How a corporate web project actually runs
Five stages from discovery to launch — which decision is made when, and how surprises are avoided.
Most delays in web projects are not technical, they are decisional. A decision made at the wrong time — or never made at all — can turn a two-week job into a two-month one. That is why we structure the process around the order of decisions.
1. Discovery
The goal is to understand the business, not the website. The questions we ask:
- Who is this site for, and what should it make them do?
- How will success be measured — form submissions, quote requests, brand perception?
- Which decisions have your competitors already made for you, and which are still open?
Output: a one-page scope and success-criteria document.
2. Information architecture
Structure before design. Which pages exist, how they relate, which path a user takes to convert. Skip this stage and the design comes out beautiful but ineffective.
3. Design
Visual language once the architecture is approved. Main templates first, then components. Building a system rather than designing pages one by one buys both consistency and speed.
4. Development
Design components become code. The performance budget and SEO structure are established here from the start — SEO added later is always a patch. More detail: how to build a fast website.
5. Launch and beyond
Launch is not an ending but the start of measurement. In the first month:
- Indexing checks in Search Console
- Real user performance data
- Verification that conversion points actually work
How to avoid surprises
Three rules are enough:
- Every stage produces a written output. Verbal agreement gets forgotten.
- Scope changes are not free — but they are not forbidden either. They just have to be visible.
- Content must be ready by the time development starts. This is the single most common cause of delay.