CMS Horror Stories, and Your Soon-To-Be “Legacy” Community Platform?

In the late 1990s the CMS invaders deployed their systems at large corporations, as managing web pages using HTML editors wasn’t scalable and non-technical folks needed to publish. In many cases after the invader left, the company’s business teams and technical web teams were stuck cleaning, fixing, enhancing, for years to come.

Unplugging web publishing systems (and community platforms) ain’t easy.

Publishing from Word Docs, ouch.
I was a web manager at a very large corporations, as such, I was the business sponsor for the website, and therefore the tools that were used to publish the website. Often, in most cases, I inherited a legacy CMS system, one that I did not choose, the underpinning structure of the site revolved around it, documents, navigation, ability to edit pages, and look and feel.

This was one of the worst implementations of CMS systems I’d ever seen, the idea was for non-technical people to edit the webpages, so the system would have the ability to check out a ‘word doc template’ filled with macros, publishers could edit the word doc, check it back into the system and a new webpage would appear. fail.

The templates were so complicated as users had to be trained on how to use the word docs, understand the styles, and all the nuances associated with the code. The linking structure linked to a primary key for a document, which also caused confusion. That’s just the publishing process, it gets worse.

The Pains of Content and Structure Coupled
The site was unfortunately designed so the structure would for the most part, remain constant. The structure of the site, and the content were coupled together, and that’s a major problem. As the site would grow and more pages were added to the taxonomy, the system became more and more inflexible. The developers had a very complicated way of managing the pages, the changes took a few days to work as the underlying code had to be changed. The simplest of web changes that you would expected to see from a web CMS system required ongoing developer support –not content changes at the business level.

I’m not going to mention the name of the CMS vendor who provided this less than stellar tool, as I believe the deployment of the system was to blame from the in house technical group –all of which happened before I got there. Whew, I feel better, that’s been pent up inside of me for a few years now.

Thinking forward: Community Systems of today, to be legacy systems tomorrow
As we deploy community solutions that have social media features, are we thinking about in a few years how these legacy systems will be inflexible, don’t talk to our other systems, cobbled together application ware that we loosely couple with our other customer facing web systems?

I also know of many business groups that are deploying community software, often by ‘notifying’ IT that they are doing it, sometimes without thinking about the long term implications of these systems not being able to migrate, talk, or share data with other websites. In many cases, the business sponsor will move on to another role, job, or company, leaving the archaic community platform in the hands of the next web strategist.

Two questions for you:

1) I’d love to hear from you about your CMS horror stories, feel free to leave a comment below, go ahead, vent away.

2) Are you deploying a community platform for your web strategy at your company? What are you doing to plan for the long term 5+ years impacts of this system in regards to the rest of the enterprise web strategy?

  • The issue I often have is the massive complexity of many CMS’s, which may be fine for large companies with a staff of IT geekazoids, but the small companies or solo entrepreneurs are often completely befuddled by the complex and/or bloated feature set. A CMS should follow the KISS principle as much as possible.

  • Ken

    Jeremiah…great post on the horror that you encountered with a CMS. I’ve been “fortunate” to only be exposed to custom-built CMS’es. The first few have been pretty good, but the one I “inherited” at a job was horrible. There were 14 steps JUST to add an image to a story. The linking feature was appalling – when you created a hyperlink, the damn thing removed the space before the word so you had textLINK in that format. I was tired of even working on updating the site…when press releases came about, I formatted them in Dreamweaver and copied and pasted the code into the WYSIWYG editor…and THAT didn’t even look right. The workflow and accessibility of this CMS was horrible.

  • I’m saddled with a “custom” CMS that I inherited — the only thing custom about it is, I think, it was the cheapest thing available in 1998. Currently kicking it to the curb with a total site redesign and change of vendor. Updating my website takes three times as long as it should because of how often I have to go back in and fix macros.

    It amazes me that CMS can be SO BAD.

  • Worse than that. I have completed competitive audits for more than 20 companies this year. Almost without exception, they have CSS issues and I am not talking about small organisations.

    There is an issue here which is that companies, even the big ecommerce sites are still not investing enough management time or cash into their online presence.

    Many then compound this with requests to get involved in social media without a strategy.

    It goes far beyond CSS….. Rant over…

  • And, at a higher level than the sheer awfulness of using particular custom CMS, don’t think that open source is necessarily a solution, either.

    A few years ago, we set up our own site and several others on the Mambo open source CMS, after reviewing the alternatives, Drupal et al.

    It was OK – at least better than hacking individual pages in Dreamweaver.

    And then it forked. Some company that owned the rights to the name Mambo reasserted those rights, supposedly on the grounds of improving the product.

    The development community took it personally and collectively went off to start the Mambo-like CMS, Joomla.

    Which way to go?

    Long story short, after switching to another (inflexible) CMS for a year, we simply integrated the site with our existing Typepad blog.

    The blog now *is* the site and that works very well for us.

  • My company specialises in deploying content management systems for both intranets and customer-facing websites. We’ve selected the open source Drupal CMS for the majority of projects. This is because we’ve found that if you deploy a modular system that is relatively well-known, keeping track of business and technical changes becomes easier to manage.

    Some of the Drupal sites we’ve encountered have been running for years. Over time, there have sometimes been changes to both the original business sponsor, as well as the internal maintenance team. Nevertheless, we’re still able to keep the sites up-to-date with the latest trends. Of course, it’s not always painless but the task is never daunting.

    Going forward, we plan to continue using Drupal. It has a solid and flexible platform with a growing developer base. This means that even if we’re out of the picture, our clients can continue to maintain the system in-house or through contractors. Should any rifts form within the Drupal community (as with Mambo), I believe that most experienced PHP developers are able to continue building on the exiting installation, if necessary.

  • We will be changing our website in the next year or so, and this post and its comments will give me a lot to talk about with the new developers.

    Our CMS is proprietary, and we have kind of the opposite problem — it’s so simplistic that it can’t easily handle a lot of changes that we would like (adding video to podcast). Changes require reposting of not just the page but every link, photo, video, pod, etc. Our current vendor has said they aren’t interested in doing the upgrades because they would have to start over with a new platform.

    Just waiting for the budget so I can make a change. Then we will do something that is social-media ready.

  • Our company had inherited a few wonky CMS sites over the years (both custom built and open source). So we decided to build our own from the ground up and use it for all of our websites (in house and client sites – approx 50). We are proud to have kept the feature set small (and easy to understand) and also make it very user friendly for non-technical people. The low startup cost is very important also for small businesses to get online easily and fast. I don’t usually plug products in blog comments but I think it’s a good fit to mention it here since people are discussing their horror stories regarding their CMS Systems and maybe it can help them in the future.

    It’s very popular with the SMB market and Entrepreneurs as well: http://www.bigsanto.com

    Best of luck with your CMS!

  • What CMS? You mean Notepad, vi, emacs, etc. with files managed in unix directories? 🙂

    @1ndus

  • Janie

    Glad it will give you team something to discuss.

  • Pingback: Blogging - opening the door to sales « Sweet!()

  • Hi Jeremiah,

    This is a very timely question. Deploying on a platforms or CMS almost always a Faustian bargain. You get some live quickly but then you are imprisoned in it for life. Your production schedule seems to become bound to the release (and delay) schedules of the CMS maker.

    At Dogster & Catster we take advantage of frameworks, but very contientiously built all interactive functionality. We could have gotten Forums, Diaries, Groups, Messaging, Groups pre-made, but then we’d be trapped within them.

    Much of the white label community services offered remind me of the ‘portal-in-a-box’ software popular in the late ’90s. How can you tailor your community experience as tightly as it needs to be when you you can’t tailor-make your community site? Even worse, if you’re community is outgrowing the hosted platform you started it on, how to do you export all your users to your custom-made platform when the ‘users’ are actually the users of the hosted service, not your community?

    And I have to agree with some smart commenters above. Choosing a hugely popular open source or community-driven CMS such as Drupal or even WordPress you are best ensuring, but not guaranteeing, that in 5 years the core software will be as fastidiously developed and enhanced as it is now.

    I’m rambling on too long, but of course it has to be noted that building your own system can just as easily translate to legacy code for future company members as using a pre-made CMS. The most important thing is a long-term commitment to the system, not just the functionality it provides. If you don’t work on the code base as much as the end-product you’ll inevitably have a legacy issue.

  • I’ve been using WordPress (2.5) for several small sites I have developed (mostly for friends) recently.

    I’m really starting to love WP as a CMS.

    I am interested in using it at work (I do online marketing for an ad agency).

    I would love to know if anyone has used WP for 100+ page sites and what sort of problems people see or forsee.

    Thanks,
    Josh

  • Ah…the CMS discussion. Finally. We’ve been using Interwoven Teamsite for the couple sites I help manage at work and its a total mess. It was built by developers, for developers – it is in no way easy or useful to the common, MS Word using common business user – which I thought was the target (or should have been) audience of these things, that is, if the Intewoven folks had any foresight and realized that the governance for most sights would move out of IT and into the hands of business folks. Its like juggling bricks when you have a set of those soft little satin pillows at your feet – you know there’s a more elegant, more comfortable, more pleasing way to do what you’re doing but you’re stuck with the damn bricks.

  • That’s an interesting topic Jeremiah, I once worked with a team that developed an in-house custom CMS. On paper it looked like a dream, modular, bespoke, anything you wanted it to be.

    In reality, customers paid through the nose for developer’s learning as they struggled to implement multiple sites that were not quite similar enough to be cost effective (and took twice as long to implement).

    It has to be said that the open-source solutions provide such a great alternative (if you can convince the corporate IT team and execs).

    Just knowing that it is not a team of ten or twenty, but thousands of smart individuals that are contributing to the constant evolution of plug-ins etc.

    Sure some will argue about the ‘robustness’ of .net et al. or say that php is ‘slow’, but at the end of the day you really can do just about anything with a core engine from one of the OS solutions.

    As with everything, it’s not the tools but the content strategy that defines the engine. And ye olde ‘one-size-fits-all’ mentality needs to be thrown onto the pile with all those legacy CMS systems that over-promised and under-delivered.

  • Jeff Lippold

    I agree with the majority of the comments above, in particular the point about CMS’ being largely a Faustian bargain.
    I’m based in Japan, and in addition to the points above, I think the biggest problem with the implication of CMS systems here is that they’re implemented as the solution to give internal types control, well before they know what they’re going to do with the site. Ultimately, most CMS systems fix designs in a manner that doesn’t allow for robustness and scalability (plus a lot of the sites look really bad), resulting in systems that simply become glorified (and expensive) news update mechanisms.
    We’ve had some luck in building some AJAX based custom CMS systems for sites we manage internally (with experienced developers, editors and programmers), but even they require constant enhancements in the face of the changing nature of social computing. After all, if we could see the future clearly we wouldn’t be congregating here, would we?
    All in all a good topic. If at the end of this you come across some examples of good robust systems and cases that have worked (especially in a global setting), that would be great.

  • Pingback: Pick A Winner: Your CMS Selection — Archive — RD2 Blog()

  • It’s very clear, after 24 hours, I’m not the only one with pain in this topic.

  • I agree that was a horrible implementation of CMS we had there. One of the fatal flaws of that implementation was an assumption that the site taxonomy was relatively stable when the reality was the exact opposite. Like clockwork, every 6 months the GSS folks would completely turn the site structure upside down. I used to joke with other developers that the users must have found the documents so they were hiding them again.

    At some point you have to compromise when the CMS serves multiple audiences and integration points. It’s no wonder why most intranets are department silos of content–getting people to agree on a format and subsequent alterations is not for the weary. Nevertheless, I’m still amazed when people architect solutions that don’t accommodate, let alone anticipate, redesigns. Change is the only thing you can count on and yet so many solutions overlook this simple fact.

    Ideally the folks who organize the content should be able to do so directly via the tool and it shouldn’t be complicated, but the implementation there was intended to be a grand unifying CMS that served all 3 nets (inter/intra/extra). It never got there and frustrated everyone along the way, eventually the developers were updating web pages by editing the word docs because no user could figure it out. There was a follow up project where we had to update all the templates for all the existing documents and we were told to enter the project name followed by our initials in the comments field as an audit trail. I had to resist the temptation to take that literally as it would have read, “CMS Phase 2 BS”.

    -Brian S

  • Brian S

    If you ever want to guest post on CMS systems for websites, let me know.

    Thanks for sharing.

  • ian

    The worst I’ve ever seen was a client who had a 5-year-old Microsoft CMS implementation that pulled in Word docs but, in conversion, added about 500kb of ‘mso:normal’ garbage to every post.

    My second-favorite: The custom CMS that didn’t allow addition of new pages.

    It continues to amaze me how few organizations consider a CMS a strategic decision. When we’re evaluating one for a client we’ll review audience, publishing schedules and plans for the next year before we even draw up a short list of systems.

  • AJ

    With regards to deploying a community platform, I’m going open source. As communities become increasingly important as a web presence, I like the opportunity open source CMS’s offer you to extend you functionality past the blog, message board and networking features that commonly define online communities. I haven’t found that white labels are there yet.
    Additionally, the community of developers ongoing participation adds services and seamless authentication options that allow you to scale and grow your community out towards the 5 year range. Open source CMS’s / community platforms seem to really harness the collective wisdom with regards to smartly designed content management. Bypassing the narrowband approach of some big co’s.

  • liz

    Thanks Jeremiah and all the thoughtful commentators.

    This post is very very timely for me.

    I’m developing a social entrepreneur project for non profits in Ireland called SKIL2.

    Basically I want to create online as well as face to face learning spaces that capture the informal, messy, tacit knowledge in the non profit workplace.

    My non profits are small to medium and don’t have the budget, inclination or time to develop learning management systems.

    Me? I love a challenge so at the moment I’m researching the kind of platform (it will be open source) I want to use to pull down the social media cloud I’ll use to create learning spaces.

    The open source CMS LMS options are not endless but they are interesting – wordpress, ning, drupal, moodle, sakai, elgg … and I know of about a dozen others.

    Over the next few months I’m going to research and test all of the above and a few more.

    Wish me luck.

    I’ll be posting my explorations at my SKIL2 social network – you’re all very welcome to visit and please add any support/advice/danger stories about any of the options I mentioned above.

    Ultimately I want to create this social media cloud space so that non profit workers can communicate, collaborate and rattle the cage of their work practice – and then – provide the best services to the communities they serve.

    Just think – even 3 to 5 years ago we wouldn’t have had the open source options or social media tools to do this for non profits.Big cheers to all the IT geeks who support open source.

    Thanks Jeremiah for your always thoughtful posts.

    best regards
    liz

  • I’m the IT Director at a medium-sized non-profit. At the moment, we’re stuck with three different legacy CMS systems- Savvy for our main website, Kintera for our fundraising and event management functions and MS Sharepoint for our Intranet. As you can imagine, this does not lead to fun times. We’ll be looking to consolidate over the next year or so. We’re an MS shop, so will be looking for solutions down that path.

    On the community front, we’ve got a lot of pent up demand to launch communities and I’m spending a lot of time putting the brakes on rogue projects while we develop a deployment strategy and identify a target for an initial iteration that will stand the test of time/be integrated with our core systems.

    It’s a delicate balance with a limited window of opportunity before people start grumbling about IT roadblocking key initiatives. Fortunately, most of the other senior staff seem willing to cut me some slack while we get our ducks in a row.

    Jeremiah, I have to reiterate that the quality of your work is tremendous. Yours is by far the most insightful blog that I follow on the web strategy and community front. Please keep it up!

  • Thanks Bill, like you, it’s my passion too, keep up the good work from the front lines!

  • Be great to know what the Top 3 CMS platforms were, as voted for by front line web editors, not by sales.

    Do you know of such a list? Let me know on my blog if you do!

    Cheers guys / gals.

  • Tom

    Great topic! I’ve seen incredibly successful CMS implementations and on occasion, epic failures. Full disclosure: I work for a commercial CMS vendor (Interwoven).

    Sometimes the software itself is the blame. All CMS vendors, including Interwoven, need to become easier to install/administer and easier to use. Although WordPress has some pretty significant limitations as an enterprise CMS, it does illustrate how easy a CMS should be to implement and use. Usability is critical as the business user of a CMS is increasingly someone in Web Marketing.

    Many of the poor CMS implementations I’ve seen have been caused by scope creep during the implementation process. CMS platforms are incredibly powerful and therefore dangerous in the wrong hands. Countless times I’ve warned customers against creating the 30 step workflow approval process or creating 100 unique layout templates for the 100 pages on the site yet the customer ultimately has the final say.

    WebChicklet says it best “A CMS should follow the KISS principle as much as possible.” The applies to the CMS platform itself and the implementation of the platform. CMS vendors should strive to minimize complexity and continue to drive towards true business user usability. Companies implementing a CMS should spend plenty of time upfront understanding the real business requirements for the CMS and err on the side of simplicity. I generally recommend starting small with a quick win; maybe a landing page or microsite. This helps flush out some of the larger issues before tackling the full site re-design.

  • Pingback: CMS - open/closed/alphabet soup « Tom Altman’s Wedia Conversation()

  • I work for a commercial CMS vendor as well. I have tried out quite a few opensource products, and most of the time they are too bloat(eg. Joomla) or too complex. I think the biggest problem is that CMS is usually very general product, but the clients are needs are quite specialized.

    In our product we have tried to keep things simple, but allow flexibility by offering components(like formatted text, navigation, html, forms, databases), actual databases and programming like ifs, querys and variables.

    What we do is that we work with the clients from the start, designing the site structure, layouts, create templates, program components or features, templates and place the initial content on the new site. That way we know what the client needs and we can guide the client to be more “reasonable”(you cannot get facebook+youtube+delicious on a CMS). We host the sites on our on servers so there isn’t any installing or administrating on the client side.

    Point is that there probably isn’t a perfect solution, you just need to find good solutions that work for the client. Highly automated or easy CMS hurts the customizing, and highly customizable or feature rich hurts the simplicity.

  • Dave

    Also very timely for me. I’ve just started a new job, managing a website that at present is 7500+ .asp files stored in a shared drive. The number of PDFs on the site is in five figures. There’s no CMS and whilst the site holds its shape, it’s not sustainable.

    There’s no doubt we could do a better job of matching content with customers, and a CMS is a potential part of that solution. I’ve only every worked with bespoke solutions (mixed success), with all this bad blood around about CMSs, the inflexibility, the difficulty in specifying a solution that’ll work in the mid-to-long-term etc etc I find myself paralysed and unable to take action confidently.

    I’d be interested in some of the open questions you think I should be asking of myself and the team that may help determine what type of approach to content management is suitable for my organisation (a non-US Govt department). Thanks.

  • I had a legacy CMS one time that would wipe out all your information if more than one person signed in at a time. The maker of the CMS conveniently forgot to fill is in on that fact. So the first time that two people were signed in the entire website went down!

  • Pingback: Excellent Web resource on Web strategy | novatownhall blog()

  • It seems like there’s a consulting opportunity to help Dave here.

  • All old CMS’s are notoriously crap. I’m not ashamed to say some of our old custom CMS installs were less than perfect.

    I’m looking forward to the day that someone takes the time to build an open source CMS that has as much care taken in the interface to the CMS being as usable as the customer facing side.

    We’re currently using Joomla and the interface is quite confusing to the less tech savvy members of the team. WTF is a mambot anyway 😛

  • Jeffrey Davis

    This is quite a timely discussion. The CMS landscape is quite crowded and determining what will be the best for your organization is a challenge. On that thought (re: the post by kenobi), anyone have any feedback regarding the main Enterprise class offerings out there (Documentum, IBM Workplace, TeamSite, Oracle UCM, RedDot, Vignette, etc.)?

    I am investigating what could serve as a true robust foundational layer for an extremely large, complicated and distributed web presence. It needs to handle pretty much anything we could throw at it… 🙂

    Thanks,

    Jeff

  • Approaching social software challenges as CMS challenges is very superficial and ultimately, if adopted, will lead many organizations to spend a great deal of limited funds without achieving intended results. {Though your analysis of this trend is spot on] Here is a thought process on a more appropriate enduring business model. http://tinyurl.com/68tfbe

  • Pingback: Modular Buildings()

  • Pingback: Joining Dots | Links last week – 080511()