Why Frontline Workers Don't Use the New App (and How to Fix Adoption)
The app went live, the training happened, and the crew is still on paper. The usual causes of poor frontline software adoption, how to spot each one in your data, and the fix that works for technicians, site crews and plant staff.
When a crew doesn't use a new app, it's rarely stubbornness. It's usually one of five practical problems: the app is slower than the habit it replaces, training came too early, training didn't match the role, the phone or the site gets in the way, or the old way is still accepted. Each one shows up differently in your data and needs a different fix. Retraining everyone the same way, the default response, only fixes one of them.
This guide is for whoever owns the rollout: the operations lead at the site, or the customer-success team at the company that sells the app.
First, look at the numbers by site and role
Before you change anything, pull three numbers from the app for each site and each role:
- Usage share: jobs, permits or checks done in the app ÷ all jobs done.
- Rework: submissions sent back for missing photos, wrong asset or skipped steps.
- Coverage: how many people in the role have logged in and completed the task at least once.
Adoption problems are almost always local. One site at 90% and another at 20% tells you the app isn't the problem. Something about the second site is.
Cause 1: the app is slower than the old way
How it shows up: people log in, then drift back to paper or WhatsApp for the same task. Usage share starts fine and falls.
Watch someone do the task in the app and time it against the old way. Frontline staff are rational: if closing a job takes 14 taps and a photo upload over weak signal, and the paper form takes 30 seconds, paper wins.
Fix: shorten the path in the app (defaults, fewer required fields, offline drafts), or make the old way unavailable for that task. Don't retrain. They already know how.
Cause 2: training happened too long before the work
How it shows up: coverage is low. Plenty of people never completed the task at all, even though attendance at training was high.
A session two weeks before go-live is mostly forgotten by the first real job. This is the most common cause, and the cheapest to fix.
Fix: put short, task-specific material on people's phones and re-send it at the moment of use: "Closing your first job today? Here's the two-minute version." A tap-through walkthrough of the real screens works well because it's quick and it looks exactly like the app.
Cause 3: training didn't match the role
How it shows up: one role lags behind. Supervisors approve fine, but technicians skip steps. Rework is high for that role specifically.
General product tours teach everyone a bit of everything. A technician needs their four tasks in the order they do them, not the reporting screen.
Fix: one walkthrough per role, per task. If you need to build them quickly, the free walkthrough maker turns screenshots into a tap-through in a few minutes. See the role-by-role example for what that looks like.
Cause 4: the phone or the site gets in the way
How it shows up: a site or shift lags with no obvious training gap, and complaints mention logins, cameras or signal.
Common culprits: shared phones with someone else logged in, no data in the plant room, cameras that won't focus on QR tags in low light, password resets nobody can do on night shift.
Fix: a site walk with the supervisor, phone in hand, doing the real task in the real place. Fix logins and connectivity before touching training.
Cause 5: the old way is still accepted
How it shows up: usage share plateaus well below 100% and stays there.
If a supervisor still signs paper permits "just this once", everyone learns that the app is optional.
Fix: agree an end date for the old process with site leadership, announce it, and stick to it. Make the supervisor the person who checks app usage on their crew. In most stalled rollouts this matters more than any training material.
A 30-day adoption routine
| When | Do this |
|---|---|
| Week 1 | Daily: usage share and coverage per site. Visit or call the lowest site |
| Week 2 | Find each lagging site's cause (one of the five above) and fix that, not everything |
| Week 3 | Re-send the role walkthroughs to anyone who hasn't completed a task yet |
| Week 4 | Old process ends. Weekly check from here on; new joiners get the walkthroughs on day one |
What not to do
- Retrain everyone at once. It costs a day of production and fixes only cause 2.
- Blame the crew. "Resistance to change" is usually one of the five causes wearing a disguise.
- Measure attendance. People attending training tells you nothing. People completing the task in the app tells you everything.
For the full plan around go-live, see the software rollout training plan and how to train field workers on a new app.
Frequently asked questions
Why don't frontline workers adopt new software?
Usually for practical reasons rather than resistance: the app is slower than the paper or WhatsApp habit it replaces, people were trained too long before they needed it, the training didn't match their role, the phone or connectivity on site gets in the way, or supervisors still accept the old way. Each has a different fix, so find which one you have before retraining everyone.
How do you measure software adoption for field teams?
Use the app's own data, per site and per role: the share of jobs, permits or checks completed in the app versus the old process, how many submissions are sent back for missing information, how long tasks take in the app, and how many people have logged in at all. Compare sites; adoption problems are almost always local.
How long does it take for frontline workers to adopt a new app?
For a well-run rollout, most crews are doing their core tasks in the app within two to four weeks. If a site is still below half after a month, something specific is blocking it, often a supervisor still accepting paper or a step that is genuinely harder in the app.
Does retraining fix low adoption?
Only when the cause is that people don't know how to do the task. If the app is slower than the old way, or nobody checks whether it's used, more training changes nothing. Find the cause first, then retrain the specific task for the specific role that's struggling.