How a construction SME should choose ERP software

Every construction ERP demo shows the same screens: a board, a BOQ, a dashboard with a rising line. They are demonstrating the software. You need to know something else, which is what happens on a wet Tuesday when a supervisor with 4 percent battery has to record 18 workers.

Six questions that actually separate systems

1. Can I import my BOQ as it is?

Not "we support Excel import". Ask them to import a real BOQ of yours during the demo, with your column layout and your section structure. If that becomes a services conversation, every future project will too.

2. Does a site cost land on a BOQ line?

If costs post to a project but not to a line, you get project-level margin and no cost control. Ask to see the path from a labour entry to a variance report on one specific line.

3. What does the supervisor screen look like?

Ask for the field app, not the web app, and count the taps for a daily attendance entry. Then ask what happens with no signal. These two answers predict your data quality more reliably than anything else in the demo.

4. Who can see finance?

Role scoping needs to be specific enough that a site engineer sees quantities but not rates, and a client sees their own project only. Vague answers here become a reason not to put real numbers in the system.

5. What happens when we outgrow the plan?

Ask exactly what occurs at the project or user limit. The good answer is a prompt to upgrade with nothing lost. The bad answer is a feature quietly stopping, or data becoming read-only.

6. Can we get our data out?

Export in a usable structure, not a PDF pack. Ask before you sign, because the answer never improves after you have four years of history inside.

What matters less than the brochure suggests

• Module count. You will use four to six modules. The other twenty are a maintenance surface.
• Dashboards. Every product has them and none of them are the reason a rollout succeeds.
• AI features, unless they remove a specific task you can name today.
• Integrations you have no current plan to use.

Run a real pilot

1. 1

   Pick one live project

   Mid-sized, currently running, with a supervisor who will tell you the truth. Not the flagship and not the disaster.

2. 2

   Import the actual BOQ

   If this takes more than an afternoon, you have learned something important.

3. 3

   Run four weeks of real logging

   Attendance and progress, every day, from the phone. Do not run the old process in parallel or nobody will use the new one.

4. 4

   Compare one report by hand

   Budget vs actual for the month, worked out manually, against the system output. Differences are the pilot findings.

5. 5

   Ask the supervisor, not the office

   The office will tell you the reports are nice. The supervisor will tell you whether this survives contact with a site.

Commercial terms worth negotiating

• Month-to-month for the pilot period, annual only after it succeeds.
• Onboarding included, with a named person and a defined scope.
• Local billing if your revenue is not in the vendor currency.
• A written export commitment, in a structured format, on request.