Latest · Interpreted · Published 11 Sep 2026 · GA Applications editorial

If the tool only watches, the tool gets abandoned.

Construction field technology fails when it extracts data from workers and returns nothing to them. The tools that survive make the worker’s day visibly easier: today’s schedule without a phone call, photos that replace re-typed reports, proof that protects them in disputes. The value-exchange map below tests any proposed tool: what the worker gives, what they receive, and when.

Read the exchange map Describe the field task

The exchange pattern is drawn from GA field-tool practice and labelled inference where it generalises beyond it.

Recognition

Why field tools die in week two

The pattern repeats across the industry. Management buys visibility. The crew experiences surveillance plus extra taps: the app needs signal the site does not have, fights gloves, demands typing on a ladder. Within a fortnight the work is back on paper and camera photos, and the office calls it an adoption failure.

Adoption is not a training problem; it is an exchange problem. A worker who feeds a system and gets nothing back is doing unpaid data entry for someone else’s dashboard. People stop. A worker who feeds a system and immediately gets something useful — the schedule, the record, the proof — keeps feeding it, because the tool is visibly theirs too.

So the question for any proposed field tool is not “what does management see?” It is “what does the worker receive, and when?”

Give, receive, when

The field-value exchange map

Six common data demands, what each must return to the worker, and what happens when the return never arrives.

The worker gives	The worker receives	When	If they receive nothing
Photos of finished work	A completed record — no end-of-day report re-typing	Immediately, on save	The camera app wins; the tool is dropped
A tap confirming a task done	Today’s schedule and job details without a phone call	At the start of each day	The whiteboard and group chat win
A measurement or reading	Instant validation that it is in range	At capture	Entries get invented to finish the form
A note about a problem	Confirmation someone saw it and acted	Same day	Notes stop within a fortnight
Location or time check-in	Proof of presence that protects them in disputes	Automatically, on arrival	Check-ins are forgotten or quietly resented
Accurate time on site	Correct pay without arguing from memory	Payday	Estimates replace records, and everyone loses them

Read any row whose third column is “never” as a surveillance feature, whatever the sales deck calls it. GA inference from field practice; test per site with the checklist below.

Constraints

Design for hands, weather and no signal

The exchange only works if the tool survives the environment:

Gloves and wet hands. Targets sized for the worst day, not the demo. One-thumb flows as the default.
No signal. Capture offline, sync when coverage returns. A field tool that needs reception has failed the first test of field work.
Sunlight and rain. Contrast and type sized for a screen viewed outdoors, at arm’s length, in a hurry.
No compulsory typing on site. Pickers, photos and defaults. Typing happens back at the ute, if at all.
Honest states. A queued item says queued, not sent. Field trust, once lost to a lying interface, does not come back.

These are the constraints GA designs to in its own field technology work. A field tool is designed for the worst fifteen minutes of the day, not for the boardroom walkthrough.

Before any rollout

The worker-value test

Ask the crew, not the vendor. Tick what the proposed tool genuinely offers.

0 of 7 complete

Every data demand has a visible return to the person entering itWalk the exchange map row by row.
Works with gloves, wet hands and one thumbTest on site, not in the office.
Works with no signal; syncs when signal returnsAirplane mode is the honest test.
Faster than the paper or photo habit it replaces, timed on siteIf it is slower, nothing else matters.
The worker can see their own records without asking the officeOwnership of their own data builds the exchange.
Failures are honest: a queued item says queued, not sentOne lie and the tool is done.
The crew helped choose it, or can veto it after a trial weekA veto you honour once buys adoption for a year.

Fewer than five ticks: the tool serves the office at the field’s expense. Expect abandonment, and budget for it.

The surveillance framing kills twice

A tool introduced as monitoring gets resisted, then gamed: invented entries, shared phones, forgotten check-ins. You lose the data and the trust in the same quarter. Introduce the exchange instead, prove the return in the first week, and let management visibility arrive as the by-product of a tool the crew actually uses.

Evidence honesty

Sources, method and what would change this

Inference
Exchange principle
The claim that field tools survive on worker-visible return is GA’s practitioner inference from field-tool scoping and rebuilds. Not a controlled study; test it per site with the checklist.
GA Applications editorial
Demonstration data
Offline and glove constraints
GA designs and operates field-connected tools under these constraints; they are observable in GA’s own field interfaces rather than cited from research.
GA Applications field practice
Verified fact
No vendor verdicts
This article offers a test, not a ranking. No product is named, recommended or criticised.
GA Applications editorial method
Proposal
What would change this conclusion
If crews in GA trials kept using tools with no worker-visible return, the exchange model would be revised. Row-level results belong in future dated updates to this piece.
GA Applications editorial

Bring the tool, or the problem

Describe the field task that is slow, invisible or disputed. GA maps the value exchange honestly — including whether the fix is technology at all.

Describe the field task See the field technology service

Sometimes the right answer is a better clipboard. The exchange map says so when it is.