Case study · Architecture PoC · CMS / DAM
Built the proof. Changed the roadmap.
Our digital ecosystem ran on a legacy Java monolith with an outdated WCMS. Publishing a simple news article took five manual steps. I built a composable proof of concept, benchmarked it against the existing platform, and the measured results informed the platform roadmap.
- Role
- Web Designer & Developer · PoC lead
- Scope
- Content model, code, API integration
- Method
- Controlled publishing test, 6 pages × 2 systems
- Stack
- Contentful · Cloudinary · Next.js
Context
Key challenges with the existing platform:
- Deployments every 6–8 weeks
- Strong dependency on developers for marketing pages
- No drag-and-drop or modular editing
- Large text fields misused for inline HTML
- Manual asset resizing in Photoshop
- Re-uploading multiple image sizes into the DAM
- No dynamic media transformation
- Limited performance optimization (static JPG/PNG)
- 01Marketing writes a Word brief
- 02Publisher downloads assets
- 03Manually resizes images
- 04Re-uploads assets
- 05Rebuilds page in CMS
The workflow was slow, rigid, and operationally expensive.
Trigger
As Web Designer & Developer, I was directly exposed to these inefficiencies daily. At the same time, technological benchmarking showed how far modern headless CMS and media platforms had evolved.
What started as frustration became a structured initiative: I proposed building a proof of concept to validate whether a composable architecture could significantly improve speed, autonomy, and performance.
Strategic Vision
We needed:
- A composable architecture (replace tools independently)
- A flexible headless CMS
- Greater marketing autonomy
- Dynamic media transformation
- CDN-based optimized delivery
- Measurable performance improvements
After auditing multiple solutions, I selected:
- Contentful for flexible content modeling
- Cloudinary for on-the-fly media transformation and CDN delivery
I built a POC aligned with our existing e-commerce content model to ensure structural compatibility.
Experiment Design
To validate impact, I ran a controlled test. Six pages were each published once in the current system and once in the POC, and I measured publishing time end to end using a stopwatch.
| Page type | Pages | Legacy | POC |
|---|---|---|---|
| News page | 2 | ✓ | ✓ |
| Athlete page | 2 | ✓ | ✓ |
| Story page | 2 | ✓ | ✓ |
Result
Publishing time was reduced by approximately 60%.
This meant:
- Same team could publish significantly more content
- Or free capacity for higher-value initiatives
- Reduced manual resizing and upload work
- Increased clarity and consistency in workflow
Additionally
- Dynamic media delivery improved Lighthouse scores
- Eliminated manual generation of multiple image sizes
- Reduced dependency on frontend developers for marketing execution
Trade-offs & Resistance
Main resistance: budget. Tool licensing costs were visible. Time savings and long-term scalability were not.
A deliberate constraint
We avoided testing directly with marketing teams to prevent frustration in case the initiative was paused. Validation was therefore conducted internally within the web team.
Strategic Insight
This project was not just a tooling upgrade. It revealed that:
- Editorial workflows can be fundamentally restructured
- Marketing autonomy can scale
- Performance and flexibility are directly linked
- Platform architecture directly impacts organizational speed
It shifted the conversation from “which CMS?” to “how do we design systems that enable velocity?”
Validated · informed the platform roadmap
The measured results were presented to stakeholders, and the composable architecture informed the platform roadmap.
Publishing times were measured end to end with a stopwatch on six pages, each published once in each system. The ~60% reduction is an internal benchmark run by the web team, not a study with the marketing teams.
Back to home