Crownpeak Software Architecture: WCM in the Cloud

Jim Howard
By Jim Howard
July 30, 2012

Crownpeak’s architecture is what makes things better for our customers. It’s the grand canyon of differentiation for us. Anybody can host a Web Content Management System (WCM) in the cloud, but that’s not the key. The cloud helps us only because we have architected our product specifically to run well there. Our differentiating advantage is the architecture we created to fix the problems that have existed in the WCM market for many years. It’s an architecture designed to deliver a global, high-availability, high-security shared service from the cloud – a native SaaS architecture. But it’s also an architecture designed to make marketers at large enterprises agile and IT teams more productive.

When we created our product, we had to take an architectural approach nobody had ever considered for WCM before. WCM prevented agility. And we knew how to fix it.

Software and Architectural Requirements:

  1. Functional parity with the top WCM systems
  2. Every customer runs on the same version of the product
  3. No customizations can be allowed. Therefore, all customer requirements must be possible via configuration
  4. No configuration can impact system performance or security
  5. Configuration must be possible by any IT organization or web development company with minimal training
  6. The configuration and deployment process must be very rapid
  7. QA for deployment must be limited to function and layout only (No QA on load, performance, security, stability, etc.)
  8. Unlimited independent teams can configure solutions simultaneously
  9. Adding new output formats are simple configuration options
  10. Every configuration is supported forever or adjusted “on our dime
  11. Developing and releasing new configuration options is possible rapidly
  12. All configurations and the product User Interface (UI) call an API so core product changes don’t impact customer configurations or UI functions
  13. Customers are absolutely isolated from each other
  14. Customers may have many different solutions and control content sharing, access and permissions independently or universally across solutions
  15. The system can scale infinitely but the costs/customer drops with scale
  16. Fast content to site visitors, globally, no matter the configuration
  17. The system has to have better than 99.9% up-time
  18. Unplanned usage spikes cannot affect performance
  19. We can never risk losing customer data
  20. We are held to higher security standards than on-premises WCM systems
  21. We must integrate directly with the other best-of-breed digital marketing and content functions in common use by enterprise customers
  22. We must provide a complete and open API for every function and direct access to the content repository for use by new or custom integrations
  23. We must be compliant with the most rigorous corporate standards
  24. We support every aspect of the system including configurations and integrations

The result of this list of requirements is our completely unique architecture – and it’s a huge advantage for us and for our customers. On-premises products can have holes that are fixed by custom development. They can require a lot of infrastructure to run at heavy load. They need the crutch of CDN to distribute content globally. They can have complicated structures that make the system difficult to maintain over time. They can require time-consuming and painful upgrades. They can have a clunky and difficult integration approach. We simply can’t.

I’ll leave the details of the architecture: levels of abstraction, approach to the configuration environment, logical and physical approach to redundancy and our global deployment environment, for a later post.

But the end-result of these architectural innovations is that marketers using Crownpeak can implement changes very rapidly, work on multiple projects simultaneously, and take advantage of innovation instantly. IT and eBusiness teams can roll out powerful, integrated solutions much more rapidly than previously possible. But we like it too because it’s our grand-canyon of differentiation and creates a nice barrier to competition from the established folks like Oracle, IBM, Adobe, OpenText and HP because to have a true SaaS, they would have to create a brand new product from the ground-up. I’m not saying they won’t, but it may take them years to get around to it.

comments powered by Disqus