Public organizations have documented QGIS adoption, ArcGIS migration and lower licensing expenditure. These are different claims. A defensible comparison identifies the organization, year, workload and exact financial category before drawing a lesson for a Latin American project.
The following primary sources were reviewed on September 23, 2026. They are external cases, not GeoSAT clients. Historical figures remain in their original currencies and periods; none is a current vendor quotation or a promised saving.
Four cases with different evidence
Dominican Republic: a documented ArcGIS to QGIS transition project
The IGN-JJHM's 2024 institutional report describes a geographic database project begun in the second quarter, centralizing geographic, cartographic and geodetic information. It includes transition from ArcGIS to QGIS as the institution's principal tool and an open-source satellite imagery server. This documents project scope; it does not prove every dependent application was replaced or quantify net savings. See printed page 39, PDF page 43 of the official 2024 report.
Colombia: Cornare documents QGIS in geoprocessing
Cornare's software review for fiscal year 2025, dated March 4, 2026, identifies QGIS for geoprocessing in its installed-software description. That is evidence of operational adoption in a Colombian environmental authority. It does not supply an ArcGIS replacement count or a before-and-after cost model. See page 3 of Cornare's software report.
United States: St. Joseph County reports a bounded savings figure
The county's 2021 GIS report, dated January 7, 2022, describes QGIS replacing ArcGIS as its standard GIS software and reports USD 7,500–10,000 in annual licensing savings. Its 2022 work list still includes continuing the transition. The figure concerns licensing; a complete transition budget and net operating saving are not established. See the county's one-page report.
United Kingdom: Barrow reports publishing and catalog savings
The LGA case, first issued in 2011 and updated in February 2012, describes GeoServer and GeoNetwork for publishing and metadata discovery. It reports GBP 26,000 annually in license and software cost savings. The case does not identify every displaced proprietary component as Esri. It is evidence for open publishing architecture, not a universal ArcGIS price comparison. See the original LGA case.
Compare what the sources actually establish
| Case | Usable evidence | Not established by that evidence |
|---|---|---|
| IGN-JJHM | Named transition project and intended principal desktop tool | Full completion, retired licenses or net savings |
| Cornare | QGIS use for geoprocessing | Replacement percentage, productivity or savings |
| St. Joseph County | Migration and annual licensing figure for the reported period | Total ownership cost or universal price per seat |
| Barrow | Publication/catalog architecture and reported software savings | All-Esri predecessor stack or current market price |
A project can use QGIS while retaining ArcGIS services. It can reduce licenses while increasing spending on operations. It can publish open data without opening all internal datasets. Treat these as separate questions when preparing a proposal or interpreting a headline.
Turn a public case into a local investigation
The following is GeoSAT's proposed evaluation method, not a claim that the cited institutions performed these exact steps.
- Match the workload. Identify whether your task is desktop analysis, an internal map, a public download, field capture or a network operation. A publication case is weak evidence for replacing specialized editing.
- Find the removable expense. Locate the actual contract line and its cancellation or renewal date. A published foreign licensing amount says nothing about the discount or bundle your organization holds.
- Price the missing work. Estimate conversion, application changes, training, support and coexistence. Mark unpublished details as unknown instead of filling them with a percentage from another case.
- Reproduce one important task. Use a representative layer, the real consuming application and a documented expected result. Include permission and recovery tests, not only a screenshot.
- Measure the new operation. Record support hours, storage, failed runs and elapsed user time. Separate salary capacity from cash payments that actually disappear.
For a Colombian authority evaluating QGIS, the transferable question from the Dominican report is how a central geographic database and analyst workflow fit together. For a public portal, the useful architectural question is how publication and discovery are operated. Neither implies copying another institution's staffing assumptions or claiming compliance with another country's requirements.
A worksheet for published savings
Create one row per financial claim with source URL, publication date, period covered, original currency, cost category, before amount, after amount, transition cost and retained capabilities. Use “not published” where needed. A range must remain a range; replacing USD 7,500–10,000 with its high endpoint silently exaggerates certainty.
Keep a separate column for your interpretation. “Licensing decreased” may be supported while “five-year return is positive” remains untested. If a case omits labor, you cannot label its percentage as total cost of ownership. Use the migration cost method for your own calculation.
What a GeoSAT implementation case would need
A publishable implementation should identify the permitted scope, starting workflow, changes made, measured period, retained systems and result. Financial claims need matching cost categories and a stated treatment of internal work. Performance claims need dataset size, versions, request type and environment. The client must authorize publication of identifiable material.
The municipality and utility exercises are explicitly fictional. Their arithmetic is useful, but they are not substitutes for a completed project reference. Use the case collection to compare the kinds of evidence, then select a bounded pilot with the phased plan.