What is your developer-dependent CMS actually costing you? Calculate your costs and get a full report
What is your developer-dependent CMS actually costing you? Calculate your costs and get a full report
Evaluating Agility
Structure a replatforming business case: measure a baseline, estimate effort from your own inventory, plan phases and roles, quantify benefits, and manage the risks.
A replatforming business case has to answer three questions for the people who approve it: why change, what will it take, and what do we get. This guide gives you a structure, a way to estimate effort from your own inventory, and the risks to plan for. It deliberately contains no stock numbers. Estimates built from your own content, team and proof of concept are the ones decision-makers trust.
A business case that gets approved usually has these sections:
You can't show improvement without a starting point. Measure before you change anything. The "Measuring Website Performance" section of Building a Website suggests combining leading and lagging measures:
| Measure | Type | Where it comes from |
|---|---|---|
| Lighthouse score, desktop and mobile | Leading | Lighthouse, on your key page types |
| Engagement: bounce rate, time on page, time on site | Leading | Your analytics tool |
| Conversion rate | Lagging | Your analytics or CRM |
| Online revenue, if relevant | Lagging | Your commerce or finance systems |
Add measures of how your team works, because that's where a CMS change shows first:
Effort comes from a handful of drivers. Count each one for your project; the counts, not a rule of thumb, are the basis of your estimate.
| Driver | How to count it | What makes it bigger |
|---|---|---|
| Content types | Distinct kinds of content (articles, products, people, events) | Inconsistent structure in the source, heavy rich-text formatting |
| Content volume | Items per type, per locale | Many locales, content that needs cleaning or rewriting |
| Page types and components | Distinct page layouts and building blocks in the design | Many one-off designs; components that behave differently per page |
| Assets | Number of images and files, and where they live today | Missing alt text, assets spread across several systems |
| Integrations | Search, forms, analytics, CRM, commerce, sign-on, translation | Custom or legacy integrations without a modern API |
| URLs and SEO | Indexed URLs, redirects needed, URL changes planned | A new URL structure, many inbound links |
| Sites and channels | Websites, apps and other destinations | Different structures per site |
| People | Editors to train, teams with different workflows | Many teams, approval rules that differ by team |
Two things sharpen the estimate more than anything else:
Agility recommends a phased, iterative approach: launch a first phase as a minimum viable product, then improve in smaller releases (Building a Website). A typical plan:
| Phase | What happens | Main roles |
|---|---|---|
| Discovery and proof of concept | Inventory, fit check, POC, business case | Sponsor, lead developer, content architect, editors |
| Content architecture | Content, component and page models; workflow and roles | Content architect, editors, admin |
| Build | Front end, preview, Web Studio, integrations, environments | Developers |
| Migration | Map, clean, import and verify content; redirects | Developers, editors |
| Training and governance | Editor training, roles, approval workflows | Admin, editor leads |
| Launch | Content freeze, cutover, DNS, monitoring | Developers, admin, sponsor |
| Iterate | Improvements in small releases against the baseline | Everyone |
Content work can start before the front end is finished. Agility recommends that editors enter content early, so developers build against real content instead of placeholders.
Tie each benefit to a baseline measure, and estimate it from your own numbers:
| Risk | How to manage it |
|---|---|
| Losing search rankings | Inventory URLs early, build a redirect map, and test redirects before launch. |
| A content model that doesn't fit | Model with real content and review it with editors in the POC. |
| Migration taking longer than planned | Migrate a sample first; budget for cleaning, not just copying. |
| Integration surprises | Build the riskiest integration in the POC. |
| Editors not adopting the new platform | Involve editors in the POC and train before launch. |
| Being locked in | Confirm how you would export everything. See Data Portability and Exit. |
AI coding tools and the Agility CMS MCP Server can shorten the build and migration phases: the AI tool can create models and seed content as well as write the front end. If you plan to rely on that, test it in the POC and estimate from what you measure. Vibe Coding with Agility describes the approach and its limits.