QField vs ArcGIS Field Maps: compare the complete cycle

The right field application preserves work and delivers usable data on return. Compare capture, project distribution, synchronization and device support. An attractive form in a connected demonstration does not establish operation in a disconnected region.

The test that changes the decision

Give two crews the same fictional batch with a large attachment, a value domain and a parent-child relationship. Test airplane mode, unexpected closure, retries and concurrent changes. Record loss, duplication and reconciliation minutes. If the team depends on specific relationships or services, retain that dependency until the alternative is demonstrated.

Compare equivalent costs

Request the complete current cost and an alternative estimate with the same scope, availability and volume. Include retained licenses, infrastructure, support, internal hours, training, integration and coexistence. Do not use another country's public list price as a Colombia or LATAM quotation. Taxes, discounts and purchased entitlements can change the result.

When to retain and when to proceed

Retain the current component when an essential capability lacks a proven replacement or transition costs exceed expected benefits. Proceed when the pilot completes the task, an operating owner exists and TCO remains acceptable under a less favorable scenario. Data ownership and provider choice also matter, but express them as concrete requirements: exports, formats and recovery times.

This comparison is an editorial evaluation guide, not a parity certification or benchmark. Check versions, licensing and documentation when defining procurement scope.

Compare two complete operating arrangements

On the QField side, name QGIS, QField, the packaging or cloud synchronization method, device management and office review. On the Field Maps side, name the ArcGIS deployment, maps, feature services, offline configuration and user entitlements. Comparing a free mobile installation with a fully operated platform produces misleading economics.

A row labeled “inspect” is intentional. Exact capabilities depend on product release, storage backend, configuration and device. The table identifies what must be proven; it is not a claim that every combination supports every workflow.

When QField is a strong candidate

QField is worth piloting when the organization already authors in QGIS, uses portable data, can define its synchronization process and wants control over project configuration. A bounded inspection program with exclusive work assignments can be a manageable first use. The QField guide describes a 20-asset exercise with a related inspection table, forms and attachments.

It can also make sense when several organizations need to exchange a reproducible project without assuming the same commercial desktop environment. That benefit depends on a complete package: project, data, symbols, fonts and licenses. A QGIS project referencing an author's local directories is not a portable field solution.

If the team lacks operations capacity, compare a managed QField arrangement with the existing service. Open software does not require self-hosting every component. Ask what the support offer includes: project preparation, synchronization monitoring, backups, device troubleshooting and conflict resolution are different responsibilities.

When retaining Field Maps may be reasonable

Retain the current arrangement while evaluating alternatives if it is tightly integrated with critical ArcGIS feature services, forms, identity or downstream review applications. Rebuilding those integrations may cost more than the mobile application itself. Identify the actual dependency rather than declaring the entire platform inseparable.

A stable existing operation also has training and accumulated configuration value. Measure whether a proposed move preserves field productivity and office review quality. A new application that reduces license expense but increases reconciliation work can have a higher total cost, as the fictional utility scenario illustrates.

Do not retain or replace solely because an application has more listed features. A feature matters when it supports a required task and works in the actual environment. Similarly, do not use a missing optional feature to reject a candidate that fully satisfies the required operation.

Run the same two-crew pilot

Use the synthetic lab and select 20 assets with stable identifiers. Prepare equivalent forms and required evidence in both candidate arrangements. Keep assignment size, devices, crew training, basemap detail and network conditions as comparable as practical. Record unavoidable differences rather than hiding them.

1. Preparation: a second operator prepares the package or downloadable area using written instructions. Record time and undocumented dependencies.
2. Offline start: crews download, enter airplane mode, close the app and reopen it. Merely losing connectivity after the map has loaded is insufficient.
3. Normal capture: each crew creates five inspections with a required value and a photograph. Include one asset with a second visit.
4. Validation: attempt a missing required value, an out-of-domain code and an incomplete record that must remain draft.
5. Conflict: both crews change the same parent attribute in a disposable test assignment. Record which value appears centrally and how the operator sees the conflict.
6. Interrupted return: interrupt a transfer, retry and then submit the same package or changes again. Check missing records, duplicate inspections and attachment completeness.
7. Office review: a reviewer queries one record, accepts another and retrieves the correct photo from a second workstation.
8. Recovery: simulate loss of the working copy after a preserved return and reconstruct the usable project without overwriting the original test evidence.

This is a proposed evaluation protocol. No comparative speed, reliability or savings result is claimed until it has been executed on the selected product versions and devices.

Measure accepted work, not only upload speed

Define a completed record as one that passes the operation's actual acceptance rules, including attachments and review where required. Count submitted, accepted, rejected, duplicated and unresolved records separately. Their relationship should reconcile with the test batch. A successful synchronization indicator alone does not prove the information is usable.

Measure preparation hours, field minutes per accepted inspection, office review minutes, conflict-resolution time, data transferred and recovery time. Record battery and storage observations under comparable conditions if these constrain the operation. Avoid extrapolating a short indoor demonstration to an entire rural workday.

For positional quality, use the agreed reference and receiver procedure. Compare measured results with the required tolerance and retain the relevant metadata. Neither application's brand establishes survey accuracy. The Field Maps high-accuracy guidance illustrates the additional receiver and location-profile responsibilities on that side of the comparison.

Calculate the cost per accepted inspection

A useful operational measure is:

cost per accepted inspection =
  (annual software + services + devices + support + field labor + office review)
  / accepted inspections in the same period

Use consistent allocation for devices and shared staff. Keep one-time migration, training and parallel operation separate, then include them over the selected decision horizon. The denominator must be accepted work, not raw submissions; otherwise a process that creates duplicates or incomplete records may look artificially cheap.

For QField, include any selected cloud subscription or managed service, synchronization operation and project maintenance. For Field Maps, use the organization's actual entitlements, service costs and storage, accounting for agreements shared with other workflows. The price of an individual component is not necessarily the amount removable from the budget.

Stress-test the estimate with slower review, a higher photo volume and a longer coexistence period. Use the TCO calculator for the overall migration and keep this per-inspection measure as a separate operational view. A lower cost is meaningful only if the accepted work and required service level remain comparable.

Migration decisions after the pilot

Can QField work entirely without mobile internet?

An appropriately prepared project can support offline work, but the complete package, basemap, forms and return method must be tested. Direct dependence on an online database or remote resource changes that answer. Verify a cold start and full working session in airplane mode.

Does cloud synchronization eliminate conflict decisions?

No synchronization label replaces the organization's rule for incompatible observations or edits. Inspect what the chosen mechanism detects or overwrites and determine who makes the final decision. Reducing overlapping assignments is often useful, but it does not remove the need to rehearse exceptions.

Specific references: QField cable packaging, QField forms, Field Maps offline preparation, and Field Maps forms.

Sources and documentation

• QField documentation
• ArcGIS Field Maps documentation

Next step

Continue in the Open GIS collection. For a specific project, use the total-cost calculator and request an assessment.

Related reading

• ArcGIS Online vs self-hosted or managed Open GIS
• GeoServer vs pygeoapi vs QGIS Server: which to use