Your Field Software Isn't Broken — Your Rollout Was
Your Field Software Isn't Broken — Your Rollout Was
I bought Jobber in March of 2022. Sent my two launch techs a credentials email that afternoon. Thought that was the rollout. Six weeks later, both of them were still writing jobs on paper and transcribing them at the end of the day.
I'd paid the subscription. Watched the onboarding video. What I had not done was change any behavior — mine or theirs. The login email was not a rollout.
If your techs aren't using the app, this is your story too.
Why Techs Don't Use the App (And It's Not Because They're Bad With Tech)
The default explanation is generational. Older guys don't like phones. They learned on paper and that's just how they are. I've heard this from owners with 25-year-olds on their crew, which should tell you something about how accurate it is.
The real explanation is simpler: you handed out a login and called it done.
The app showed up with no changed workflow, no reason to use it today specifically, and no consequence for skipping it. So the clipboard — which your tech already knows, which has never crashed, which doesn't require cell service in a basement — stayed the rational choice.
I recognized this the same way I recognized my intake failure in month two of running Reeves Electric. I had planned for everything: the pricing model, the CRM, the van graphics, the website. What I hadn't planned for was the gap between "tool exists" and "behavior changed." The problem wasn't the app. It was me. I had no rollout. I had an email.
What a Real Rollout Actually Looks Like
After my NPS hit a 4 in month two, I didn't fix intake by reading about intake. I fixed it by answering every call myself for ninety days and taking notes on what was actually breaking. The behavior became automatic because I forced it — on myself, every day, no exceptions — until it was.
Same logic applies here.
The structure is ninety days because two weeks is enough time for a tech to white-knuckle through something they hate and revert the moment you stop watching. Ninety days is enough time for the app to feel slower than the clipboard on day one and faster than the clipboard by day sixty. That's actually what happens once the muscle memory builds.
Weeks one and two. Pull the app usage report from your platform — Jobber has activity reporting, Service Fusion has a tech performance dashboard. Establish where you actually are. No consequences yet. This is your baseline number. You need it before anything else matters.
Weeks three through six. Required use on every job close. The tech cannot mark a job complete without closing it in the app. GPS check-in required at job start. Photos required on any job above $400. No paper backup. If a job isn't in the app, it doesn't exist for dispatch and invoicing. This is where the real friction surfaces and where you learn what's actually broken.
Weeks seven through twelve. Full workflow in-app: job notes, parts used, customer signature, photo attach. Deviation is a documented conversation. Second deviation in the same period is a formal write-up. The consequence structure is real, stated upfront, and applied consistently.
Check the usage data every week. Not monthly. Weekly, as a line item on the numbers review you're already running.
This Isn't a Training Problem, So Stop Buying Training
When adoption stalls, most owners call the vendor and buy the onboarding package. Or schedule a lunch-and-learn. Or send the crew a YouTube link. I've done all of these. None of them moved adoption more than about 10% because I was solving the wrong problem.
Your techs can figure out the app. It's a phone, not a panel upgrade. What they can't do is prioritize the app over the clipboard when there's no reason to and no consequence for skipping it.
If there's no consequence for skipping the app, skipping the app is the rational choice. That's not resistance. That's logic.
Owners feel like they've rolled out the software because they paid for it and talked about it in the crew meeting. That feeling is not data. Pull the usage report. Look at who's actually logging jobs and who isn't. Until you've done that, you have a guess, not a rollout.
The variable that's missing in most failed rollouts isn't better training materials. It's a consequence structure that makes the app the path of least resistance. Same reason the call recording habit stuck at my shop — I tied it to the Friday numbers review, attached it to booking rate, and made skipping it more work than doing it.
What Actually Happened When I Ran This for Real
When I added trucks three and four in 2023, I ran the protocol instead of the credential-email version. It mostly worked. It also broke in ways that told me exactly where the resistance actually lived.
About three weeks into the required-close phase, one of my techs had perfect app compliance on paper. Every job closed, GPS logged, photos attached. Then my dispatcher mentioned she'd seen him sitting in the parking lot in the truck for ten minutes after jobs. I pulled the timestamps. He was closing jobs in the parking lot — logging the GPS check-out and uploading a photo he'd taken inside before leaving — so the report looked clean without changing how he actually worked at the job site.
The resistance wasn't about the phone. He clearly knew how to use it. The friction was in the mid-job documentation. Specifically, the photo requirement at the job site. We talked about it. The issue was upload time on jobs where he was in a concrete basement with one bar. Real problem, fixable — I added a wifi-upload-only setting and the behavior changed within a week.
The bigger fix was getting my dispatcher into it. I was checking the reports myself, which meant I was checking them when I had time, which meant inconsistently. She already runs the Friday call recording review with me. I added app compliance to that same checkpoint — she flags anything that looks off during the week, we look at it together Friday, I have the tech conversation Monday. Same meeting, nothing new added.
That structure, more than anything in the written protocol, is what made it stick.
What the Software Actually Has to Do
A forced-use protocol will fail if the app creates more friction than the clipboard. Before you run the 90 days, be honest about whether your platform clears that bar.
The minimum: mobile job creation, GPS check-in and check-out, photo attach, digital customer signature, parts and labor line items, and a status push to dispatch that doesn't require a separate action. If your techs have to take three taps to log a status change and one to write it on the paper sheet, you will lose.
For a five-truck residential service shop, the mobile experience gap between Jobber or Service Fusion and the cheaper alternatives is real and it matters. The cheaper apps' field UX is noticeably worse, and you're asking techs to change their behavior on the back of that experience every single day. Start with a platform whose field app was actually built for field use.
The other thing I see constantly: shops that have layered in too much software and blame the rollout for the confusion. Your Zapier or Make automations need to be invisible at the field level. If your tech is seeing error notifications from a webhook, or being asked to take a step that only exists because of something you built in the backend, something's misconfigured. The tech's job is to run the job. When that line gets blurry, he stops trusting the app and goes back to the clipboard — and honestly, that's a reasonable response to a broken setup.
What to Do Monday Morning
Here's the sequence. Specific tasks, not concepts.
Monday of week one: Pull the usage report. Find out what percentage of your jobs last month were fully logged in-app from check-in to close. That number is your baseline. Write it down.
Tuesday of week one: Have the conversation with your techs. Not a training session. Not a lunch. A short, direct meeting: starting in two weeks, every job is closed in the app or it doesn't exist for invoicing. Here's what that means. Here's what happens if it doesn't. I tell my guys straight — no softening, no "we're going to try to move toward." The expectation is clear or it isn't an expectation.
Week two: Let them use the app with no consequences while you watch the data daily. Note every friction point. Fix the ones you can fix before enforcement starts. Sync issue? Solve it now. GPS burning battery on an older phone? Solve it now.
Week three: Enforcement starts. Jobs not closed in-app get flagged same day by your dispatcher, not by you. You have the follow-up conversation with the tech, but your dispatcher is the daily eyes. Add a compliance line to your weekly numbers review — it sits below booking rate and average ticket, because it affects both.
Month two: Full workflow required. Compliance below 90% in any given week is a documented conversation. Hold the line.
Month three: The behavior is default by now. You'll still have the weekly line item. You'll spend thirty seconds on it instead of ten minutes.
The Monday action is the usage report. Pull it before anything else.
Hard Cases
My techs are older and genuinely not comfortable with smartphones.
The timeline stays the same. What changes is weeks one and two — skip the YouTube link, skip the vendor video. You or your dispatcher sits next to them in the truck and walks through a job close together. Do it in person, more than once. Get comfortable with the app before enforcement starts, then run the same structure.
My best tech refuses to comply and I can't afford to lose him.
Before you treat it as defiance, have the friction conversation. Every time I've heard "he just won't do it," there's been a real friction point in how that tech specifically works that nobody asked about. Ask first. If it turns out he understands the requirement, can do it, and won't — that's a management problem, not a software problem. Making an exception tells every other tech the requirement isn't real. You know where that ends.
We tried this six months ago and it fell apart after week three.
That's when the owner is still checking compliance personally and life gets busy. A couple of jobs slip through unchecked, and within two weeks the structure has quietly dissolved. The fix is handing daily flagging to your dispatcher before week three starts. If you're the single point of attention on this, it fails the first week you're slammed.
How do I know if the problem is actually bad software rather than a bad rollout?
Run a real 90 days — baseline pulled, expectations stated clearly, consequences applied consistently. If you're still below 70% compliance at day ninety and the failures are spread evenly across your whole crew, the platform friction is real. If it's one or two techs, that's a people issue. Either way, you need the 90 days of data before that question is answerable.
Enjoyed this article?
Get articles like this in your inbox every Monday. Free, no spam.