Construction Progress Tracking Challenges: Why It Breaks Down and How to Fix It
By Kush Agarwal, President and CEO, YelloSKYE. Published .
Executive Summary
- The Question: Why does construction progress tracking keep failing on real projects, even when teams work hard at it, and what fixes each failure?
- The Answer: Tracking usually breaks in five places: subjective percentages, stale reports, inconsistent capture, work hidden before it is recorded, and evidence split across too many tools. Each has a specific fix, and they work best together: agreed definitions of done, a short capture cycle, a fixed capture routine, mandatory records before closing up, and one shared home for site evidence.
- The YelloSKYE Angle: Construct IQ gives a project one dated, visual record from drones, 360° walks and fixed cameras, and DEX compares it with the schedule so status is read from evidence rather than estimated.
Construction progress tracking breaks down for a handful of predictable reasons: the figures depend on who made them, they arrive after the site has moved on, the evidence behind them is gathered differently each time, some work is covered before it is recorded, and what does exist is spread across too many places to check. None of these is solved by working harder at the same routine. Each needs a change to how evidence is gathered and who can see it.
This guide takes the five failures one at a time: how to recognise each on your project, why it happens, and what fixes it. It is written for project managers, PMOs and owners' representatives who suspect their progress figures are less reliable than they look. Where it helps, we show how Construct IQ, YelloSKYE's construction monitoring solution, deals with the same problem.
The five failures at a glance
| Failure | How it shows up | Root cause | The fix |
|---|---|---|---|
| Subjective percentages | Two people give different figures for the same floor | Status is an estimate made from memory | Written completion rules checked against visual evidence |
| Stale reports | Meetings argue about conditions that have already changed | Long gap between observing and reporting | A short capture cycle and reports generated from the capture |
| Inconsistent capture | This month's photos cannot be compared with last month's | No fixed routine for where, how and when to record | Fixed routes, viewpoints and settings, kept for the whole job |
| Hidden work | Nobody can prove what is behind a finished ceiling | Recording is optional and happens after closing | A mandatory capture step before any element is covered |
| Too many tools | Status has to be rebuilt by hand before each review | Evidence lives in chats, spreadsheets, inboxes and phones | One shared record linked to the schedule and existing tools |
1. Percentages that depend on who is asking
What it looks like
The site engineer reports level 4 partitions at 70 per cent. The drylining contractor's claim says 85. The project manager, after a quick look, thinks 60 is fair. All three walked the same floor in the same week. The meeting spends twenty minutes on the number and none on what to do about it.
Why it happens
A percentage for a trade on a floor is a judgement compressed into one figure. It depends on what the person saw, which rooms they skipped, what they count as finished, and, for a contractor, what they hope to be paid. Nothing is wrong with judgement, but when it is the only input, there is no way to settle a disagreement except by another visit. We cover the mechanics of how figures drift in our guide to percent complete in construction; the point here is the tracking failure it causes, which is that status cannot be checked by anyone who was not there.
The fix
- Write down what complete means for each key activity, in terms you could see on a photograph: "boarded both sides and taped", not "substantially done".
- Base every figure on a dated visual record of the zone, so a reviewer can open the evidence and reach their own view without travelling.
- Separate the person who claims from the person who confirms. The contractor can propose a figure; someone else checks it against the record.
On YelloSKYE projects, captures are aligned with the planned schedule so planned versus actual is shown by zone, and the same record is used for contractor billing verification. Our aerial progress capture use case describes a weekly drone programme where payments were released on visual evidence and payment disputes fell to zero.
2. Reports that describe last week's site
What it looks like
The weekly report goes out on Friday, built from walks on Monday and Tuesday. By the time the owner reads it, a pour has happened, a crew has left and a delivery has slipped. Decisions are made, confidently, against a site that no longer exists. The most expensive version is a trade crew booked into an area that is not ready for them.
Why it happens
Every manual step between seeing and sharing adds delay: walking, writing up, transferring figures into the programme, compiling the report, waiting for the meeting. On a busy site those steps are the first to slip, so the gap grows exactly when the site is changing fastest.
The fix
- Shorten the cycle between capture and review. Set the capture interval by the pace of the work, not by the reporting calendar.
- Let the record produce the report, so the figures are read from the latest capture instead of being retyped.
- Use continuous capture for the zones that move fastest. Fixed cameras record selected areas at set intervals all day, which shows activity between scheduled site-wide captures.
- Set alerts for the events that matter. With Construct IQ, reaching a key milestone triggers a notification, and so does a stop in activity in any priority zone, which means the news does not wait for the next report.
3. Capture that changes every time
What it looks like
The project has thousands of photos and still cannot answer a simple question: when did the east stair core reach roof level? Each week's pictures were taken by a different person, from different spots, of whatever caught their eye. Comparing two dates means hunting for two photos that happen to show the same thing.
Why it happens
Capture is treated as a habit rather than a procedure. Nobody defined the route, the viewpoints, the camera height or the cadence, so each person improvises. Staff changes make it worse, because the routine leaves with the person who carried it.
The fix
- Define the capture plan at mobilisation: zones, routes, viewpoints, cadence and who is responsible for each.
- Keep the settings constant for the whole job. A drone flown on the same route and height each time produces maps that line up; a 360 walk that starts at the same point on each level produces views you can step through by date.
- Write the plan down so it survives staff changes, and check after each cycle that every zone was covered.
The payoff grows with scale. In our portfolio monitoring use case, every site was captured with identical drone, 360° and fixed-camera parameters, so leadership could go through 10+ sites in half an hour. Our guide to construction photo documentation covers what is worth photographing as the build moves through its phases.
4. Work that disappears before anyone records it
What it looks like
A leak appears above a finished ceiling a month before handover. The question is simple: which services run through that void, and were they installed as drawn? Nobody can answer without opening the ceiling, because the last person to see the void was the installer.
Why it happens
Progress tracking tends to follow what is visible. Services in slabs, insulation behind cladding, waterproofing under screeds and fixings inside walls are complete for only a short time before the next trade covers them. If recording is optional, it is skipped at exactly that moment, because the next trade is waiting.
The fix
- Make capture before covering a hold point in the programme. No ceiling is boarded, slab poured or riser closed until the space has been recorded.
- Record the whole space, not a single detail. A 360 walk or laser scan of the full room shows routing and position; a single close-up rarely does.
- Pin each record to its location and date, so it can be found from the plan years later, not only from a folder name.
- Keep these records with the handover package. They are as useful to the people who maintain the building as to the people who built it.
The warehouse operator handover use case shows the long-term value: drones and 360° cameras logged MEP placements across the facility before they were enclosed, and maintenance teams later used that history to locate utilities hidden behind concrete.
5. Progress spread across too many tools
What it looks like
Photos are in three chat groups and on a dozen phones. The programme is in scheduling software, quantities are in a spreadsheet, issues are in emails and the drawings are in a document system. Before every progress meeting, someone spends half a day pulling these together, and the result is out of date before the meeting starts.
Why it happens
Each tool was adopted to solve one problem, usually by a different person. None of them was chosen as the place where progress lives, so evidence collects wherever it was easiest to put it at the time.
The fix
- Choose one home for site evidence. Every capture, from every sensor, lands in one place with a date and a location.
- Connect it to the tools you already use rather than replacing them. Progress data should flow into the scheduling and document systems without retyping.
- Retire the side channels for evidence. Chat groups are fine for conversation; they are not a record.
In DEX, drone maps, 360° walks, fixed-camera images and BIM models sit in one time-linked environment, and Construct IQ connects to Procore, Autodesk Construction Cloud and Autodesk BIM 360, so progress data and flagged deviations reach the team's existing workflows without anyone typing them in again.
How the failures feed each other
These problems rarely appear alone, and they compound. Inconsistent capture makes subjective percentages impossible to check. Scattered tools make reports slow, so they go stale. Stale reports encourage people to estimate rather than look, which makes figures more subjective. Hidden work is never captured because nobody owns a routine that would have caught it.
That is why a single fix often disappoints. A new reporting template does not help if the evidence behind it is irregular. A drone programme does not help if its results sit in a folder nobody opens. The repairs work as a set: a fixed routine produces comparable evidence, one record makes it findable, a short cycle keeps it current, and agreed definitions turn it into figures everyone accepts.
Which failure to fix first
Start with the failure that is costing you most. These symptoms point to the likely cause.
| Symptom on your project | Most likely failure | First step |
|---|---|---|
| Long arguments over percentages in progress meetings | Subjective percentages | Agree visible completion rules for the five most disputed activities |
| Crews arriving at areas that are not ready | Stale reports | Shorten the capture cycle in the zones that feed the next trades |
| You cannot say when a milestone was actually reached | Inconsistent capture | Fix routes, viewpoints and cadence, and assign an owner per zone |
| Ceilings or walls opened up to answer questions | Hidden work | Add a capture hold point before every covering activity |
| A day lost before each review assembling information | Too many tools | Name one home for site evidence and link it to the schedule |
| Payment claims accepted or rejected without evidence | Subjective percentages and stale reports | Time a capture just before each claim period closes |
If you recognise most of the table, begin with capture consistency. Every other fix depends on having evidence that can be compared from one date to the next. Our guide to automated construction progress tracking explains how much of the routine can then run without manual effort.
A one-month plan to repair your tracking
- Week 1: audit the current state. List where progress evidence lives today, how old the latest record is for each zone, and which activities caused disputes in the last three payment claims.
- Week 1: define complete. For the activities that caused disputes, write completion rules that a reviewer could check from a photograph or scan.
- Week 2: fix the capture plan. Set zones, routes, viewpoints, cadence and an owner for each, and add capture hold points before every covering activity in the next lookahead.
- Week 2: choose one home. Decide where all site evidence will live and how it connects to your schedule and document tools.
- Week 3: run the new cycle. Capture, review and report from the same record, and time one capture to the next payment claim.
- Week 4: check the result. Ask whether two reviewers now reach the same status from the same evidence, and whether the latest record for every zone is less than a week old. Adjust the plan where the answer is no.
Frequently asked questions
What are the biggest challenges in construction progress tracking?
The common ones are percentages that rest on one person's judgement, reports that are already out of date when they are read, capture that changes from week to week, work that is covered before anyone records it, and progress scattered across too many tools. Most projects suffer from several at once, and each makes the others worse.
Why are construction progress reports so often wrong?
Usually not because anyone is careless. Reports are built from estimates made on a walk, typed up later and passed through several people, so they carry the reporter's view and lag the site by days. Without a dated visual record behind the figure, nobody can check it quickly, so errors survive until a payment or a handover exposes them.
How do you track progress on work that will be covered up?
Make capture before closing a mandatory step in the programme. Before a ceiling is boarded, a slab is poured or a riser is closed, record the space with a 360 camera or laser scan and tie the record to its location. That gives you proof of what was installed, and a reference for maintenance teams later.
Is it a problem to use several apps for progress tracking?
The number of apps matters less than whether progress has one home. Photos in chat groups, figures in spreadsheets and issues in email mean someone has to rebuild the picture before every meeting. Pick one record for site evidence and connect it to the scheduling and document tools you already use.
How can you tell whether your progress tracking is working?
Ask three questions. Can two people reach the same status for the same zone from the same evidence? Is the latest record less than a week old? Can you find what a space looked like on any past date in under a minute? If any answer is no, you know which part to fix.