Insights17 November 2026

Migrating from Optimizely to Umbraco: keep the capability, rethink the cost

Moving away from Optimizely doesn’t have to mean giving up personalisation, experimentation or enterprise standards. It can be an opportunity to build a more focused digital platform and put more of your budget into improving it.

By Oliver Bailey

Back to all posts

Moving away from Optimizely doesn’t have to mean giving up personalisation, experimentation or enterprise standards. It can be an opportunity to build a more focused digital platform and put more of your budget into improving it.

The renewal arrives. Someone asks which features the organisation actually uses. Someone else asks why a relatively straightforward website needs such a substantial platform budget.

Then comes the difficult question: what would we lose by moving?

For organisations considering a migration from Optimizely to Umbraco, that deserves a proper answer. Licence cost matters, but so do editorial freedom, security, performance and the ability to keep improving the experience.

The opportunity is to understand which capabilities earn their place, choose how to deliver them and remove the complexity that no longer serves the organisation.

Start with what your organisation actually uses

Optimizely brings together content management, experimentation, personalisation and other digital capabilities. For an organisation actively using that breadth, an integrated platform can make sense.

But owning a capability and benefiting from it are different things.

Before deciding to migrate, examine what happens in practice. Which personalisation rules are live? How many experiments have produced useful decisions? Which integrations are essential? Where do editors still depend on developers?

Some features will be business critical. Others may have been bought for a roadmap that never happened.

That distinction shapes the migration. It prevents a team from spending money recreating functionality nobody uses, or discovering too late that an apparently minor feature supports an important operational process.

It also makes the commercial comparison more useful. You can compare the cost of delivering your actual requirements.

Headless gives the CMS a more focused job

In a headless architecture, the CMS manages content and makes it available through APIs. A separate application controls how that content becomes a website or another digital experience.

Editors might manage product information, articles and landing pages in Umbraco, while a Next.js front end presents them to visitors.

This separation gives teams more freedom over presentation, delivery and integration. It also allows content management and the public website to evolve at different speeds.

Umbraco’s Content Delivery API provides structured content to that separate front end. With Umbraco Cloud, the content management platform sits within a managed hosting environment, while the public website can use infrastructure suited to its own requirements.

Headless is an architectural choice rather than a reason, by itself, to migrate. Optimizely also supports this approach. The question is which combination of platform, architecture and operating cost best supports your organisation.

Licence savings need a complete cost comparison

One attraction of moving to Umbraco is the opportunity to reduce recurring platform costs. But the useful comparison goes beyond the licence line on an invoice.

At the time of writing, Umbraco publishes the following starting prices for its managed Cloud platform:

Umbraco Cloud planPublished price
Starter€45/month
Standard€280/month
Professional€730/month
EnterpriseCustom

These are platform prices, not complete project budgets. Resources, environments, support requirements, usage and additional services affect the final cost. Prices checked November 2026.

An enterprise implementation should be sized around its actual requirements. The cheapest entry-level plan is not a meaningful comparison with a platform supporting demanding availability, governance or integration needs.

Optimizely packages its plans individually, so a credible comparison needs your actual contract and renewal terms.

The calculation should include migration, ongoing operations, front-end hosting, testing tools and the cost of maintaining connections between them.

A lower annual platform bill is valuable. A lower bill combined with a platform your team can improve more easily is a stronger business case.

It can also change how the digital budget is used. Money previously committed to platform overhead can become available for research, accessibility, better customer journeys and improvements that people actually notice.

Personalisation and testing can live outside Optimizely

Moving the CMS does not mean abandoning optimisation. It means deciding where those capabilities should sit.

For some organisations, keeping personalisation close to the editorial experience makes sense. Others benefit from a dedicated experimentation platform that works across their wider digital estate.

Two options worth evaluating are:

ApproachWhere it can fit
Umbraco EngageTeams wanting content personalisation and A/B testing close to their Umbraco editorial workflow
AB TastyOrganisations needing dedicated web experimentation, personalisation or server-side feature experimentation

Umbraco Engage supports headless personalisation, segmentation and A/B testing through its headless package, with documented feature limitations. Its suitability needs to be checked against the chosen delivery architecture.

AB Tasty offers web and server-side experimentation capabilities. It can be evaluated separately from the CMS, allowing content management and optimisation to have their own development paths.

Neither approach should be treated as a drop-in replacement without checking workflows, integrations, data requirements and commercial terms.

The important distinction is between showing different content and running a trustworthy experiment.

An experiment needs consistent allocation, reliable exposure tracking, meaningful outcome measures and enough evidence to support a decision. A personalisation programme needs useful audience definitions, appropriate data handling and someone responsible for maintaining its rules.

Buying another tool does not supply those disciplines automatically.

The migration is a good moment to establish which activities are producing value, which need improvement and which can be retired.

A leaner stack can still meet demanding requirements

For a financial services client, we’re developing an architecture that combines Umbraco Cloud for content management, a Next.js static export for the website and Cloudflare for public delivery.

Content is retrieved during the build. Pages and their required media are prepared ahead of time and deployed as website assets. Visitors can then read that published content without the CMS processing each request.

This creates a useful separation: the CMS supports publishing, while the deployed website supports public access.

If the CMS becomes unavailable, the existing deployed pages remain available. Publishing new content still depends on a successful build, and search, forms and other integrations have their own availability requirements.

For a content-led website, this can reduce the amount of processing required to serve visitors and make public delivery less dependent on the availability of the publishing platform.

The architecture is in development, so these are design decisions rather than claims about measured production results. The principle is straightforward: avoid making every visitor request depend on systems that do not need to be involved.

Security comes from the implementation

Separating public delivery from content management can reduce direct exposure of the CMS. It does not make a website secure automatically.

A dependable implementation still needs controlled editorial access, protected previews, secure deployment credentials, dependency maintenance, monitoring and a tested recovery process. These should form part of your Umbraco maintenance and support plan.

The build pipeline matters too. If publishing content triggers a deployment, the code being built must already have passed the appropriate review and checks.

Personalisation introduces another consideration: caching. A page intended for one audience must not accidentally become the shared cached response for everyone. Sensitive information and preview credentials must never enter public build output.

These details deserve attention during architecture and delivery, alongside the choice of CMS.

Enterprise capability is demonstrated through these controls and the way the service is operated.

Static where it helps, dynamic where it matters

Pre-rendered content can be an efficient choice for articles, information pages and many marketing journeys.

It cannot, on its own, deliver genuinely live availability, authenticated account data or decisions made for each individual visitor.

Those requirements need an appropriate runtime service. That might be an API, an edge function, a specialist platform or a server-rendered part of the application.

The aim is to introduce that processing where it adds value. An informational page and an authenticated application do not have to carry the same infrastructure burden.

The same applies to experimentation. If a test requires a decision for each visitor, the architecture needs to support that decision, record the exposure and measure the outcome. It should be designed into the experience rather than bolted on after launch.

Treat the migration as a service change

Content transfer is only one part of moving from Optimizely to Umbraco.

A successful migration also protects search visibility, redirects, forms, analytics, accessibility and the editorial processes people rely on. Existing personalisation rules and experiments need explicit decisions: migrate them, redesign them or retire them.

Editors should test realistic publishing tasks before launch. Technical teams should rehearse deployment and rollback. The organisation should know what success looks like beyond getting the new site live.

That might mean shorter publishing times, fewer support requests, faster pages, lower operating costs or an increased capacity to run useful experiments.

At Absurd, we approach replatforming through service design and Umbraco development together. We examine what customers need, how teams work and which technology supports both.