Web GIS delivers geographic data and analysis through the browser — maps as applications rather than documents. A web GIS stack pairs spatial databases and services on the server with interactive mapping clients on the front end, letting thousands of users view, query, edit and analyse shared authoritative data without installing anything. It is how modern organisations actually consume GIS: dashboards, portals, field apps and embedded maps.
The standard architecture
Under every web map sits a familiar stack: a spatial database (PostGIS, enterprise geodatabase) holding authoritative data; a services layer (GeoServer, ArcGIS Server, cloud APIs) publishing it as tiles and feature services; and a client (MapLibre, Leaflet, OpenLayers, ArcGIS JS) rendering and interacting in the browser.
Around that core: authentication and roles, catalogues for discovery, editing workflows with validation, and integration APIs tying maps to the systems where business actually happens.
What distinguishes web from desktop
Desktop GIS optimises for the analyst: deep tools, heavy data, single user. Web GIS optimises for the organisation: one authoritative version, controlled access, zero-install reach, real-time currency. Analysis migrates too — server-side geoprocessing, cloud notebooks — but the web tier’s first job is trustworthy shared truth.
The two are complements: production and hard analysis in desktop and pipelines; consumption, collection and coordination on the web.
Why portals succeed or die
Failed portals share a biography: launched around an event, loaded once, never refreshed, quietly abandoned. Succeeding ones are owned products — data pipelines keep layers current, performance is engineered (tiles, caching, indexes), users see their workflows (not the GIS team’s), and governance assigns every layer a custodian.
The technology is mature; sustainability is organisational. Budget the operating model, not just the build.