CMS Marketplace
A Look at the CMS Marketplace
by Tony Byrne
24-Jun-2004
Having just completed version 6 of The CMS Report, it's time once again to take a look at the Web content management marketplace as a whole.
Despite some recent turbulence at the enterprise level, the Web content management software industry still supports a plethora of healthy, growing players, in North America, Europe, as well as Asia-Pacific. The good news is that buyers have more solid choices than ever; the bad news is that distinguishing among the different offerings remains as challenging as ever. Still, there are major differences among CMS vendors when you dig deeply.
Here is a summary of just a few of the trends we identified in the latest version of the Report.
Rising Tide: lifting all vendors
Despite frequent analyst judgments that Web content management is passé, we have never seen the marketplace more vibrant. In the face of constant predictions that the vendor community is due for consolidation, new players continue to join the scene while existing ones routinely press forward and -- in many cases -- truly thrive. Sure, EMC bought Documentum, but the latter's WCM software didn't go away. Products are not disappearing.
Vendors report a substantial number of first-time CMS implementations, largely in the mid-market, but also among major enterprises. Many enterprises still see multiple installations using different products internally. Whatever heartburn this causes the CIO, it reflects, we think, the diversity of use cases within a distributed enterprise, as well as the substantial differences among vendors once you get beyond basic capabilities.
We also see a lot of system replacement going on out there. We're working with a firm that's about to embark on its 3rd CMS. The first one was a failure. The second one worked reasonably well for 3 years, but has run its course. The company is now putting out feelers for a 3rd try in lieu of upgrading automatically with the incumbent. They know what they want, and are going to press bidders for just the right solution.
Over the past couple of years, we have also catalogued the rise of "regional players." The CMS Report briefly examines about 20 of them -- but there are dozens more. These are usually 8- to 15-person companies that have a regional (North America) or national (Europe and Asia-Pac) focus. They often boast more contemporary technical architectures, and can provide more intimate care and feeding of their clients than major vendors. Nationally-focused vendors are doing particularly well in Europe, where the expansion plans of many traditional CMS players have been dashed on the rocks of stubbornly national marketplaces.
We should also note that the open-source community remains fertile -- if somewhat fractured itself. New open-source CMS packages come out monthly. We're not sure this is a good thing for potential buyers, but is still a reflection of the market's vibrancy.
So on the whole, vendors are smiling. But there's one potential pitfall: vendors typically assume that their rising sales reflect competitive prowess, and therefore risk becoming complacent.
Software Licensing: all over the map
A couple of years back, it appeared that CMS software licensing was converging around per-CPU charges. Not anymore.
2004 has seen substantial experimentation on licensing, with vendors in some cases tinkering with existing approaches, and in other cases, coming up with wholly new models as they deliver major software releases. (This has led to some difficult conversations with their installed base, but that's another story.) A brief tour of pricing models would include:
- per CPU
- per server
- per domain
- per contributor
- per power contributor
- per content item,
- per developer
- per year, and, frequently...
- ...some combination of the above
Yeesh. You can eventually get to an apples-to-apples comparison of pricing among your vendor short list, but you will have to work hard at it (our report has some pointers).
A steady return to per-user pricing reflects in part a return to thicker clients -- more about that below -- which, among other things, require ongoing support. In other cases, vendors are looking to capture incremental value from buyers who start modestly (almost always a good idea), but then expand usage over time.
Usability: beginning to surface
In the latest version of The CMS Report, we added "usability" as an evaluation category. Vendors often report little interest in usability among prospects and buyers beyond the basic authoring interface. We don't believe that. The problem remains that end-users are too often shut out of buying decisions.
To be fair, usability can be highly subjective and situational. Vendors point out that casual contributors and power users often bring different needs and perspectives. And of course there is more to usability than interfaces. We've come to focus on some key attributes as representative of the usability of a product as a whole:
- the availability, quality, and customizability of any help subsystem;
- the intuitiveness of the work queue or task inbox;
- the availability and quality of real management reports;
- the richness of the CMS search capabilities;
- the customizability of the user interface as a whole.
This isn't an exhaustive list, but a starting point to considering a vendor's orientation. For example, Interwoven Teamsite 6.x, whatever its weaknesses, has a very nice approach to customizing user interfaces. PaperThin's CommonSpot 4.0, while attractive in many respects, does not provide a help system. Serena Collage sports a very handy reporting subsystem. RedDot, an otherwise quite capable product, still pops too many new windows. And so on.
Metadata: still a weak point
Most CMS vendors still provide only very basic abilities to apply metadata. They tend to argue that buyers talk more about metadata than actually implement it. That may be true, but perhaps it's because the CMS doesn't provide enough capabilities to offer a sufficient return on all the extra investment that adding attributes and categories requires.
To be sure, there has been some progress. Even with low-end packages, you can now add custom metadata fields. The larger issues revolve around whether and how you can manage controlled vocabularies, and then leverage those attributes for linking, navigation, and search. Even some sophisticated CMS products cannot support a multifaceted categorization scheme like the one in the screen below (from CrownPeak, showing one tree from among 3 facets).
Thicker Clients: they're back...
After years of struggling with the limitations of browser-based interfaces, many CMS vendors are returning to thick, or at least "thicker" clients. By this we mean more than just a rich text editor for WYSIWYG editing. Rather, the entire screen is generated by a Java applet, ActiveX control, or in some cases, a desktop client application. Here are just a few examples:
- Mediasurface recently developed a slick Windows client called Morello (see screen below);
- Documentum Web Publisher bundles a (rather clumsy) Java applet-based authoring interface;
- Percussion uses an applet to generate its main control panel;
- Ingeniux ships with ActiveX-based and native OS X clients.
We tend to distrust thick clients: they can be difficult to customize and lock you in more tightly to the vendor. Current web browsers are not ideal, but many interface developers are doing great things with DHTML and JavaScript, although that approach can be buggy on some platforms. So it's your call.
One area where thick clients do make sense is on the Macintosh, where browser support is getting increasingly difficult. Vendors are beginning to take advantage of some advanced GUI capabilities in OS X to develop highly functional interfaces for what is often a very influential community of users.
Hosted Vendors: leading innovation
If you're a (responsible) software vendor, you have to worry about regression-testing new releases against all your customers' crazy combinations, flavors, and configurations of operating systems and application servers. This can be a significant drag on productivity in a small or mid-sized vendor -- and none of the CMS players are particularly large.
Freed from this burden, hosted or "ASP" CMS vendors seem to be able to innovate faster. In the CMS Report we profile Atomz, Crownpeak, and Clickability (whose screen, below, shows the ability to manually order article headings on an index page), but there are several others. Not only do hosted vendors appear to roll out new features faster, they can also propagate them immediately to their entire customer base.
To be sure, a hosted solution is not for everyone. It's particularly problematic if you have a significant amount of integration with other internal systems. But in an environment where nearly all the CMS vendors are doing well, the ASP vendors seem to be truly thriving.
Better RFPs: describe and test
We still see a lot of execrable RFPs and tenders out there (a subject for a future article). But on the whole, they are getting better. Buyers are beginning to anticipate more what their system implementation and usage will look like, and are coming up with more measurable criteria to judge vendors.
In particular, we see a healthy trend towards eschewing check-box RFPs in favor of a more intelligent examination of how vendors implement a particular function. Smart buyers are also including more narrative descriptions -- call them use cases or scenarios -- to be very clear about what they need in terms of business improvement, as well as to be able to compare vendors with specific tests. The CMS Report describes this phenomenon in more detail and includes a sample scenario.
Another trend is growing more slowly than we'd like: actually testing packages head-to-head in competitive "bake-offs." Admittedly, this can be hard to arrange:
- most packages are complex enough that you need vendors to do the actual prototyping;
- some vendors are reluctant to do this;
- it takes time and effort on behalf of your internal team to participate.
But the results are usually worth it. In a real bake-off, the team (you have a selection team, right?) can preview the merits and blemishes of the winning system. But perhaps more importantly, the competition results may surprise you. They often surprise us.
In any case...good luck!