Spinning lathe chuck next to a new spare part

Artikel

DXP

Storyblok

Migration

Relaunch

Moving without downtime: how your website switches to a DXP

Moving from a long-grown CMS to a digital experience platform does not mean taking your website offline. The new platform is built in the background, and the old one keeps running until everything has been checked and approved. The switch happens on an agreed cut-over date.

The question usually comes from management, right after the one about budget: will we be offline for a while? IT follows close behind, wanting to know when it has to change something on the domains. And customer service does not want a service technician to open the manual at a customer's site and land on an error page.

At Werbeberg, a DXP consists of the headless CMS Storyblok, where your editorial team manages content, and a custom-built frontend that delivers this content as a website. Headless means content and presentation are technically separate and come together through an interface (API). The way there looks similar in every project, even though scope and duration vary a lot.

Most important facts

  • //Die alte Website bleibt online, bis die neue geprüft ist, umgestellt wird an einem vereinbarten Stichtag.
  • //Ein DXP-Projekt läuft in sieben Phasen, von der Bestandsaufnahme bis zum laufenden Betrieb.
  • //Ein vollständiger Weiterleitungsplan sichert alte Links und Google-Rankings über den Umstieg hinweg.
  • //Ihr Team wird zu planbaren Zeitpunkten gebraucht, am meisten bei Inhalten und Freigaben.
  • //Verzögerungen entstehen selten durch die Technik, meistens durch späte Inhalte, lange Freigaben und Sonderwünsche.

// 1

Running in parallel

Your customers will not notice the rebuild

While the build is under way, the project team works on a separate test environment called staging. That is where the frontend and content are created, where the product manager clicks through the first pages, and where IT finds bugs before anyone outside sees them. The public website is not affected.

The switch only happens once content, functions and redirects have been checked and approved. Technically, the switch is then a change to the domain settings (DNS) at an agreed time. We like to put that date on a quiet day. The week before the Hannover trade fair is a poor choice, and so is the day the new spare parts catalogue is meant to go live.

The only tricky part is the time in between. Anyone who still changes something on the old site has to make the same change on the new platform, otherwise it gets lost in the switch. So during the project months there is a simple rule: touch only what is necessary on the old site. Shortly before go-live comes a content freeze, a short phase in which nobody changes content.

// 2

How the project runs

From the first workshop to a trained editorial team

It starts with an inventory. We record which pages, content types, languages and forms exist, which interfaces are in place and how often each page is visited. We also talk to marketing, IT and the specialist departments. They usually know exactly which pages nobody has opened for years and which ones sales needs every day.

This analysis produces the content model. It defines which building blocks a page consists of and which fields the editorial team fills in, for a product for example name, data sheet and technical data. In parallel, we design the layouts for the most important page types. This phase is where most of the decisions are made that save time and money later.

Then comes the build. Development codes the frontend, sets up Storyblok with components and languages and connects the interfaces. Roles and approval workflows can also be set up in Storyblok. Before the start, we go through together which features your Storyblok contract includes.

How long all this takes depends on the number of pages, languages and interfaces, and on how well the content is prepared. As a guide: [PLATZHALTER: typische Dauer je Phase für Analyse, Konzept und Design, Migration, Schulung und Go-live].

As soon as the first page types are in place, content migration begins. We move well-structured content by script, and the rest is entered by hand. Texts that needed revising anyway go into the system in their new version straight away.

At the same time, the editorial team gets to know the system, working on its own content rather than a demo project. Anyone who creates the data sheet for the next machine during training finds their way around faster in daily work afterwards.

Shortly before go-live, a topic that is often underestimated comes to the fore: the old addresses. A new website usually gets a new URL structure. Without precautions, links from search engines, newsletters or partner sites then lead nowhere, and visitors and rankings are lost with them.

That is why every migration includes a redirect plan. It starts before the relaunch with a complete list of all existing URLs, plus the pages with the most visits and inbound links. Every old address gets a new one, set up as a 301 redirect, a permanent redirect that search engines understand as a move. Page titles, meta descriptions and structured data go into the content model so that the editorial team can maintain them later.

On the cut-over date come acceptance, the final test of the redirects and the domain switch, and the new sitemap is submitted to Google Search Console. In the days after, we watch error pages and rankings and add any missing redirects. Then the platform goes into operation. Ongoing support covers updates, security patches and help, and you commission further development as needed. Where the frontend runs and who holds the hosting contract is agreed together before the project starts.

What happens at each stage

Seven phases in three stages

Icon: blueprint

Plan

Inventory, content model and design. This is where the structure of the new website is set and who maintains what.

Icon: building blocks and hammer

Build

Frontend, Storyblok setup and interfaces, plus content migration and training. The old website stays online during this time.

Icon: start switch

Launch and run

Go-live with a redirect plan on the agreed cut-over date, then ongoing support and further development as needed.

// 3

Your team in the project

Who from your company is involved, and when

A DXP project cannot be fully outsourced. The effort for your team can still be planned if it is clear early on who is needed when. Most of the work falls to marketing, which leads the project on your side. It is involved in every phase, with the biggest chunks in content and approvals.

Management is needed at the start, when goals and budget are set, and again to approve the design and the go-live date. IT is needed in the analysis, because it knows the existing systems, interfaces and single sign-on, and later for hosting and the switch of DNS and domains.

Purchasing negotiates the quote and contract with us before the project starts. Your company signs the Storyblok licence directly with Storyblok, and we help you choose the right plan. For the time after launch, support is added, and depending on what was agreed in advance, the hosting contract too. Specialist departments and product management supply and check content, especially during migration. The editorial team joins with the training and fills the last pages in the weeks before go-live.

The switch happens on a planned cut-over date that you help decide.

Schedules rarely fail because of technology. More often, things get stuck on issues that a little lead time would avoid. New texts, images and product data are finished later than planned because they are created alongside day-to-day business. So clarify early which content is carried over and which is rewritten, who supplies it and by when.

In a group of companies, approvals add to this. If design and content go through several levels, possibly up to the parent company, that takes time. Agree in advance who approves and how long an approval may take.

That leaves new requests in the middle of implementation. Every additional requirement shifts the schedule, even if it looks small. Collect such ideas in a list for after go-live. A DXP is designed to grow step by step, and the list then becomes the plan for the first extensions.

// 4

To forward

The short version for your colleagues

For management, IT and purchasing in five points:

  • The existing website stays online until go-live. The new one is built on a separate test environment, and the switch happens on an agreed cut-over date.

  • The project runs in seven phases: analysis, content model and design, development, migration, training, go-live and operation. Where the frontend is hosted and who holds the hosting contract is agreed with Werbeberg before the project starts.

  • A redirect plan for all old URLs protects Google rankings and existing links.

  • On our side, marketing leads the project. Management, IT and purchasing are needed at clearly defined points.

  • The biggest schedule risks are late content, long approval chains and new requests during implementation.

One-pager as a PDF (in German)

Project roadmap to pass on

All seven phases on one page: what happens, who from your company is needed and which preparation pays off. For everyone involved in the decision internally.

DXP with Storyblok

What would your migration look like?

Werbeberg has been a Storyblok Certified Partner since 2017. See how we build digital experience platforms.

Free initial meeting

Discuss your project plan

Planning a relaunch and want to know what the process would look like for you? In an initial meeting, we look at your existing website and clarify which phases need the most attention in your case.

No obligation and no preparation needed.

Planning a relaunch and want to know what the process would look like for you? In an initial meeting, we look at your existing website and clarify which phases need the most attention in your case.

No obligation and no preparation needed.

More articles to read