TL;DR
- A construction firm was burning 9 hours per day on admin transcription across 12 sites — pure re-entry of paper clipboard data.
- We replaced clipboards with a Power Apps canvas app connected to SharePoint, with offline support for unreliable site signal.
- A Power Automate flow watches the list and sends immediate Teams alerts for any High or Critical issues — average response time dropped to 4 minutes.
- Data lag dropped from 26 hours to zero. Transcription time dropped to zero.
At a construction firm we worked with, every site supervisor carried a clipboard. At the end of each day they handed it to the office admin. She typed every entry into a spreadsheet the next morning. By Wednesday, the operations manager was looking at Tuesday's data. Decisions were being made on information that was 24 to 48 hours old.
That's not a technology problem. That's a process problem that technology can solve. Here's what we found when we looked closely, and what we built to fix it.
The Clipboard Problem in Numbers
Before proposing anything, we spent two days measuring the actual cost of the current process. The numbers were stark.
The office admin spent 45 minutes per site per day on transcription alone. Across 12 active sites, that was 9 hours of admin time per day on pure data re-entry — work that added no value beyond moving information from one medium to another. That's more than a full-time salary spent on transcription.
Error rate from handwriting misreads: 8.4% of entries contained at least one error when we checked a sample of paper records against the typed spreadsheet. Incorrect headcounts, wrong material quantities, missed issue flags. Small errors that compound over time into bad project cost estimates.
Data lag: average 26 hours between an event occurring on site and that event appearing in the operations manager's spreadsheet. The operations manager put it plainly: "We're managing projects in the rearview mirror."
The daily cycle: paper clipboard on site, admin re-entry the next morning, operations manager working from data that was already 26 hours old.
What We Built
A Power Apps canvas app connected to SharePoint lists. Supervisors open it on their phone or tablet on site. No special hardware. No custom-built app store submission. Power Apps runs through the free Microsoft Power Apps mobile app on both iOS and Android.
The form captures: date and time (auto-populated from the device — no manual entry), site (dropdown populated from a SharePoint list of active sites), activity type (dropdown), headcount, materials used (structured entry with quantity and unit), issues flagged (free text with a severity dropdown: Low, Medium, High, Critical), and photos (camera capture via the Power Apps camera control, uploaded directly to a SharePoint document library). Submit. Done.
Data appears in SharePoint immediately. The operations manager's Power BI dashboard refreshes every 15 minutes. By the time the supervisor walks back to the site office, their report is already visible in the dashboard.
We deliberately kept the form to 7 fields. The original request had 19 fields. Every field we removed was a point of friction for a supervisor standing in the mud in steel-capped boots. Short forms get filled in. Long forms get "I'll do it back at the office" — which means paper first, digital entry later, which means delay.
The Offline Problem — And How We Solved It
Mobile signal on construction sites is unreliable. Half-built buildings block signal. Basements and underground sections have none. Any solution that required a live internet connection would fail the moment a supervisor walked into a stairwell.
We built the app with Power Apps offline capability. When the device detects no network connection, form submissions save to a local collection on the device. A visual indicator in the app shows "Offline — data saved locally." When signal returns, the app detects the connection and automatically syncs the locally stored records to SharePoint. The supervisor doesn't manage this. It happens in the background while they're doing something else.
We tested the offline scenario deliberately during UAT: cut signal for 6 hours, submit 14 forms, restore signal. All 14 records synced to SharePoint within 30 seconds of reconnection. No data loss. No duplicates. The supervisors didn't notice anything different — which is exactly what good offline handling should feel like.
Offline mode saves data locally on device. Signal restored triggers automatic sync to SharePoint. Critical issues trigger an immediate Teams notification to the operations manager.
What the Operations Manager Got
A Power BI dashboard connected directly to the SharePoint list via the SharePoint Online List connector. Real-time headcount across all 12 sites. Materials usage versus budget by site and by material type. Open issues grouped by severity, with the site name and the time of submission. A timeline view showing when reports were submitted each day — which also doubled as a way to see if any sites had stopped reporting.
The first morning after go-live, she came in and said "I've been waiting 3 years for this." That's not a testimonial we asked for. It's the kind of thing people say when a tool actually solves the problem they've been living with.
The Number That Surprised Everyone
We expected the big win to be data speed. It wasn't. The biggest win was issues getting caught earlier.
Before, a site issue reported on paper on Tuesday might not reach the operations manager until Thursday. By then, what started as a medium-severity problem had sometimes grown into a costly delay. Two days of lag in a construction project can compound fast.
After go-live, a Power Automate flow watches the SharePoint list. Any new record with the Severity field set to "High" or "Critical" triggers an immediate Teams notification to the operations manager and the relevant project lead. The notification includes the site name, the issue description, and the supervisor's name. No hunting for context.
Average time from issue reported to management being aware: 4 minutes. That number came from testing, not from a sales pitch. We measured it.
The 8.4% error rate from handwriting misreads isn't unique to this client. It's typical for any process where handwritten records get typed into a system by someone other than the person who wrote them. Every re-entry step is an error multiplication point. The only way to fix it structurally is to eliminate the re-entry entirely.
Key Takeaways
9 hours per day of admin time was spent on pure transcription. That cost disappears completely when data enters the system once, at source, by the person who has it.
Offline capability isn't optional for field use — it's a requirement. Any mobile form that needs signal to submit will fail on a construction site at least 30% of the time.
The biggest ROI wasn't speed — it was issue escalation time. Catching a High-severity problem 4 minutes after it's reported vs. 48 hours later changes project outcomes.