Are web maps getting worse?

Helmut Barz argues that #GoogleMaps is suffering from #enshittification, as a profit-driven pivot toward #AI-derived map data and a focus on lucrative urban markets leaves rural areas stranded in data deserts while the community that built the map is abandoned. He makes the case for reclaiming geographic data as a decentralised #publicgood.
Author

Ralph Straumann

Published

June 30, 2026

I couldn’t find conclusive evidence about this supposed phenomenon, but there is “anecdata”1 that Google Maps data may have worsened, at least in some parts of the world, over the last few months:

My2 search was triggered by a LinkedIn article written by Helmut Barz: When the algorithm suggests free fall: The enshittification of Google Maps. Helmut lists some examples from his experience as well as some cases of bad navigation (data or functionality) that made headlines:

It used to be different. For more than a decade, Google Maps was my personal gold standard of navigation across Europe and the US. But lately, the system’s failures are piling up, and we are long past the point of funny anecdotes.

Helmut attributes this to the process of enshittification3 and a shift from painstakinlgy collected ground truth to stronger reliance on data derived from satellite imagery using AI or GeoAI4. From Helmut’s post:

The focus today is on paying logistics companies, Uber drivers, delivery services, and urban agglomerations. That is where the ad money flows, and where data harvesting is lucrative. If you live in the countryside, drive a camper outside the metropolises, or seek nature on a bicycle, you fall through the cracks. Rural areas are becoming victims of “digital redlining” — they are degenerating into data deserts.

The most cynical part? We, the users, made Google Maps great. We fed the map for years with Point-of-Interest (POI) data, photos of restaurants, and hazard warnings for free. Now, the very community that built the system’s wealth is being abandoned.

What might the ad revenue be here?

The article goes on to point to some alternatives: (Google-owned) Waze for car navigation and Komoot for cycling. On that latter platform, however: The popular sports planning platform Komoot has been sold to private equity last year. In his article When we get Komooted, Josh Meissner gives an insightful account of the sale of Komoot and the often disappointing mechanics behind similar community-powered services.

My main take-away from the first post is the data politics perspective offered by Helmut:

But the medium and long-term solution must be to reclaim geographic data as a decentralized public good. (…) It’s time to stop feeding free data to a system that wants to send us — literally and metaphorically — over a cliff. Let’s rather put our data where our interest is — and use it for the common good.

I would add as another pillar: official data, sustainably, responsibly, and transparently collected in guaranteed homogeneous and suitably high quality.

Footnotes

  1. Anecdotal evidence.↩︎

  2. Admittedly: brief.↩︎

  3. “Enshittification” is a term coined by writer Cory Doctorow for the mechanisms prevalent specifically in two-sided online platforms and services. Think, for example, web search which serves two sides of the market with – often – conflicting interests, namely the (end-)users and (enshittification says: over time, more and more) the advertising customers.↩︎

  4. The application of AI on geospatial data and for answering spatial questions.↩︎