The problem: we can't draw the line anymore

Ten years ago, the question hardly came up. Either you wanted a website, or you wanted software. Today, the web has gained so much functional depth that the line between "site" and "software" has become blurry. A site can have a member area. Software can run in a browser. A blog can embed a simulator. A newsletter can generate personalized PDF on the fly.

Hence the constant ambiguity: is my project a site or a web app? And behind this question, budget gaps of 1 to 10.

This article gives a simple 5-question test to decide, illustrated with real cases.

The 5-question test

For each question, answer yes or no. Count the "yes" answers at the end.

1. Will your users have a personal account?

Not a contact form, not a newsletter — a real account with:

  • A username and password (or SSO)
  • A personal space they access after login
  • Data that belongs to them and that they find on every visit

A WordPress site can technically add accounts via a plugin. But as soon as accounts become a central element (10% of user time spent logged in), you're in web app territory.

2. Will your users enter and store business data?

Not filling out a contact form whose result is sent by email. Entering structured and persistent data:

  • Creating product entries, contact entries, appointments
  • Completing a record, tracking the progress of a request
  • Filling out a questionnaire whose answers are kept and processed

If your users will produce structured content that your system has to process, you need a business database — it's a web app.

3. Is there business logic specific to your activity?

Logic that exists nowhere else: a validation workflow, a specific quote calculation, a matching algorithm, a scoring system, an approval path, a recommendation engine.

If you can describe your business logic by saying "like [some well-known software/site]", you're probably still on a site with some custom. If you have to explain in detail because nobody does it the same way, you're on a web app.

4. Will multiple users interact with each other through the platform?

Exchange messages, see others' actions in real time, collaborate on a document, take part in a shared game, apply to a job offer, respond to a public request.

This interaction implies: data shared between users, notifications, fine-grained permissions (who can see what, who can do what), sometimes real-time.

5. Does the interface need to transform significantly depending on the user?

Not just "show the first name" and "personalize the dashboard". A real transformation:

  • The menu, available options, accessible features differ by role
  • An admin sees X, a client sees Y, a subcontractor sees Z
  • The central content (the "displayed data") is driven by what the user has entered/done

If the experience of a logged-in user is radically different from a public visitor, you're on a web app.

The decision rule

  • 0 to 1 yes → you're on a site. WordPress (classic or Headless + Next.js) is plenty. Budget: €2,250 to €8,000, delivery 3-8 weeks.
  • 2 yes → gray zone. Probably a Headless site with a bit of applicative logic (advanced form, simulator, light member area). Budget: €5,000 to €15,000, delivery 6-10 weeks.
  • 3 yes or more → web app. WordPress is no longer the right tool. Stack Next.js + PostgreSQL + custom admin. Budget: €15,000 to €80,000+, delivery 2 to 6 months depending on complexity.

Concrete examples

Case 1 — Site, no hesitation

A law firm wants a new site that presents the team, specialties, news and a contact form for appointments. 0 out of 5 yes. Optimized classic WordPress is enough, budget €2,250 (Classic tier).

Case 2 — Headless site

An independent media outlet publishes 15 articles per month, wants maximum SEO score, a well-structured editorial blog, themed dossiers. No reader accounts, just a newsletter. 0 out of 5 yes but maximum performance need. Headless WordPress + Next.js, budget €4,000 (Headless tier).

Case 3 — Gray zone

A consultant wants a site that presents his offers AND an eligibility simulator for a tax scheme that takes 12 parameters and generates a personalized PDF. 2 out of 5 yes (specific business logic, but no accounts or user interaction). Headless site + Next.js with an applicative module, budget €6,000 to €9,000.

Case 4 — Pure web app (Panorama Pub)

An editor wants to launch a B2B directory of promotional product suppliers. Suppliers create accounts, manage their listings. Buyers search, compare, contact. The admin drives everything. 4 out of 5 yes. Custom web app, Next.js + PostgreSQL, autonomous admin. Delivery 2 months.

Case 5 — Custom mobile app (Hermitage)

A woodland estate wants a mobile treasure hunt for visitors: geolocation, contextual riddles, scores. No accounts (progress is local), but exclusively mobile and offline usage. 2-3 out of 5 yes + strong mobile need. PWA Next.js, local persistence, native geolocation.

A question nobody asks

What if I still hesitate after the test?

Three principles to decide in the gray zone:

  1. Start small. A web app can almost always start with an MVP that looks like a site with some custom. If the project takes off, evolve. If not, you haven't spent €50k.
  2. Avoid defensive over-architecture. "In case we need…" is the main source of budget waste on web projects.
  3. Think data model. If you can describe your business data model in less than 5 minutes (entities, relations, rules), a web app makes sense. If you struggle, it's probably because you're on a site with a light functional layer.

Going further

For a personalized project diagnostic: run the form or book a call.