Walk into any mid-market construction firm and ask the office team to show you how the business actually runs. You won't be shown software. You'll be shown spreadsheets.
That's not a criticism. Spreadsheets are the default tool for a reason. They're free, they're flexible, everyone in the office can use them, and they let you change anything at any time without asking permission. For most small operations problems, a spreadsheet is genuinely the right answer.
The trouble starts when a spreadsheet becomes the system of record for a process that's outgrown it. The spreadsheet that worked great when you were doing three quotes a week is the same spreadsheet you're now running twenty-five a week through, with three different people editing it from different devices and nobody quite sure which version is current.
Here are five spreadsheets most construction businesses I've come across are running, and where each one starts to break.
1. The quote and enquiry tracker
What it does well: gives the office somewhere to log enquiries and track which quotes are out. Anyone can add a row. Anyone can update a status.
Where it falls down: nobody updates the status. The “quoted” column is sometimes filled in, sometimes not. The “won/lost” column lags by weeks because the estimator forgets. The follow-up date isn't a real reminder, it's just a date someone typed. When the MD asks “where are we with that quote for the school job,” nobody can answer without checking with the estimator.
What it should be: a tool the whole office uses to actually move quotes through the chain. Every quote has a current status anyone can see. Follow-up reminders fire automatically. When a quote wins or loses, the win/loss rate calculates itself, rather than being something nobody ever has time to work out.
2. The variations log
What it does well: captures the fact that extra work has been agreed mid-job. (Variations is the industry word; extra work is the plain version.)
Where it falls down: variations get logged inconsistently. Some get a row, some don't. The “client signed off” column is sometimes ticked, sometimes left blank. The link between a logged variation and an invoice is a manual step that gets forgotten. By the end of a long job, nobody can confidently say which variations got billed and which didn't.
What it should be: a way to track extra work from “we've agreed to do this” through to “we've billed for this.” With approval status visible to the office, and a clear handoff to invoicing built in. The test isn't whether variations are listed somewhere. The test is whether you can answer, on any given day, how much agreed extra work is currently unbilled.
3. The subcontractor compliance tracker
What it does well: lists who's on the approved books, with columns for insurance dates, accreditations, and contact details.
Where it falls down: dates go stale because nobody chases them. Insurance certificates are in a folder somewhere on the shared drive, in three different versions, none of which clearly match the row in the spreadsheet. When a subcontractor is needed at short notice, half an hour gets spent verifying that their paperwork is current. Sometimes it isn't, and a decision gets made anyway.
What it should be: a system that knows when each subcontractor's insurance and accreditations are due to expire, sends the chase automatically, and won't let an out-of-date subcontractor be assigned to a job without flagging it. The paperwork lives in the same place as the record of the paperwork.
4. The plant and equipment tracker
What it does well: usually exists as a whiteboard rather than a spreadsheet, which means anyone in the office can see it. That's a real advantage.
Where it falls down: nobody on site can update it. Equipment gets returned later than planned and nobody corrects the board. Hire costs roll over because nobody noticed the return date passed. The whiteboard is only as accurate as the last person who walked past it remembering to update it. And when something needs to be on hire for an extra week, the conversation about whether to extend or swap supplier happens in a WhatsApp message that's not on the board.
What it should be: a record the site team can update from a phone, with hire periods that flag automatically when they're about to roll over. The cost of each piece of equipment per day visible alongside the status, so the decision to extend or return is informed rather than reactive.
5. The snag and outstanding-items list
What it does well: captures the long tail of small things still to do at the end of a job.
Where it falls down: lives on whoever's laptop made it. Gets emailed back and forth in different versions. The contractor, the client, and the office all have slightly different copies. Some snags get fixed and never crossed off. Some get crossed off but never actually fixed. Final sign-off slips by weeks because nobody can confidently say which items are still outstanding.
What it should be: a single shared list everyone is looking at the same version of. Site team can mark items done from their phone. Client can see the live status. The list naturally shortens to zero and then the job closes.
The pattern across all five is the same. Spreadsheets work fine when one person is using them. They start to fail when the process they're tracking involves multiple people working at different times, often from different places. The spreadsheet can't show status to everyone who needs to see it. It won't enforce a step that needs to happen. It can only hold what someone remembered to type into it.
This isn't a case for ripping every spreadsheet out and replacing it with software. Some of these spreadsheets are doing their job and would cost more to replace than to keep. But not all of them are working. Some have status that's unclear. Some are letting things slip. Some have the office team doing chase work that no one was ever supposed to do. For those, there's a better shape.
The simplest test: pick the spreadsheet that frustrates your office team the most. Ask the person who maintains it how much time they spend each week keeping it accurate. If the answer is more than two hours, that spreadsheet has already outgrown itself. It just hasn't been replaced yet.
If you can already think of which spreadsheet that would be, that's the place to start. Get in touch.