: Executives want to offer an excellent customer experience while reducing costs. One way to do that is to use SaaS products. A SaaS-based CCMS offers a one-stop shop for software, hardware, upgrades, infrastructure, server, peripherals, consulting services—all conveniently bundled at a predictable level of expenditure. Many companies don’t have the necessary experience to go out and acquire all these components on their own and to choose the correct amount of infrastructure to support the anticipated enterprise-wide scale of the project.
Another area challenging C-level executives is the mobile revolution. In the future, few enterprise applications will remain on the desktop. Most will be distributed on mobile platforms. So when you research CCMSs, it makes sense to look for systems that enable people to produce, review, edit, approve, and publish content using mobile devices—smartphones, tablets, etc.
Beyond the ability to push content out, executives are looking for useful ways to bring customer feedback in. They want to engage with customers. While I can’t speak for other CCMSs, the Astoria CCMS delivers this level of engagement by enabling your company’s customers to provide feedback that helps writing teams identify the most important topics. Astoria gathers search criteria, search results, and prompted feedback, and then provides this correlated information to the writers. The writers can then improve the topics their customers care about based on what those customers need.
Why does this kind of capability matter to leaders in the C-suite? According to a report released by Oracle Corporation in December 2012, 81 percent of the consumers they surveyed said they would pay more for a superior customer experience. Writing teams that have customer feedback data correlated with specific content topics are in the best position to improve customer experience.
TCW: Astoria is owned by TransPerfect, a major player in the translation-services industry. Recently, TransPerfect bought Vasont, a competitor in the CCMS market. What does that mean for customers and prospects?
MAR: The Astoria system and the Vasont system serve different types of customer. The Vasont CCMS offers a department-level solution for small and medium-sized businesses. The Astoria CCMS serves the distinct needs of large companies—like Cisco, Siemens, General Electric, and Micron—at the enterprise level. It handles massive amounts of global data and provides access to unlimited users with fast response times.
For more, see the 27 January 2014 TechWhirl interview with Phil Shawe, co-CEO of TransPerfect: http://techwhirl.com/transperfect-vasont-systems-merger-path-one-stop-solution-global-clients/
TCW: What mistakes do you see companies making as they explore content management options?
MAR: One mistake is scheduling demos with vendors before discussing their true pain points with those vendors. This approach puts the company’s people in the difficult position of “translating” the features they see in a demo to their imagined resolution of pain points in their current processes. Unfortunately, we have seen prospects ask questions that reveal a fundamental misunderstanding of what a CCMS can do.
A wise manager approaches vendors with pain points. The vendor needs to understand company goals, directives from senior executives, business processes, and tools. The vendor needs to cover the company’s issues, successes, and failures. The kind of “discovery call” I’m talking about has no demonstration component, just honest discussion between equal partners. If there’s a good fit between a company’s pain points and the vendor’s solution, then it makes sense to schedule a demonstration.
A company’s preparation for this kind of discovery call with a CCMS vendor is an introspective exercise in honesty. Before reaching out to vendors, companies need to assess their people, their processes, their operating environment, their hopes and dreams—everything that goes into the craft of building, managing, and delivering product information—to determine what, if anything, they want to change.
Another mistake I see is that companies fail to do proof-of-concept trials to test the system for several weeks and discover how it works for them. I’m talking about doing a trial. Not just scheduling a trial. We see plenty of companies scheduling trials. They assemble their evaluation criteria. They identify their evaluation teams. They mark the start dates and end dates on their calendars. That’s scheduling the trial.
Few companies get around to the doing. They don’t allocate time for people to run through the criteria. They don’t have regular calls with the CCMS vendor to discuss and resolve issues discovered in the trial. They don’t test the vendor’s responsiveness and appreciation of their needs. That’s unfortunate, of course, because doing the trial helps both the customer and the vendor. Doing the trial reveals many things about the fit between the CCMS vendor and the company. Doing the trial reveals the character of the product and service beyond superficial examination and light testing of features and functions.
If companies want to prove out the benefits of a CCMS, they need to commit to doing a proof of concept.
One more mistake: some people focus on price when they should focus on return on investment. A CFO will approve the purchase of a CCMS if presented with a rigorous ROI analysis. I can’t speak for all vendors, but at Astoria Software, we help people produce such an ROI. We help our prospects focus on verifiable benefits and the monetary value of those benefits to the organization. Our ROI-discovery process strips out fantasy projections and nebulous benefits, boring in on the elements that drive cost savings, productivity improvements, and the ability to embrace future initiatives. These are the issues that executives want to hear about. Our ROI helps documentation teams make the right case to those leaders. This process presents the financial backing to make the business case to the CFO, who will see the value of spending x to get 3x back within three years. Nevertheless, it’s surprising to me how infrequently that information is packaged and delivered to budget-wielding senior management.
TCW: The discipline of technical communication is experiencing a reset. Gone are the days that writers were paid top dollar to “document the system.” Outsourcing, off-shoring, and other business models for creating content are shaking things up. Today, the top-paying jobs require far more than writing. What skills do technical communicators need to develop to remain relevant?
MAR: The traditional view of the technical communicator as product documentor, at least as well-paid product documentor, may be disappearing. But I see great opportunity ahead for those who step up to the content challenges facing companies today. Companies need skilled technical communicators now more than ever.
Specifically, companies need technical communicators who know their own discipline and understand business goals. Who can explain how things work and think strategically about content. Who can think postsale and presale. Who can create information that works for end users and for translators. Who can create content worth reusing and can structure that content for reuse.
I encourage all technical communicators who want to improve their value to their employers and clients to participate in events that can help them grow in these directions. The Intelligent Content Conference and CIDM (Center for Information-Development Management) events come to mind.
Technical communicators who expand their skill sets in the ways I’m talking about will lead their companies straight to the business advantages that content management systems are designed to deliver. Those technical communicators will find themselves in demand for years to come.
