Interview with Noz Urbina, Mekon UK

TCW: Noz, thanks for taking time out of your busy schedule to chat with us today. Before we get started, can you tell our readers a little about yourself and the company you work for?

NU: I’m a Senior Consultant at Mekon Ltd., a UK-based specialist consultancy house supporting organizations with content issues. I’m myself a bit of a content solutions Swiss Army Knife, or Swiss Army Noz, if you’ll let me get away with that one. Mekon’s got a diverse range of services including consultancy, systems implementation integration, workflow automation, and training around content technologies, so they keep me well stocked with juicy problems to solve.

Generally, I am most useful in situations where several stakeholders in a content solution – management, different departmental heads with their own goals, technical communicators, engineers, and 3rd party vendors – all need to come together to work out a way forward that is feasible, cost effective and that delivers on an organization’s business goals. I spent the 5 years before Mekon at a major XML Authoring tool vendor working with Mekon and other consultants that implemented integrated content solutions for clients around Europe.

TCW: So you’ve moved from software sales to pure consultancy. What do you see as the big difference between providing consultancy as opposed to selling and consulting working for a vendor?

NU: Content solutions, especially things like XML and Component Content Management, have had their ‘hay day’, where the market was so small and bleeding edge, that if you had the skills (or the software) you could get pretty far without having to be scrutinized on things like ROI and incremental performance increases. Projects were often led by ‘technical visionaries’ who were inspired by the technology and what it offered in theory. Lots of big projects were launched, a few succeeded, but many failed.

I spend a lot of time telling people what they can’t actually achieve within their budget, despite what a software vendor might have told them, and keeping their enthusiasm up for all the great things that they can still accomplish if they’re organized, pick their battles wisely, and scope their phases properly. Failing to set up and to communicate the right scope and project outline are the things I’ve seen kill the most projects over the years.

TCW: Mekon has had a long time association with the X-Pubs brand. Tell us a little about X-Pubs and let our readers know how they can participate in its annual conference and keep abreast of your other events.

NU: In a sentence – I’d say X-Pubs is a low-tech technology conference, focused on helping people get the most from their information assets. I know a low-tech high-tech conference sounds ridiculous, but in a market so fraught with both confusion and enthusiasm, there was precious little in terms of events focused on tackling pragmatic business issues.

Sometimes we even have presenters who don’t present on XML. Why? Because if you’re really going to tackle the how-tos and pros and cons of XML head-on, you have to hear about things which aren’t specific to XML, like Emma Hamer’s very well received presentations on handling disruptive / innovative IT implementations, or from people like the guys at Author-it, who feel that native XML isn’t necessary for many applications.

We wanted to answer the questions people weren’t answering, like:

Can a small company show ROI on XML or is it just for the big guys?
How do I get this stupid project off the ground already?
How can a technical communicator get this “C-level attention” I’ve been told I need?
So… how does one “use XML to bridge engineering, training and tech comms” exactly?

Every year we mix global community pillars in the DITA, S1000D and content management worlds like JoAnn Hackos, Michael Priestley, Ann Rockley, Svante Eriksson, with new thought-leaders, and many actual users like Schlumberger, the Irish Government, Bombardier, BMJ Group, ITT Flygt and so on.

This year we’re encouraging more government and training content that previous years, so we’re looking at the DITA-training subcommittees, online help community, and the public sector for speakers. We get everyone from microchip manufacturers, to banks, to people who make brain scanners and nuclear missiles all in a room together and hash out the problems that are common to anyone with complex content challenges.

TCW: That’s exciting! We’ll continue to update our readers on X-Pubs conferences and webinars. Can readers of The Content Wrangler receive a discount on registration? If so, how do they claim it?

NU: Webinars! Good point! The annual pre-conference webinar series is a great way for newbies to get a ‘sneak preview’ of what’s at the conference, as well as get up to speed and get the most out of the day itself. Your readers can get a discount code either by attending one of them, or directly from The Content Wrangler newsletter (free subscription required). These codes are then usable when registering on http://www.X-Pubs.com.

TCW: Mekon helps organizations who value their content as an asset find better ways of creating it, managing it, and delivering it to those who need it. What are some of the primary business drivers for your clients? And, what types of problems are they attempting to tackle and how does Mekon help?

NU: That’s a mouthful, Scott! What’s been interesting for me lately, and I hope interesting for your readers is seeing how the market has developed in the last 5 years. Because of the place we managed to secure in the UK marketplace, the Mekon project list can be used as a sort of ‘barometer’ for what’s going on in the XML / CMS world. Unlike the years pre-2001/2, What we’re seeing now is we’re being asked into projects that are already at every possible point along their lifecycle; from initial requirements analysis, to migration from an existing XML or CCMS solution that has been outgrown, or worse, failed. This has forced us to refine our methodology to suit organizations big and small, and hit the ground running no matter what this rapidly evolving market chucks at us.

In the last year:

We’ve helped several clients, for example Surf Control software and Elekta Medical Devices, get the ball rolling with their DITA / XML content reuse projects. We validated their initial thoughts and goals, formulated initial requirements lists, helped structure their business case and presented it with them to management to start the projects off. For companies just starting out we have a package called a ‘Content Strategy Audit’ which helps them dip their toe into both content solutions, and into working with us.
For STMicroelectronics and a few others, we were brought in to an already established business case and asked to choose the right teams, workflows, content and scope to make sure a pilot project would be a success. Our ‘Scoping and Planning’ workshops are designed for this. Whereas 5 years ago we were just scoping out XML content reuse systems for tech communicators (using pretty techie interfaces), now we’re outlining solutions that take content in DITA format from it’s birth in an Engineer Specification straight to Web 2.0 knowledge portals and machine-specific PDF at the same time.
We’ve had three companies already using XML, and some even already using a component content management system they didn’t like approach us for guidance on migration to a NEW way of using XML, as they’d gone it alone the first time and were struggling to maintain it all. At one of these, Micro Focus software, XML content creation and publishing were already in place based on custom DTDs. We helped them evaluate the suitability of DITA and wade through seas of CMS vendors out there by providing a recommended short-list based on their needs. After they had decided to ‘go DITA’, we helped them convert their content over. We used a process we’ve defined for technology evaluation to help them choose something that could grow with them for the next 5 to 10 years.
Two large organizations, one major UK medical publisher and one major software developer, that had both already selected and installed the X-Hive Docato component content management system asked us to develop customizations and enhancements on the system to help them get the most value from their tools.

As you can see Scott, this market, and therefore our work, is getting more diverse. When Julian, our Managing Director, was developing his keynote for the DITA Europe conference, as part of his presentation prep we did a little ‘round up’ on the companies we’d helped recently. They ranged from 100-60,000 employees, and did everything from manufacture medical devices and warships, to publish software, to drill for oil or manage government policies.

TCW: The market in the UK is not the same as the market in the US. Regulations, standards, and business drivers are often different. What differences can you see between the US and UK markets for XML authoring and content management? Are knowledge workers in the UK solving different types of problems or just approaching them differently?

NU: Like all companies, our clients are trying to get more with less: More customer satisfaction, higher quality, more accuracy and consistency in communications of all kinds (manuals, tech bulletins, websites and KBs); less time, less cost, less pain. Unlike in North America where downsizing is famous as a management goal, our European clients tend to drive towards avoiding new headcount, and that means getting organized, collaborating, and squeezing the most of all the information assets you have. We have a lot of tech comms clients who feel squeezed, let me tell you!

TCW: Noz, you and I are often referred to as “structured content evangelists” because we promote the value of moving away from unstructured content and try to help organization see the many benefits structured content can provide them. However, moving to structured content usually means moving to XML. And, XML is still a scary and often misunderstood markup language, especially in the technical communication and training spaces, where the popularity of the Darwin Information Typing Architecture (DITA) has created even more confusion.

What are some of the primary challenges you see that are preventing folks from moving to XML…and, by extension, DITA?

NU: I actually see XML or ‘traditional XML’ as I call it, and DITA having some different challenges, but more on that in a minute. For ‘traditional XML’ I would list these major stumbling blocks:

Badly scoped migration or implementation plans—Sure this is one of the main problems of any major IT solution, so it makes sense it would be top of my list, but what’s interesting is that XML is not driven by IT departments, it’s driven by departments like tech comms, training, and product management, that want to single source content. These guys aren’t as experienced as IT at writing technology migration plans, so when tech and non-tech collide, you get disaster. With XML product vendors all tripping over themselves to make their solutions all seem easier to adopt than their competition, the poor customers are left with a very skewed impression of what can be realistically be done quickly, what should be done with caution, and where a bit of planning can make their whole implementation faster and more painless. I could write whole article on this one alone, so onto...
Too much flexibility, combined with lack of understanding of how to apply the ‘Semantic Mark-up’ bit of XML—As you said, Scott, it’s all a bit scary and a lot misunderstood. Writing a DTD is easy and I could teach my Mom to write one in an afternoon. Writing a good DTD is very, very hard! (And yes, FrameMaker users, this means EDDs too.) Just because you can edit them doesn’t mean you should be designing them from scratch. The huge rift between these is where people tumble in screaming. They so often don’t know how to model information in XML such that it’s both flexible enough to be usable, yet imposes enough of the company’s business rules that it actually adds the real value they were after. My advice on this one: don’t bother. If you’re a tech communicator, trainer, or IT head, your management wouldn’t ask you to fix the building’s plumbing or run the in-wall electrics. These are key to the office infrastructure, but they’re specialist tasks and even if you managed them, in the long term you could be in for some ugly surprises! You could accuse me of trying to shill our services here, but you just have to ask companies who have done it whether they would do it again 5 years later. Why do you think DITA’s so popular? The siren song of custom DTDs are a factor. Don’t get smashed on the rocks.
Cheap, usable authoring tools— Yes, they’re more cheap and usable than 5 years ago, especially now with DITA on the scene, but it’s still often not enough to allow XML to ‘break’ into some of the use cases I’d like to see. This is historical effect caused by the fact that while all the CMS + Authoring Tool combinations were being forged, it was only commercially compelling to developed pre-cooked integrations which you could be assured were going to sell into your core markets. When your core market changes or their usage patterns evolve, you don’t go dropping your prices or re-coding a new solution, you try to sell them what you already have at a price you know you can get. I think if the authoring tool vendors offered floating licenses more easily to allow clients to access their tools on demand instead of on a per-seat basis, we’d see ‘natural growth’. Organizations could then increase usage at an organic pace, instead of having to justify an outlay for ‘seats’ every time they want to try a new workflow with new users. On the usability, there’s a big difference tool to tool, but most need to put in real effort here.

For DITA, as opposed to ‘traditional XML’, it removes many X-factors from XML, and that always helps. However, the above 3 are all only reduced, not negated. Additionally, DITA introduces much more strict guidelines on modular writing. Unplanned reuse is difficult to manage, especially without a CMS, and planning componentized reuse is foreign to most authors. Even those that think they’re writing in a ‘topic oriented manner’ really aren’t, and there is an often irritating paradigm shift when you move to enforcing these kinds of creation guidelines. We’ve seen this as the main area of pain for organizations making the change.

TCW: In your experience, what are some of the common misconceptions surrounding DITA?

NU: One of the problems is actually how effective DITA can be. If you’re looking for reuse, localization ROI, or easier adaptation of content, so many companies are in such a bad state that DITA feels like it’s working great even when they’re only getting 30% of the benefits. Here’s my pet peeve list:

“I don’t have to work out an information model or writing guidelines, the DITA standard provide them.”—No no no! DITA is very wide and very flexible, as all broad-reaching standards need to be. To get benefits like readability, consistency and especially reusabililty, you still need to have strong guidelines on how DITA should be used.
You need to define:
- What are your title capitalization rules?
- What tense are you going to write in?
- How are your map templates going to be structured and relate to one another?
- How are you going to categorise and sort all these myriad topics on your file system/CMS once you loose the handy ‘book’ wrapper that made things so clear (i.e. what’s your metadata model)?
- How deep will you nest sections/topics – and very tricky on this, are your staff really going to know the difference between sections and topics?
- Many more...
If you don’t have these types of questions resolved as part of your information modeling process, then you’re going to end up with a bunch of non-reusable topics that you’ll never be able to find when you need them, CMS or no.
“I need a CMS to use DITA.”—Again, for Julian’s keynote at DITA Europe our industry survey results combined with our own client base showed a 50/50 split on DITA off the file system (sometimes just using source control for version control) and actual CMS packages commercial or otherwise. Both work, and you need to really think about it if you’re going to make the leap. This is why we have defined such a detailed methodology for the early stages of decision making. If you start up the wrong road, changing the inertia is expensive and painful.
“Conref!”—Like all flexible powerful technologies, conref is again misunderstood. It is equated to FrameMaker Text Insets, to variables (in FrameMaker and elsewhere), to xInclude, HTML includes and so on. The fact is it is like all of these – it is a multi-purpose reuse feature that allows content of any size to be shared between two different topics. CMS vendors struggle with implementing conref because for technical reasons a database struggles allowing all that flexibility. They had pre-cooked variable and fragment reuse functions and most will pigeon-hole the mighty conref into the little boxes that work nicely in their system. Debate also rages against whether conrefs should each be hived off in their own file (or managed as sets, like all your warnings and cautions in one ‘source topic’) to be reused across the data set.
“Information Architecture and Taxonomies!”—These terms conjure up words like ‘scary’, ‘complex’, ‘expensive’, ‘going to get me fired for not knowing...’ and more. True, doing great information architecures and taxonomies is hard, and specialists are valuable here, but whomever you are, try your best to sit down and make a plan of your information as a content set. Think out of the publications box and try to get into this Lego mindset, where you have category labels (info types and metadata taxonomies), and define what blocks you’re going need and how you’re going to sort them. This is a huge boost to your chances of getting the most out of DITA, and as a huge side-benefit, sometimes if you do a good job of defining blocks and instructions, you can for the first time hand off responsibility of whole categories of block to other people to do for you (or maybe just to do the drafts?). IA is starting to become less obscure, and I think 2008 is the year of the Taxonomist. We’re going to be having a few talks at X-Pubs to ‘de-black’ this black art.
“With DITA I’ll be able to know where I’ve reused things and it’ll be easy to make new information products.”—See my first pet peeve.
“The DITA open toolkit will provide me link management, relationship visibility, and broken link reporting.”—You’d think so, but no. If you want this then you’ll have to get a third party tool. Unfortunately I’ve not seen anything in the Open Source world that does this for DITA yet.
“I am going to have to rework all my content into topics before I start to migrate to my DITA tools and processes.”—Well, no. Purists, many of whom I respect a lot, may say yes, but in fact you can load your content into ‘dumb’ DITA, where topics aren’t properly topics and things that should be conrefs aren’t. This at least gets you over the format hump, and once you’re converted you can often get high priority publications out the door and in parallel be working on your all-important information architecture plans.

TCW: We both work frequently with software vendors who provide authoring tools that help writers create structured XML content. What are some of the biggest flaws in current authoring software packages? Are there features your clients want but no vendor seems to offer?

NU: Phew! You knew I’d love this one. I’m going to be gentle and not stab at any tools by name, but I just give the top items that cheese me with most of them:

XML is a hierarchical structured language, if I have complex structures like sections, bullet lists, or task steps, with lots of sub-elements (paras, list items, list items with paras, etc.), I should be able to easily reorder these by just clicking buttons that say ‘move up’, ‘move down’ – there could even be shortcut keys for this. This simple restructuring is almost always convoluted and annoying in the big tools.
Sort out the stupid insertion context! It’s been decades and still almost all the tools just tell you what you can insert at the cursor. Few tools even give you a good sense of where you are in the tag hierarchy much less give you the options you want, e.g., if you’re at the end of a topic, and you want to insert a sibling topic, most tools make you insert a subtopic and then promote it to a sibling instead of providing an ‘insert after’ or ‘insert before’ list of elements.
Multiple views - rendered with tags, plain text with validation optionally on and off, publishing preview and a fully ‘non-tagged’ – are really important. All tools should have these.
Ability to open up a branch of a DITA map in a single editing window. There’s no real reason other than the fact that they’re not built this way now, that they can’t render multiple topics in context providing users more visibility on groups of related topics and the ability to, temporarily at least, treat them as a ‘set’. DITA doesn’t prescribe that authors have to create topics in an information vacuum just because each is ‘stand-alone’. In fact, seeing other topics in the info set prevents you from writing redundant stuff without having to keep the whole info architecture live in your head.

TCW: Do you see a markup language like DITA playing a role in video? Movies are modular – each segment is filmed independent of one another. These video content components should – hypothetically – be able to be managed like components of text content and reassembled dynamically, on demand. Should we be able to pull video components from a repository and reassemble them as needed? Or, can we do that now?

NU: I’m running out of time, but I will say briefly that yes, and in fact I’ve seen implementations of video indexing and time-stamping used to trigger the display of matching indexed topics on the fly creating a ‘multi-path’ video narrative.

TCW: Anything you’d like to add? Perhaps a shameless promotional link you’d like us to pimp for you?

NU: Well, I’ve already mentioned X-Pubs.com, so I’ll just mention it again! We’d love to meet your readers there!

TCW: Thanks, Noz. We really appreciate your time and effort.

NU: Thanks Scott. It’s been a pleasure as always.