How Blend Chooses a CMS
Over its history, Blend has developed a pattern of being an early North American partner for a number of European CMS systems. Blend was an early North America Partner for eZ Publish (now Ibexa), Episerver (now Optimizely), and Umbraco (still Umbraco!).
While all of these CMS vendors have gone international long ago, one has to wonder how a Midwest web shop continues to help introduce European vendors to North America. It brings up an interesting question: how does Blend choose what CMS systems to work with?
Balancing exclusivity and variety.
First, it’s important to understand the relationships web shops have with their CMS vendors.
On one hand, there are web firms that strive for exclusivity. They work with one system, and they are an exclusive partner for that vendor. This provides a kind of authority on that one system, and it provides a strong partnership for referrals and information transfer.
On the other hand, some web firms try a different system with nearly every project. This variety allows them to stay agile. They look for bleeding edge options in order to stay as far ahead of the technology curve as possible.
At Blend, we try to land somewhere in the middle. To us, it’s strategically important to work with more than one solution. If all you have is a hammer, everything is a nail, and it becomes hard to grow and add new techniques. By working with multiple systems, we can compare their approaches and cross-pollinate ideas between frameworks.
However, it’s also important that we’re deep experts on the systems we work with. We want to work with the same systems often enough to develop best practices and see how design decisions affect the site in the long term.
In doing so, we balance expertise and history with the ability to shift and adapt.
Experimenting outside the core.
Of course, there’s a lot that goes into onboarding a new content management system. It’s not as simple as downloading some files and running wild.
At Blend, we’ve developed a process where we work with new CMS systems on internal, low-risk, or pro-bono projects and roll them out to client solutions over time. This allows us to make low-stakes decisions while also vetting for fit. Not every CMS will fit our model, our people, or our clients, and we want to know this before we walk into a major project.
Even then, we’re selective in which systems we work with. We look for the following things when evaluating a potential new CMS:
- Good code and developer experience. The most important characteristic of any software is that it’s well-written. If the CMS has been designed and executed well, it will be easy to work with and extend to build anything we might need to add for a project.
- Good workflow. It’s important with a CMS to be able to work through multiple environments, so you can work locally to build features but still bring down the production content to test or diagnose problems.
- An active developer community. Whether a solution is commercial or open source, we want to be part of an active community. This shared experience is invaluable — from the community’s ability to find and quickly diagnose problems, to the pool of talent willing to mentor and educate. An active community within a newer system is usually a sign that code and developer experience are heading in the right direction.
- A strong content model. A content model is structure within in the CMS that allows you to assign fields and information to documents, pages, and files. We look for CMS vendors that provide a healthy selection of methods for storing data (like text, dates, numbers, and maps), as well as a way to extend those methods in case something isn’t covered.
- An implicit content tree. While the content model defines the shape of the information on your site, the content tree defines how it’s organized. Some systems (notably Wordpress and Drupal) put most of their content in a big sorted list, relying on menus to help navigate and find pages. In contrast, implicit trees define nearly all content as being part of a large site map that covers the whole site. Our projects tend to lean toward an implicit tree structure, a structure both Umbraco and Optimizely employ, with flexibility to allow for the ‘big sorted list’ method for specific sections of the site.
Making the choice.
None of the items we’ve mentioned above are explicitly European. Which is to say: we don’t limit ourselves to fledgling European CMS vendors, even if the evidence says different.
No, this is a simple coincidence.
The actual choice is based on our past experience with content management systems. We’ve worked in the field long enough to know what works for the types of projects we take on. And that trio of European CMS vendors — EZ Publish, Optimizely, and Umbraco — all fit the requirements we mentioned above.
Working with a controlled set of systems means that we get to be deep experts in those systems, and being intentional about which vendors we choose — while constantly evaluating other systems — allows us to find the best ideas and make changes when new systems come along.
Make no mistake: being early North American partners with both Optimizely and Umbraco has allowed us to find uncovered opportunities, all while introducing some great tools to our clients.
But that’s really easy to do if the tools are great to begin with.
Resources on the discovery process.
We’ve written at length, both here and beyond, on research and discovery.
A Better Budget: How Tech Planning Improves the Scoping Process
One of the toughest parts of any web project is estimating how much the project will cost and how long it will take. Technical planning helps create better budgets, plan for more realistic timelines, and improve the efficiency of the entire project.
Episode 8: Gather Insight From Your Metrics (w/ Jon Crowley) Off-site link
Corey and Deane talk about the first time they tracked analytics on their blogs in the early 2000s. Then, Jon Crowley, Senior Vice President of Strategy at Diamond Marketing Group, talks to us about the balance between data and insights — how to focus on questions rather than raw numbers, how to look for answers rather than “trying to be correct,” and a when we can take data at face value. (He also gives us a tour of his shoe collection.)