Construction management in Excel: a project and job-cost tracker that works, where it breaks, and what replaces it
Most small contractors run their jobs on construction management in Excel. A programme in one tab, a budget in another, a list of subcontractors, a change log that is mostly emails. For a builder running a few jobs at a time with one person keeping the numbers, a well-built spreadsheet is enough — and it is more flexible than a lot of software sold to them.
Then the jobs overlap, a second site manager needs to update the tracker, a client disputes a variation, and the job that looked profitable in March turns out to have lost money in September. It goes wrong in the same five places every time.
This is the honest version: a construction project tracker in Excel that is worth copying, exactly where each part of it fails, and what your options are when it does.
The construction spreadsheet worth copying
Most free construction management Excel templates are a Gantt chart with a budget column bolted on. They track what you planned to spend and what you have spent, and miss the number that decides whether a job makes money: what you have already committed to spend. Start with six small sheets instead.
Sheet 1 — Jobs. One row per job.
| Job | Client | Original contract | Start | Status |
|---|---|---|---|---|
| J-2107 | Harlow Dental | 180,000 | 2026-03-02 | On site |
| J-2112 | Mercer House | 94,500 | 2026-05-18 | On site |
Sheet 2 — Cost codes and budget. One row per job per cost code. This is your estimate, broken down the way you will track it.
| Job | Cost code | Description | Budget |
|---|---|---|---|
| J-2107 | 03-100 | Concrete and formwork | 22,400 |
| J-2107 | 09-200 | Drylining and ceilings | 18,750 |
Sheet 3 — Commitments. Every purchase order and subcontract, the moment it is agreed — not when the invoice arrives.
| Ref | Job | Cost code | Supplier / sub | Committed | Status |
|---|---|---|---|---|---|
| PO-588 | J-2107 | 03-100 | Northside Concrete | 19,800 | Open |
| SC-041 | J-2107 | 09-200 | Kestrel Interiors | 17,200 | Open |
Sheet 4 — Costs. Every invoice, timesheet and material receipt, each against a cost code and, where there is one, the commitment it draws down.
| Date | Job | Cost code | Against | Amount | Type |
|---|---|---|---|---|---|
| 2026-04-11 | J-2107 | 03-100 | PO-588 | 9,900 | Material |
| 2026-04-12 | J-2107 | 03-100 | — | 1,640 | Labour |
Sheet 5 — Change orders. One row per variation, with a status. Nothing goes into the contract value until it is approved.
| CO | Job | Description | Amount | Status | Approved on |
|---|---|---|---|---|---|
| CO-07 | J-2107 | Additional steel lintel, rear wall | 3,850 | Approved | 2026-04-20 |
| CO-08 | J-2107 | Upgrade to acoustic ceiling | 6,200 | Pending | — |
Sheet 6 — Daily log. One row per site per day: weather, who was on site, what was done, what went wrong, and a link to the photos.
The numbers that matter are then derived, never typed:
Revised contract = Original contract + SUMIFS(CO[Amount], CO[Job], job, CO[Status], "Approved")
Committed to date = SUMIFS(Commitments[Committed], Commitments[Job], job, Commitments[Cost code], code)
Spent to date = SUMIFS(Costs[Amount], Costs[Job], job, Costs[Cost code], code)
Projected final = MAX(Committed to date, Spent to date) + uncommitted remaining work
The commitments sheet is the one most spreadsheets skip, and the reason jobs "suddenly" go over budget. By the time an invoice arrives, the money was spent weeks ago when the subcontract was signed. Tracking commitments turns an overrun you discover at the final account into one you see the day you sign.
Where it breaks
1. Change orders happen on site, not in the spreadsheet
A client asks for an extra door on a Tuesday walk-round and the site manager agrees. The work is done, the invoice goes out, and the client says they never approved it. In the spreadsheet the variation is either missing, or it went straight into the contract value while it was still only a conversation. Real change management keeps a variation pending until it is priced and approved, and records who approved it and when, so the argument at the final account is about the price, not whether it happened.
2. Nobody knows which drawing is current
The spreadsheet can say "Rev C issued 14 April", but the drawing itself lives in an email attachment, a shared folder and the back of a van. When the architect issues Rev D, half the trades are still building from C. Drawing revision control — one current version, old ones kept and marked superseded — is something a spreadsheet can only describe, never enforce.
3. Every trade sees every trade's prices
The moment you share the tracker with a subcontractor so they can update their progress, they can see every other subcontractor's contract value, rates and retentions. So you don't share it. You copy updates across from emails instead, and the tracker is always a week behind the site. A shared project where rival trades cannot see each other's contracts is not something a spreadsheet can do.
4. The site doesn't have a laptop
Daily logs, photos and snag lists are captured on phones, often with poor signal. A spreadsheet on a phone is miserable to use and useless offline, so the diary gets written up on Friday from memory and the photos stay on someone's camera roll. When a delay claim or a defect dispute needs evidence of what was on site on a particular day, it is not there.
5. The job cost is always last month's
Costs reach the spreadsheet when invoices are keyed in, which is weeks after the work. Labour comes from timesheets someone types up. By the time the cost report is complete, it describes a job that has moved on, and it disagrees with the accounts because the two were updated at different times from different copies of the same invoices.
What a real one has that a spreadsheet doesn't
We keep a hand-counted inventory of what construction management products actually ship, built from 60 products across 120 official source pages — the full list, with a test for each feature, is in construction management software features:
| Capability | Products (of 60) |
|---|---|
| Reports, scheduled distribution and exports | 55 |
| Tasks, issues and field coordination | 54 |
| Drawings, markups and revision control | 46 |
| Procurement and purchase orders | 45 |
| Mobile and offline jobsite access | 44 |
| Document repository with version custody | 43 |
| Budget, job cost, cost codes and variance | 42 |
| Commitments and subcontracts | 37 |
| Users, roles, companies and scoped permissions | 36 |
| Change orders, approvals and impact | 34 |
| Daily logs, diaries and site conditions | 32 |
| Master schedule, Gantt and critical path | 31 |
Read it against the five failures and two things stand out. First, the Gantt chart — the thing most people picture — is near the bottom, at 31 of 60. The capabilities at the top are the ones that keep a job coordinated: reports that go to the right people, tasks and issues, drawings under revision control, and access from the site itself.
Second, look at the permissions row: 36 of 60 products document users, roles and companies. In construction, access is not just per person but per firm, because a project is shared by businesses that are also competitors. That is failure 3, and it is the part of a construction system that is hardest to bolt on later.
Your three options, honestly
Keep the construction spreadsheet
Correct for a small builder: a few jobs at a time, one person running the numbers, subcontractors who send updates by phone. Use the six sheets above, and treat the commitments sheet as the most important one. It will do more for your margins than any software you do not yet need.
Stop when you cannot share the tracker without exposing prices, when variations start arriving faster than you can price them, or when the cost report is always a month behind.
Buy software
Correct for most contractors past that size. Construction management software gives each firm its own access, keeps drawings under revision control, and takes daily logs and photos from phones. If the ledger side — retentions, subcontractor tax deductions, payment applications — is the harder problem, see construction accounting software, which covers where generic accounting stops.
The real cost is fit and setup. These products are powerful and take real configuration, and many are priced for larger firms. Check what you pay per project or per user before you commit, and whether your subcontractors will actually use it. If you are starting on a free plan, free construction management software covers what each kind of free costs and how to test one.
Build it
A coding agent can build a construction project system in days. It will build the Gantt chart and the project dashboard first, because those are what it has seen. What it tends to skip is what the table above says matters: commitments against cost codes, change orders with an approval state, per-company access, and drawings with a current revision. When four AI models built the same construction spec, keeping one firm's contracts away from another's was the defining test.
If you build it, build these first
- Jobs, cost codes and budgets, so every cost has a home the moment it is incurred.
- Commitments before costs. Purchase orders and subcontracts recorded when agreed, with invoices drawing them down.
- Change orders with states — pending, approved, rejected — and nothing reaching the contract value until it is approved.
- Access per company, so each subcontractor sees their own work and contract, and nothing of anyone else's.
- A daily log from the phone, with photos attached to the day and the job, that works with a poor signal.
Drawing management, scheduling, estimating and integrations with accounting are real scope, and worth adding once the money side is right.
If you are building with a coding agent, the Construction Management planner is the spec we hand ours: the job-cost chain from estimate to commitment to actual, change orders with approvals, and access that keeps rival trades apart on a shared project.