Thank you for this piece, more and more people are joining the daunted hunt for the right CMS daily.
@gsal: While I have to agree that technology can be no “silver bullet”, there are certain companies which have recently rolled out the red carpet for content management systems that have cutting-edge curation analytics capabilities. I am a tech comm and my company just adopted MindTouch 2010 and it’s already given us helpful feedback on our online documentation. MindTouch offers a free trial system so you can walk away with no strings attached if you believe that your content issues aren’t “technology-deep”. But MindTouch might just give you the feedback you need on really whatever content issues you’ve been harboring. Here’s the article on Mindtouch 2010’s curation analytics that won us over: http://www.cmswire.com/cms/enterprise-20/mindtouch-2010-includes-curation-analytics-for-content-and-documentation-008045.php
This was a great article and I think most of us agree with all of your points. The ultimate challenge is getting the stakeholders to invest both time and money into content strategy rather than just signing on the dotted line for a tool.
I have learned that you really have to be blunt in explaining the important of developing content first. I have actually told a friend of mine, “your site sucks because your content sucks” and he quickly understood. That’s more difficult to tell a client who is paying thousands of dollars for your services.
It’s nice to see some more focus on content modeling, and the value of creating a semantic structure for the content in a CMS. As a vendor, we’ve had semantic content and content modeling technologies as one of our most important areas of R&D for several years now(http://www.webnodes.com/ontology-engine), as we see it as one of the most important factors in creating good websites, both for visitors and editors.
However, I do not agree with the presumption that:
“Any web project more complex than a blog requires custom CMS design work.”
Of course, we might have different interpretations about what constitutes custom CMS design work. What do you define as “custom CMS design work”? Could you give an example? To me customizations are changes to the CMS that are outside the scope of the intended built-in flexibility of a CMS.
Such customizations should only be done as a last resortIMO. Most likely, you have selected the wrong CMS, if it isn’t possible to solve your problems within the built-in flexibility.
I’ve seen so many customizations over the years done to different content management systems and other IT systems that have turned out to be extremly costly over the lifetime of the solution. This has mainly been due to upgrade problems, and a limited number of people that knows how the customizations are set up.
The ultimate challenge is getting the stakeholders to invest both time and money into content strategy rather than just signing on the dotted line for a tool.
What do you define as “custom CMS design work”? Could
you give an example? To me customizations are changes to the CMS that are outside the scope of the intended built-in flexibility of a CMS.
I go into a bit more depth later on, starting with, “A product can’t fix content problems out of the box.” Whether or not customizations are within the scope intended by a software vendor, web teams need to spend time customizing tools to their needs. Software vendors tend to talk about “solutions”, which is presumptuous: solutions to which problem?
The “intended built-in flexibility” of proprietary tools is beyond the scope of this article. If you approach a software vendor/tech team/open source community with a clear idea of your content model and task flows, they should be able to estimate how much time and money is required to customize the tools for the project’s needs.
If you approach a software vendor/tech team/open source community with a clear idea of your content model and task flows, they should be able to estimate how much time and money is required to customize the tools for the project’s needs.
Sure, they should be able to estimate how much time it would take to customize the tools for the project’s needs, but there are several potential pitfalls. Yes, you can hack most CMS tools to support a given data model, but there’s a BIG difference between supporting a data model within the “intended built-in flexibility” of the data model, and hacking together support in a system never intended for such a data model.
How much will it cost to upgrade when the next version arrives? How well will it work? How polished will the customization be compared to the built-in functionality? How well will it scale with lots of traffic?
Thought provoking article, but the choices we really have are way more limited than you’d suggest.
Most web development firms are working with just three of four different platforms, representing different bands of the cheap ‘n’ simple, to enterprise-level spectrum. The choice will always be between your favourite 3 (or so) CMS because:
1) You can’t be really good at more than 3 or 4 systems if you’re undertaking lots of custom implementations
2) Supporting too many platforms as a supplier introduces risk and sustainability: how do you retain the knowledge of a system if you have any staff turnover?
3) If you choose a platform on project fit alone, and it isn’t something you know well, you won’t be building something that meets all the best practices for that tool – these essential nuances to a build are developed over time
When choosing a CMS the real question is, which of my (3 or 4) favourite systems is a good fit. And failing that, should I build the required features in a custom application?
As the article points out content modelling is essential in understanding this.
Buts let’s keep it real, there are never 1000s of options on the table, but a maximum of 5. Maybe a few more if you include low cost, hosted solutions like Ning, Shopify etc.
I heard a talk about Expression Engine not too long ago and it seemed to be highly versatile. I’ve been using Wordpress since its conception many years back and have watched it grow from a little blog tool to one of the most powerful free CMS’ out there.
I am doing an hour long talk about history of Wordpress and where 3.0 is at today, MU, Buddypress, etc. I feel with the hundreds of thousands of plugins and themes already available it is hard to promote anything else.
Thanks for a great article and discussion piece! It’s refreshing to see this perspective clearly articulated. I work for a custom web dev and design shop who, for the last 12 years, have focused on always building custom solutions for our clients. It’s worked well because, before anything, we work with our clients to determine their needs and present our plan for a solution before we even lay down any code. No crowbars to hack up other, pre-fab CMS’s but instead a unique product created just for them.
Anyone can buy a frozen dinner and try to doctor it up with some salt and pepper but wouldn’t you rather have a chef cook to your own tastes?
When choosing a CMS the real question is, which of my (3 or 4) favourite systems is a good fit. And failing that, should I build the required features in a custom application?
I think your logic is backwards. Of course I’m not suggesting we should survey all of the thousands of tools out there—that would be impossible. But focusing on “building something that meets all the best practices for that tool” sounds to me like a classic distraction from the content strategy question. It’s an excuse to stick with tools you know rather than involving yourself in a strategic process. The tool isn’t the thing: the content is.
Copy & paste the code below to embed this comment.
Eric Sol
The focus on content in your article is much appreciated, as well as your basic outline of content modelling (I was trained as an information analyst and produced many diagrams and flowcharts). However, on the web things change much much faster than in most administrative worlds so you cannot rely on your initial analysis to last for more than maybe a year. Your carefully selected CMS should stick five years. Therefore I don’t think a content model should be translated into code or a database tables. Ideally webmasters/editors should be able to change not only the descriptive and administrative metadata but also the structure of the content and even workflow – within the CMS. Systems do exist that have those capabilities, e.g. Drupal. And yes, that is a feature ;-)
I’m the “Chicago Humanities Festival’s”:http://www.chicagohumanities.org webmaster. Last year, we rolled out our new site that is similar to the hypothetical site described in this article, with the addition of serving as a multimedia archive. We didn’t develop the site in-house, and my role during the development was, in part, client liaison to the development team. My background is in web design and development, so I enjoyed the unique experience of sitting on the fence between client and firm, with my legs dangling over the client side.
Our development partner had a good internal understanding of a lot of the concepts discussed in this article. One part of the trip that was pretty rocky, however, was in the evolution of the content model beyond the initial model. The article suggests “Don’t Design the Perfect Model,” positing instead that content models will evolve, and iteration should be allowed for.
While I agree that striving for perfection in the initial model is futile, my experience in overseeing the development of our site suggests that an explanation of how content models evolve — and how this evolution is managed — is warranted.
Left on their own, content models will evolve (or devolve) independently on both sides of the client/firm fence. The client’s model tends to expand, bounded only by imagination. The firm’s model tend to become more concrete, or even whittled down, as the development team uncovers challenges and butts into the (perhaps hitherto undiscovered) limitations of the given CMS. Sometimes the client’s imagining results in a possible enhancement, which then has the potential to become a change order, which is great for the firm. However, there is also the possibility that the client’s imagination and exploration will uncover a flaw in the initial content model, or that an unforeseen limitation in the CMS breaks the model.
How can this be avoided? What kind of communication and management strategies need to be in place to effectively manage model evolution? How can a firm establish scope of work and yet leave the content model open to iteration?
Lots of great comments here agreeing that content modeling is important, and several pointing out that ACME product can handle any content model you throw at it. There’s less talk about the task analysis piece, which is telling: it’s the messy, politically challenging piece which technology people have generally managed to ignore up until now.
So I’ll say it again: it doesn’t matter how smart your content model is, and how well your chosen tool is able to model it—if the organization’s editorial staff can’t perform their key tasks in the time they have available, the project will fail. If you haven’t done task analysis, and run it past the actual editors who’ll use the system, the content strategy will be hampered and the project won’t achieve its goals.
an explanation of how content models evolve — and how this evolution is managed — is warranted. … How can this be avoided? What kind of communication and management strategies need to be in place to effectively manage model evolution? How can a firm establish scope of work and yet leave the content model open to iteration?
Fantastic questions, which I didn’t have space to address in the article. Not that I have all the answers!
This would be a great topic for an article. Why don’t you write it? ;)
There are a lot of good points in here most of which I have experienced first hand. I do not understand where clients get the idea that Wordpress is an enterprise level solution for their gigantic website. Even for smaller sites I still prefer to use Drupal or Expression Engine.
I think many developers make the mistake of expecting a client to understand whatever platform they choose for their client. I’ve had many clients come to me and asked me to switch them to a better platform that was easier for them to manage their content.
I really enjoyed that. Very interesting article, which has got me thinking – what is the most important aspect of a Content Management System? Well, to me personally, it is to be ease of use. I couldn’t bare to use a CMS which has a bucket load of features yet isn’t easy to navigate and post from. That would drive me crazy. A CMS has to be easy to use or the whole concept becomes pointless in my opinion.
i know quite a lot about this stuff as I myself is a developer but reading such great posts always helps one to in improving his/her skills. It has helped clearing few doubts i had about using CMS, from my mind.
This was a very interesting read, but I think the design side needs more focus. Design by itself is a much broader entity than art direction—it includes art direction. Art direction is a form of design. I think you need to focus your definition of design before the distinction between the two becomes useful
I can’t help but draw a wider circle, and see mishandled approaches to CMS integration as part of a global phenomenon where IT makes the call and manages the solution. Probably what’s fundamentally changed in business practices over the last 20-30 years is the way technology (be it the workstation on the desktop of the corporate intra- or internet website) has given stationery suppliers, actuaries and filing clerks a one-way ticket to redundancy. All three (I’m sure someone can add to the list) once had a secondary role, a lot further down the chain of command. IT and its mystic, inaccessible ways came along and required specialists to install and run it. Unfortunately, at some point they managed to convince management that they also were the only people technically competent enough to make decisions, and took on a management and deciding role that no procurement, statistical or storage division within a company would ever have been given. Maybe it’s time MBAs came with a compulsory unit of study on “understanding technology and keeping it in its place”…?
Our clients often have more questions than answers about selecting their Web Content Management Software. Many see Drupal, or Wordpress as a simple solution because of the ever increasing visibility. Thank you for taking an approach that does not start with a solution and work backwards such as touting why one CMS is better than another!
Our clients often have more questions than answers about selecting their “Web Content Management Software”:http://www.duoconsulting.com/ Many see Drupal, or Wordpress as a simple solution because of the ever increasing visibility. Thank you for taking an approach that does not start with a solution and work backwards such as touting why one CMS is better than another!
Thank you dor this article which is very deep. The quality of contentis very important but sometimes it’s hard de write. I often spin my content with online tools. I save lot of time.
34 Reader Comments
Back to the ArticleSDtechcomm
Thank you for this piece, more and more people are joining the daunted hunt for the right CMS daily.
@gsal: While I have to agree that technology can be no “silver bullet”, there are certain companies which have recently rolled out the red carpet for content management systems that have cutting-edge curation analytics capabilities. I am a tech comm and my company just adopted MindTouch 2010 and it’s already given us helpful feedback on our online documentation. MindTouch offers a free trial system so you can walk away with no strings attached if you believe that your content issues aren’t “technology-deep”. But MindTouch might just give you the feedback you need on really whatever content issues you’ve been harboring. Here’s the article on Mindtouch 2010’s curation analytics that won us over: http://www.cmswire.com/cms/enterprise-20/mindtouch-2010-includes-curation-analytics-for-content-and-documentation-008045.php
Liz
blainsmith
This was a great article and I think most of us agree with all of your points. The ultimate challenge is getting the stakeholders to invest both time and money into content strategy rather than just signing on the dotted line for a tool.
I have learned that you really have to be blunt in explaining the important of developing content first. I have actually told a friend of mine, “your site sucks because your content sucks” and he quickly understood. That’s more difficult to tell a client who is paying thousands of dollars for your services.
vidar_webnodes
A very good article!
It’s nice to see some more focus on content modeling, and the value of creating a semantic structure for the content in a CMS. As a vendor, we’ve had semantic content and content modeling technologies as one of our most important areas of R&D for several years now(http://www.webnodes.com/ontology-engine), as we see it as one of the most important factors in creating good websites, both for visitors and editors.
However, I do not agree with the presumption that:
“Any web project more complex than a blog requires custom CMS design work.”
Of course, we might have different interpretations about what constitutes custom CMS design work. What do you define as “custom CMS design work”? Could you give an example? To me customizations are changes to the CMS that are outside the scope of the intended built-in flexibility of a CMS.
Such customizations should only be done as a last resort IMO. Most likely, you have selected the wrong CMS, if it isn’t possible to solve your problems within the built-in flexibility.
I’ve seen so many customizations over the years done to different content management systems and other IT systems that have turned out to be extremly costly over the lifetime of the solution. This has mainly been due to upgrade problems, and a limited number of people that knows how the customizations are set up.
Jonathan Kahn
Thanks for the kind words, guys!
Jonathan Kahn
@blainsmith:
Exactly!
Jonathan Kahn
@vidar_webnodes:
I go into a bit more depth later on, starting with, “A product can’t fix content problems out of the box.” Whether or not customizations are within the scope intended by a software vendor, web teams need to spend time customizing tools to their needs. Software vendors tend to talk about “solutions”, which is presumptuous: solutions to which problem?
The “intended built-in flexibility” of proprietary tools is beyond the scope of this article. If you approach a software vendor/tech team/open source community with a clear idea of your content model and task flows, they should be able to estimate how much time and money is required to customize the tools for the project’s needs.
vidar_webnodes
Sure, they should be able to estimate how much time it would take to customize the tools for the project’s needs, but there are several potential pitfalls. Yes, you can hack most CMS tools to support a given data model, but there’s a BIG difference between supporting a data model within the “intended built-in flexibility” of the data model, and hacking together support in a system never intended for such a data model.
How much will it cost to upgrade when the next version arrives? How well will it work? How polished will the customization be compared to the built-in functionality? How well will it scale with lots of traffic?
marksteven
Thought provoking article, but the choices we really have are way more limited than you’d suggest.
Most web development firms are working with just three of four different platforms, representing different bands of the cheap ‘n’ simple, to enterprise-level spectrum. The choice will always be between your favourite 3 (or so) CMS because:
1) You can’t be really good at more than 3 or 4 systems if you’re undertaking lots of custom implementations
2) Supporting too many platforms as a supplier introduces risk and sustainability: how do you retain the knowledge of a system if you have any staff turnover?
3) If you choose a platform on project fit alone, and it isn’t something you know well, you won’t be building something that meets all the best practices for that tool – these essential nuances to a build are developed over time
When choosing a CMS the real question is, which of my (3 or 4) favourite systems is a good fit. And failing that, should I build the required features in a custom application?
As the article points out content modelling is essential in understanding this.
Buts let’s keep it real, there are never 1000s of options on the table, but a maximum of 5. Maybe a few more if you include low cost, hosted solutions like Ning, Shopify etc.
yellowboxstudio
I heard a talk about Expression Engine not too long ago and it seemed to be highly versatile. I’ve been using Wordpress since its conception many years back and have watched it grow from a little blog tool to one of the most powerful free CMS’ out there.
I am doing an hour long talk about history of Wordpress and where 3.0 is at today, MU, Buddypress, etc. I feel with the hundreds of thousands of plugins and themes already available it is hard to promote anything else.
kevinfreitas
Thanks for a great article and discussion piece! It’s refreshing to see this perspective clearly articulated. I work for a custom web dev and design shop who, for the last 12 years, have focused on always building custom solutions for our clients. It’s worked well because, before anything, we work with our clients to determine their needs and present our plan for a solution before we even lay down any code. No crowbars to hack up other, pre-fab CMS’s but instead a unique product created just for them.
Anyone can buy a frozen dinner and try to doctor it up with some salt and pepper but wouldn’t you rather have a chef cook to your own tastes?
Sarah Daniels
Thank you for sharing your thoughts about this topic. Making us remember what really a CMS is.
Jonathan Kahn
@marksteven:
I think your logic is backwards. Of course I’m not suggesting we should survey all of the thousands of tools out there—that would be impossible. But focusing on “building something that meets all the best practices for that tool” sounds to me like a classic distraction from the content strategy question. It’s an excuse to stick with tools you know rather than involving yourself in a strategic process. The tool isn’t the thing: the content is.
Eric Sol
The focus on content in your article is much appreciated, as well as your basic outline of content modelling (I was trained as an information analyst and produced many diagrams and flowcharts). However, on the web things change much much faster than in most administrative worlds so you cannot rely on your initial analysis to last for more than maybe a year. Your carefully selected CMS should stick five years. Therefore I don’t think a content model should be translated into code or a database tables. Ideally webmasters/editors should be able to change not only the descriptive and administrative metadata but also the structure of the content and even workflow – within the CMS. Systems do exist that have those capabilities, e.g. Drupal. And yes, that is a feature ;-)
dulybookmarked
I’m the “Chicago Humanities Festival’s”:http://www.chicagohumanities.org webmaster. Last year, we rolled out our new site that is similar to the hypothetical site described in this article, with the addition of serving as a multimedia archive. We didn’t develop the site in-house, and my role during the development was, in part, client liaison to the development team. My background is in web design and development, so I enjoyed the unique experience of sitting on the fence between client and firm, with my legs dangling over the client side.
Our development partner had a good internal understanding of a lot of the concepts discussed in this article. One part of the trip that was pretty rocky, however, was in the evolution of the content model beyond the initial model. The article suggests “Don’t Design the Perfect Model,” positing instead that content models will evolve, and iteration should be allowed for.
While I agree that striving for perfection in the initial model is futile, my experience in overseeing the development of our site suggests that an explanation of how content models evolve — and how this evolution is managed — is warranted.
Left on their own, content models will evolve (or devolve) independently on both sides of the client/firm fence. The client’s model tends to expand, bounded only by imagination. The firm’s model tend to become more concrete, or even whittled down, as the development team uncovers challenges and butts into the (perhaps hitherto undiscovered) limitations of the given CMS. Sometimes the client’s imagining results in a possible enhancement, which then has the potential to become a change order, which is great for the firm. However, there is also the possibility that the client’s imagination and exploration will uncover a flaw in the initial content model, or that an unforeseen limitation in the CMS breaks the model.
How can this be avoided? What kind of communication and management strategies need to be in place to effectively manage model evolution? How can a firm establish scope of work and yet leave the content model open to iteration?
Jonathan Kahn
Lots of great comments here agreeing that content modeling is important, and several pointing out that ACME product can handle any content model you throw at it. There’s less talk about the task analysis piece, which is telling: it’s the messy, politically challenging piece which technology people have generally managed to ignore up until now.
So I’ll say it again: it doesn’t matter how smart your content model is, and how well your chosen tool is able to model it—if the organization’s editorial staff can’t perform their key tasks in the time they have available, the project will fail. If you haven’t done task analysis, and run it past the actual editors who’ll use the system, the content strategy will be hampered and the project won’t achieve its goals.
Jonathan Kahn
@dulybookmarked:
Fantastic questions, which I didn’t have space to address in the article. Not that I have all the answers!
This would be a great topic for an article. Why don’t you write it? ;)
adamfullerdesign
There are a lot of good points in here most of which I have experienced first hand. I do not understand where clients get the idea that Wordpress is an enterprise level solution for their gigantic website. Even for smaller sites I still prefer to use Drupal or Expression Engine.
I think many developers make the mistake of expecting a client to understand whatever platform they choose for their client. I’ve had many clients come to me and asked me to switch them to a better platform that was easier for them to manage their content.
gary223
I really enjoyed that. Very interesting article, which has got me thinking – what is the most important aspect of a Content Management System? Well, to me personally, it is to be ease of use. I couldn’t bare to use a CMS which has a bucket load of features yet isn’t easy to navigate and post from. That would drive me crazy. A CMS has to be easy to use or the whole concept becomes pointless in my opinion.
maneet puri
i know quite a lot about this stuff as I myself is a developer but reading such great posts always helps one to in improving his/her skills. It has helped clearing few doubts i had about using CMS, from my mind.
Indeed helpful!
Rosing
This was a very interesting read, but I think the design side needs more focus. Design by itself is a much broader entity than art direction—it includes art direction. Art direction is a form of design. I think you need to focus your definition of design before the distinction between the two becomes useful
websitebuilder
I can’t help but draw a wider circle, and see mishandled approaches to CMS integration as part of a global phenomenon where IT makes the call and manages the solution. Probably what’s fundamentally changed in business practices over the last 20-30 years is the way technology (be it the workstation on the desktop of the corporate intra- or internet website) has given stationery suppliers, actuaries and filing clerks a one-way ticket to redundancy. All three (I’m sure someone can add to the list) once had a secondary role, a lot further down the chain of command. IT and its mystic, inaccessible ways came along and required specialists to install and run it. Unfortunately, at some point they managed to convince management that they also were the only people technically competent enough to make decisions, and took on a management and deciding role that no procurement, statistical or storage division within a company would ever have been given. Maybe it’s time MBAs came with a compulsory unit of study on “understanding technology and keeping it in its place”…?
DuoConsulting
Our clients often have more questions than answers about selecting their Web Content Management Software. Many see Drupal, or Wordpress as a simple solution because of the ever increasing visibility. Thank you for taking an approach that does not start with a solution and work backwards such as touting why one CMS is better than another!
DuoConsulting
Our clients often have more questions than answers about selecting their “Web Content Management Software”:http://www.duoconsulting.com/ Many see Drupal, or Wordpress as a simple solution because of the ever increasing visibility. Thank you for taking an approach that does not start with a solution and work backwards such as touting why one CMS is better than another!
Le Caviar
Thank you dor this article which is very deep. The quality of contentis very important but sometimes it’s hard de write. I often spin my content with online tools. I save lot of time.