CMS Watch

As readers of the SharePoint chapters of our Web CMS Report and Enterprise Search Report know, there are many good online and book-length resources about MOSS 2007 in general, but almost all of them treat the platform's WCM and Search services in a fairly cursory fashion. That's too bad, because those parts of MOSS are brand new and -- customers tell us -- frequently quite puzzling, even after wading through all the Microsoft documentation and webinars.

So I was pleased when Alan pointed me to this 4-day training course exclusively about Web Publishing in MOSS. "Publishing sites" in MOSS look and behave quite differently from your typical SharePoint team sites. This kind of training is, I think, long overdue in the marketplace...

The changes at WebTrends continue.

A Friday afternoon email from the company explained that CMO Tim Kopp would be leaving at the end of this year. This follows the Halloween Day exits of CEO Greg Drew and 3 other managers and promotion of ClickShift co-founders John Rodkin and Leo Chang to engineering and hosted operations management. They served as co-founders of ClickShift which was acquired by WebTrends in 2006, providing the foundation for the WebTrends Dynamic Search product.

How does this affect you if you're a WebTrends' customer? Let's consider 3 areas of potential concern:

• Technical Support: There's been a change of management for hosted services. WebTrends' has many licensed customers that it would like to move to the hosted services model. If you are a licensed customer, I suggest you speak to your account manager regularly to stay current on potential service changes meant to "encourage" you to switch to the hosted model.
• Product Development: If you are considering the purchase of Score or Visitor Intelligence, get a clear understanding regarding pricing, support and the product development path. The management who left the company were key figures in bringing these deeper analytics solutions to market. It's not clear at this point what the new CEO and CMO will want to do with these offerings. Given the promotion of Rodkin and Chang, it seems logical that WebTrends will focus on doing more to sell WebTrends Analytics and Dynamic Search.
• Customer Commitment: If you are an enterprise customer of WebTrends, and have a significant investment in using their solutions, speak with the new CEO and understand the company's commitment to your organization going forward. It may be the best way to understand Francisco Partners' longterm strategy for WebTrends.

While there's no way to predict what will happen with WebTrends, I doubt they'll be acquired Omniture.

Are you a WebTrends customer? I'd love to get your thoughts on the current changes. Drop me a line at pkemelor@cmswatch.com

While writing reviews for the new Enterprise Search Report, I found myself frequently saying you should test the effectiveness of a given product against your own corpus of content, which is reiterated in the Report's "Advice" section. But I can't help but wonder how often an actual bake-off between vendors on a shortlist is organized.

I was giving a presentation on the "enterprise search technology landscape" at cmf2007, along with practitioner's cases presented by Brian Schurmann Michels (Novo Nordisk) and Carsten Suhr (DSB, Danish Rail). Both were using search engines to sift through their intranets, but circumstances and solutions were on the opposite ends of the spectrum: Novo Nordisk has a Google Appliance index their relatively structured content (most of which is tagged with keywords), while DSB uses Autonomy to federate disparate content sources (and to allow for their ambitious future plans). This made sense and they both seemed quite content with the technology they were working with in their respective settings.

Both companies were thorough in their implementation. Brian (who has the official job title of "searchmaster") described the detailed comparison they had made of the results of the Google Appliance vs. Microsoft's SharePoint at Novo Nordisk. Carsten talked about the search scenarios and personas DSB had developed: how could the search engine effectively help employees perform their daily tasks? Both were excellent examples of how I would suggest you'd go about selecting and implementing search technology.

Of course, the $100,000 question (putting the estimate at the low end) for me was how they had arrived at the choice for their respective vendors. Why would Novo Nordisk consider trading in their Google Appliance -- which by all accounts they were rather content with -- for SharePoint search? Well, because "SharePoint is being rolled out anyway, whether we want it or not." And why did DSB shell out for Autonomy licenses? Well, because "people in our IT department had experience with Verity's products" (and Verity was acquired by Autonomy).

So, for all their intensive testing and planning -- and DSB did perform a thorough PoC with Autonomy before committing -- the main question remains: what do you do if the only entry in your bake-off wins second prize? When the race is run, it's too late to enter a new horse -- which is why I'll stick to my mantra of investigating the alternatives in the early stages of any search project.

CMS Watch is frequently -- and I think justifiably -- critical of Vignette's "VCM" web content management platform (c.f., my review of their new management interface).

But in fairness, there is one thing that the product does fairly well, though it's not always easy to get a handle on it.

Here's what it is. After a history of poor metadata and taxonomy support, Vignette implemented a real classification system on top of its traditional "channels" motif of binding content items to your site hierarchy. Also, editorial teams can organize content items into arbitrarily-structured "projects" on the back end. Combined, these three ways of organizing content can be quite useful.

First, it allows you to have "placed" and "placeless" content on the same site, or even the same page. That is, sometimes contributors need to actively place content in a particular spot or spots and know it will appear there. In other cases, you want content to appear according to its classification (you tagged a content item "Manitoba", so it will appear in the "Canada" section of your site). With most Web CMS tools, you have to have to choose between placed or placeless approaches, and many customers are quite rightly reluctant to turn their entire site over to a placeless, metadata-driven model where authors and managers alike can never be quite certain where a new page will appear. With Vignette, you can tag content against a hierarchical taxonomy, as well as assign items to particular "channels" which are representations of website locales. In a well thought-through publishing regime, this combination can offer a lot of power.

On the back-end, being able to organize content items against an arbitrary scheme that doesn't actually reflect the published site is also handy. It makes it easier for individuals and teams to work on ad-hoc projects, sorting and storing information in a way that makes sense to them, before it gets published.

Note that Vignette salespeople don't always describe these features very well. "We have three taxonomies," one rep explained rather unhelpfully at a customer pitch I saw recently. To be sure, for a simple site, three organizational models is more of a liability than an asset. But for complex publishing operations, I think this kind of flexibility is quite handy.

In news from Tel Aviv, Israel, a local software company called Vizrt has agreed to acquire Norwegian Web CMS vendor Escenic. Vizrt is focused on the media space, with products for media asset management and broadcast graphics. Judging from a recent company presentation (PDF), Vizrt is active around the world with about USD 62 million revenue for the first 9 months this year, compared with Escenic, which had about USD 6 million revenue in the same period.

While Escenic has been a bit quiet recently, it has been an interesting Web CMS vendor to follow. They are a global player in the media space, but have struggled to move outside this niche, with limited e-government success at home in Norway. Unlike most other vendors who follow a more horizontal approach to the market, Escenic has clearly been engineered and refined for journalists and editors working 8 hours with the system every day. With Vizrt they seem to have found a larger global player who shares the same passion for the media space.

This acquisition comes at the closing of a busy year on the Scandinavian CMS scene. First in Denmark in March Synkron was sold to Dynamicweb and then in August Swedish vendor EPiServer changed owners. Some may say it is not unusual for the Norwegians to follow in the footsteps of their regional neighbors, but as a global vendor with a strong footing in the media space, they may actually be a bit better positioned for international growth. As with any merger or acquisition, buyers should carefully study the impact on the product roadmap, as well as keep close and frequent communication with their vendor contacts.

We have just a few spots remaining for a class I'll be teaching on Information Organization and Access (IOA) in New York from December 10-13, 2007. Please join me for either the 2-day Practitioner Workshop (12/10-12/11) or the full 4-day Master Workshop (12/10-12/14).

The 2-day Practitioner Workshop includes a strategic overview: building a business case, leadership, and information governance. The track then continues with the nuts and bolts of content inventorying, taxonomies, metadata, document modeling, search techniques, and findability.

The 4-day Master Workshop includes the Practitioner sessions, then outlines how to justify and conduct a proper information organization or access project, including requirements gathering and best practices in search implementation. The workshop concludes with an advanced look at security, governance, and case studies.

At the end of the day, technologies like Content Management Systems or a Search tools constitute only a part of what you need to successfully manage content and make it findable. We developed the Information Organization Access course in conjunction with AIIM to address the common issues enterprises face when taking control of their content and making it more findable.

Sign up in November and you will receive a free copy of the book that every good Information Architect keeps by their bedside: O'Reilly's Information Architecture for the World Wide Web -- a.k.a., "The Polar Bear Book," by Lou Rosenfeld and Peter Morville.

Please make plans to join me in New York this December and learn the skills you'll need to optimize findability in your enterprise in 2008...

Hope to see you there!

Microsoft has been making noises about its free web analytics solution since January, 2007. The company finally announced the beta of Gatineau at the end of October.

Focused primarily on those who use Redmond's AdCenter service, Gatineau will draw comparisons with Google Analytics. However, Gatineau offers some interesting differentiation, such as the ability to derive demographic data from site visitors who have signed up for a Live ID through Microsoft's Hotmail or Messenger. Microsoft claims that all this data is made anonymous before being passed to Gatineau. While it's true that visitors who meet this criteria represent a only sample of visitors, the ability to get real market segmentation may be quite useful to those running campaigns or testing content.

Today, all web analytics solutions tout the ability to perform "behavioral" segmentation, but this is merely filtering based on how visitors interact with the site; it's not true segmentation in the marketing sense of the word.

Another feature of interest is the ability to define a custom taxonomy based on your CMS while you're tagging the site, and then see this recreated in Gatineau reports. Although it's not clear at this point what level of effort this will entail, this may address content hierarchy reporting challenges common in other web analytics solutions.

Perhaps the least flashy aspect to Gatineau, but one that may prove to be of high value to users, is that Microsoft includes an online help system with real people actually available to answer questions. Of course, time will tell whether they can handle the volume once the product goes into general availability, but it's an encouraging step.

You can sign up for a beta, but you may have a lot of people in queue ahead of you. If interested, go to: http://advertising.microsoft.com/gatineau

Last week at cmf2007 Raimond Kempees, a developer and consultant at Radagio, gave a nice talk assessing MOSS 2007. Among other things, Raimond relayed some of the challenges he and other developers have experienced with Web Parts. It all reminded me of the love/hate relationship customers endure with portlets in the Java portal world. On the one hand, the Web Parts framework is very attractive for plugging widgets into dashboard-type pages. In addition to what ships natively with SharePoint, you can find lots of 3rd-party Web Parts. But their quality and safety varies substantially, and inevitably you need to extend them...and then how do you handle support? There is also perennial confusion and inconsistency about how much logic to put in presentation tier, and how much to place in objects behind the scenes.

So, Microsoft's portal is no more or less complicated than, say, Sun's.

What was more interesting to me was the crowd's reaction. Some end-user customers nodded their heads, but some Redmond integrator partners seemed apoplectic that someone would criticize MOSS. A typical riposte (which I have heard from MOSS developers around the world) went, in effect, "A .NET developer who really knows how the product works can fix all that!" And what about the lack of a published roadmap and consistent best practices? "If you're a Microsoft partner, you'll be the first in the know!"

SharePoint may be marketed to LAN Administrator, but it is really a boon to consultancies. What you might dislike about MOSS, a consultant might actually favor. One integrator at the session remarked lovingly on a 4-month / 4-developer MOSS customization project they just completed. Of course, many ECM/WCM/Portal projects take far longer with more bodies. But it puts to rest any misapprehensions that MOSS is plug-and-play for anything serious beyond simple collaboration.

I happen believe the world really needs a .NET development platform, but is that what you thought you were getting when you installed MOSS?

I lost my voice last week at cmf2007, and while I've almost fully recovered, seeing 43,000 delegates at Oracle's annual lovefest, Oracle OpenWorld 2007, gave me new appreciation for conference organizing on a massive scale. Today's sessions offered interesting insights into the future Oracle's two portal products, as well as a new related collaboration offering:

• Oracle Portal: Together with the release of Oracle 11g in 1H 2008 quite a bit of back-end improvements are slated for Oracle Portal, including better content integration, support for custom languages, enhanced portlet standards support, and improved management and administration of the portal.
• Oracle WebCenter Suite (WCS): Definitely still seems like Oracle's portal platform for the future. Even the morning keynote by Charles Philips, Oracle's President, showed a detailed and lengthy demo based on WebCenter. While the detailed product roadmap demonstration showed a few bugs and annoying user interface limitations, Oracle is very careful to position WebCenter suite as a composite application platform. In other words WCS is more than an enterprise portal.
• Oracle Beehive: A brand new enterprise collaborative application and platform based on a desktop client for users. This new client will aim to merge the many existing apps used for collaboration (e.g. Outlook, Office, workspaces, instant messaging clients) into one consolidated interface. It is quite ambitiously positioned to replace Outlook and due out in early 2008.

During the Oracle Portal roadmap session a few interesting questions came up. First a delegate asked if the product was about to be discontinued. Later another delegate asked when the product would be replaced by WCS. Both questions were firmly answered by a slightly annoyed Rahul Patel, VP Server Technologies, who clearly stated that Oracle Portal is not going away. According to Oracle the two portal products are intended for different use cases.

While I've been advocating for a while that you may indeed need multiple products, I remain unconvinced that Oracle's strategy is very clear here. I've previously speculated that Oracle is switching portals. However, in the short-term clearly WCS will remain more a developer framework than anything else, and far from a real replacement to Oracle Portal. Even the WebCenter Quick Start carries quite a steep learning curve. Interestingly WCS has wiki functionality, but all you'll find on the Quick Start guide is a link to Wikipedia. Also for portlets you are taken to a Oracle Portal section and the page shows that to understand the tool and the services you need a primer on many different Oracle technologies and products. Finally the product is not in the Oracle Store, which means license info is hard to come by. It feels like a slow ramp-up to me, and product customers should recognize their early adopter status, with all that entails.

In the long-term Oracle may offer a smooth migration from Oracle Portal to OWS, or indeed both products could well survive. Judging from the questions at OpenWorld, Oracle still has some explaining to do. (Thanks to Sten Vesterli from Oracle partner Scott/Tiger for valuable input.)

This past week at the Enterprise Search Summit in San Jose, CA, I had a chance to meet up with some representatives of Mondosoft. In the wake of their bankruptcy announcement and SurfRay gobbling up the pieces, they appear to have emerged relatively unscathed under new SurfRay ownership, even if many customers, no doubt, were rattled.. Indeed SurfRay took the unusual step of formally announcing that there will be no job losses at Mondosoft and that business will "continue to support all Mondosoft customers and that existing contracts may be extended without changes".

It seemed though that this message did not get through to other speakers at ESS who loudly used them as the poster-child for the death of Enterprise Search as a "cottage industry." Well the cottage industry and Mondosoft still have legs.

Has enterprise search finally reached the tipping point? Yes, seemed to be the conclusion of many of the experts here at the Enterprise Search Summit West this past week in San Jose. Of course these are people with a vested interest in believing this to be the case, but there was palpably more energy, commitment, and enthusiasm for enterprise search than in previous years. I chatted with Sue Feldman of IDC and she confirmed my feelings that something has changed, that enterprise search is entering the mainstream, and that there is no turning back.

If that is the case, then it was also clear that enterprise content management is only a few steps behind. The seniority, awareness levels, and scope of the attendees at our KM World ECM track was dramatically different from previous years.

Only time will tell, but the sense here in Silicon Valley is that the long awaited shake out, shake up, and redefinition of search from it's indie niche existence to the mainstream of the infrastructure is well underway.

At the same time, there also seemed to be a new honesty about the limitations of enterprise search and the need to manage content properly. Even Google stated that search is no replacement for properly managed content, and that the accuracy and relevance of search tools is dependent on the quality of the content and structure of content searched.

Microsoft has announced an update to the mouthful "Microsoft Office SharePoint Server for Search 2007" -- which is now called "Search Server 2008", but is still the somewhat stripped-down version of the full search engine included in MOSS 2007. In addition to those two versions, however, Redmond now also offers Search Server 2008 Express -- a download available for free.

So what's the catch? The "Express" version is fully featured, and has no document limit or any of the other usual crippling limitations. Of course, there is a such a thing as a free lunch: you'll have to buy a drink to go with it. In this case, if you want to use more than just one server for your search engine -- and high volume usage might quite quickly warrant scaling your infrastructure -- you'll have to shell out the money for the full version. By that time, you're probably already relying on your search engine and will have no alternative than to pay up. And maybe you'll upgrade to MOSS while you're at it -- or so Microsoft hopes.

So as a trial version, Search Server 2008 Express will be great to get your feet wet -- just don't start getting too cozy with it if you're not planning a budget for when you'll hit the one-server limitation.

So is there anything else in the 2008 versions that warrants the updated yearly version number? Well, I'll let you know after the 220Mb download finishes; in the mean time, I'll go buy myself a drink, without the free lunch.

A Web CMS survey conducted by the University of California at Davis (preliminary results of which were released October 22, 2007) made official what many of us suspected all along: Even the best minds in the world can't agree on how to do content management. Of the 81 (out of 129) respondents who are currently using a Web CMS, roughly 20% (18 respondents) have rolled their own solution. The other 63 institutions are using 39 different branded solutions. The fragmentation of this market is really quite stunning. Can you imagine 63 colleges using 39 different word processor programs?

Quantitatively, the UCD numbers are a bit on the thin side. What survey lacks in statistical significance, however, it makes up for in qualitative poignancy. The comments page makes interesting reading. The usual themes emerge: Know your requirements up front. Get buy-in from all constituencies, not just IT. Budget for training. Expect things to take longer than you thought.

An interesting bit of subtext that seems to weave together many of the survey respondents' comments is that when rolling out a Web CMS, the cost of failure is often better-defined than the cost of success. If a system rolls out within budget and fails due to (say) poor adoption, the cost of the exercise is well known. If, on the other hand, a system rolls out to great fanfare, unanticipated costs can surface.

"Adding a CMS changes how people think about the web," one survey respondent said, pointing out that when authors are able to post content in real- or near-real-time, it gives them a new sense of what's possible. "In the days when you needed a programmer to do just about anything in a site, people settled for getting very little done. Now they may assume they can have a lot for very little cost, so expectations management becomes a real educational endeavor."

One survey respondent said: "Be very conscious that if you do succeed with a vision of centralized management of a decentralized publishing system . . . the infrastructure (server and database) may not be able to handle the load of success, even with a scalable system."

Several of those surveyed warned against buying more functionality than needed, lest users add unexpected administrative (or other) burdens. This corroborates our research suggesting that most Web CMS customers face a greater risk of over-buying than under-buying. As one person said: "Getting the right system to meet your needs is the trick, not getting a great system."

When I'm teaching or writing about content management, one of my mantras is that, to the extent usability is fundamentally "fitness to purpose," then you cannot declare any software tool inherently more usable or "intuitive." In a joint presentation with Janus Boye this morning at cmf2007, we discussed the ubiquitous complexity of the technologies we follow and concluded, "Simplicity takes work." That is, you can't expect it from your vendor, but you have to actively build it into the applications you launch using packaged software. At a keynote after lunch, Stanford's BJ Fogg put it, well, much more simply: he said, "Simplicity lives outside of the product."

Today we released the Enterprise Search Report 2008, evaluating 18 major search vendors. We'll be discussing more about different marketplace trends and vendors in the coming weeks. For now, our initial release focuses on the stunted promise of hosted search, as customers look to appliances for simpler needs, and more sophisticated, on-premise solutions for connecting to enterprise repositories. You can download a free chapter, which includes our (I think rather path-breaking) review of Google's Search Appliance.

If you are a full subscriber, you should receive your copy shortly; if you're a previous report buyer, you'll receive an e-mail a bit later this week outlining discount eligibility.

Web CMS vendor Vignette recently released a new, dashboard-type interface featuring several Ajax controls. CMS Watch founder Tony Byrne demonstrates some of the ups and downs of this new approach, and poses some larger questions about usable interfaces...

Earlier this week Long Island, NY-based Web CMS vendor FatWire announced that it has acquired Infostoria, a Long Island-based wiki vendor.

Also on Long Island is Computer Associates, and while Inforstoria was only a tiny start-up with 12 employees, it was founded by CA alumni. Among the CA alumni is also FatWire's relatively new CEO Yogesh Gupta, who came onboard in August.

The last time FatWire completed an acquisition, it was from bankrupt vendor divine, where FatWire bought its current flagship product Content Server. It seems unlikely that this acquisition will have the same impact.

Nevertheless, this news would confirm that wikis are very hot these days, with the major MOSS 2007 wiki integration news being only a week old. Where most Web CMS vendors (e.g. Microsoft, Stellent) have built their own wiki functionality, FatWire may have taken a short-cut with this acquisition. To me though it seems rather hard to fit a wiki into FatWire's traditional emphasis on persuasive content and object-oriented content management, but perhaps more change is coming for FatWire?

Well that was the news out of Portland late yesterday from a piece in the Portland Business Journal. Also gone are Jason Palmer, vice president of product management; Tore Steen, vice president of business and corporate development; and Hamid Bahadori, vice president of product development and hosted operations.

There's not a lot of details coming out just at this time. However, if you have anything to share, please shoot me an email.

WebTrends has seemingly been on the come back trail with features such as Score and Visitor Intelligence. On the other hand, their WebTrends Marketing Warehouse has been slow to catch on with customers, and it's probable that their highly publicized effort to scoop up HBX customers hasn't paid off.

The new CEO is Bruce T. Coleman. Coleman is CEO of El Salto Advisors, a consulting firm that provides interim management to computer software and service companies.

Some former employees have speculated that Francisco Partners, the private equity firm that owns WebTrends is eager to unload the firm.

From an investor's perspective, this would seem to make sense. Web analytics is hot right now. But from a customers' perspective, quick sales don't often bring good tidings.

The offer from Oracle to acquire BEA expired this weekend, as BEA refused to meet the proffered terms and countered with a higher price.

This means that BEA remains independent, and, at least for now, still moves forward on important matters such as product roadmaps. In a series of blog postings from Josh Lannin, senior product manager, you can read interesting details on the future for WebLogic Portal, including more on content publishing, standards, single sign-on, customization, and much more. The roadmap extends to late 2008.

As a licensee, it is always quite helpful to know the roadmap for the portal product, given the somewhat turbulent landscape. Historically upgrades have been time-consuming and difficult. With proper roadmap information you can at least prepare and plan ahead. This open style of communication from BEA is quite welcome. The next few months will tell whether BEA will continue on its own or whether new owners may come in and change the course.

Our latest web content management technology research suggests that Big Blue is falling behind its major competitors in the Web CMS marketplace. To quote: IBM's Workplace Web Content Management (WWCM) product remains a generation behind, and it is conceivable that IBM may simply elect to acquire a WCM vendor to fill the gap. CMS Watch cautions that IBM WWCM customers may risk the kind of painful upgrade or even replacement of the kind endured by Microsoft Content Management Server (MCMS) licensees when Redmond rewrote that tool under SharePoint.

Read more here.