Guide · Construction
Eight automations that take reporting out of Excel at companies with 30–150 employees: where to start, how long each takes and how they link up in six months.
In short
Updated September 2026 · 13 min
Someone asks for last month's profitability report, and the finance team spends three days reconciling numbers from five Excel files. We see the scene at almost every construction or manufacturing company with 30–150 employees that comes to the construction solution. This guide describes eight automations that take reporting out of Excel, says which one to start with and how they link up into a complete system in six months.
Excel works well at ten employees and becomes a problem between 30 and 50, when the files multiply and the report depends on one person. Excel is not a bad program; it is a good tool for one-off analyses and small-scale calculations. The problem starts when it becomes the main data management system of a company that is growing.
The cycle has four stages and repeats almost identically from one company to the next. At ten employees a few files are enough: one person keeps them, everyone knows where things are, reporting takes a couple of hours. Between 30 and 50 employees, departments, projects and new data sources appear, and the files multiply.
That is when the parallel versions appear (“Final_Report_v3_ACTUAL_Final.xlsx”), the formulas broken by an added column and the email chains trying to reconcile two different numbers. Errors are not an exception but a property of the material. In a study of 50 operational spreadsheets, between 0.8% and 1.8% of formula cells contained errors, and some of them changed the final result (Powell, Lawson and Baker, 2008).
The breaking point comes when reporting takes days, decisions wait for data and a single person knows how the master file works. One wrong formula propagates into every report linked to it, and an audit becomes a search through folders. From here there are two roads: you link the systems together, or you put ever more people on maintaining files.
The Excel chaos cycle
At ten employees, a few files and one person who keeps them; reporting takes a couple of hours.
Between 30 and 50 employees, new departments and projects appear; files, versions and reconciliation emails multiply.
Above 50 employees, reporting takes days, decisions wait for data and a single person knows how the master file works.
Either the existing systems are linked and the reports produce themselves, or ever more people maintain files.
Automation links the systems you already have, accounting, timesheets, purchasing, so the data reaches the reports on its own, without manual copying. We are not proposing replacing your infrastructure or a two-year ERP (Enterprise Resource Planning, the integrated management system) implementation. Fewer than half of the enterprises in the European Union use an ERP, 46% in 2025 (Eurostat, 2025), and reports are usually still built from exports into Excel.
Modern automation for a small or medium company means four things. The existing systems are connected, so data flows without copying. Manual reports are replaced by dashboards that update themselves.
The reports are accessible to everyone who needs them, without email attachments. And the alerts come when something needs attention, not in the report at the end of the month. Anyone who wants a list of costs and steps finds it in the business process automation guide.
What automation is not: enterprise software costing hundreds of thousands of euros, replacing employees with artificial intelligence, or an implementation that takes years. Nor is it a one-size-fits-all solution that ignores how your company works. The goal is simple: the repetitive data work does itself, and the team deals with decisions.
Reporting in Excel
Automated system
The eight automations cover project costs, stock, receivables, labour, sales, compliance, suppliers and the executive report; each one replaces a specific Excel file. The implementation times in the table are our 2026 estimates, for a company whose source data already lives in accounting, timesheets and purchasing. Without them, the first weeks go on cleaning data, not on automation.
| # | Automation | Who it is for | The Excel problem | Implementation (our estimates, 2026) |
|---|---|---|---|---|
| 01 | Project profitability | construction, manufacturing | costs live in three files and in accounting; the margin is seen monthly | 2–3 weeks |
| 02 | Stock and goods movements | distribution, manufacturing | warehouse stock does not match accounting | 3–4 weeks |
| 03 | Cash and unpaid invoices | any company | payment status is updated by hand; the forecast is optimistic | 1–2 weeks |
| 04 | Timesheets and labour cost | companies with shifts or project work | overtime shows up only at payroll | 2–3 weeks |
| 05 | Sales across channels | companies with distributors, retail, online | every channel comes with its own export | 3–5 weeks |
| 06 | Compliance and e-invoicing | construction, manufacturing | documents are hunted down before an inspection | 4–6 weeks |
| 07 | Supplier performance | companies with dozens of suppliers | late deliveries are remembered, not measured | 2–3 weeks |
| 08 | Executive dashboard | management | the management report gathers two-week-old data from five sources | 4–6 weeks |
Material costs sit in a purchasing file, labour hours in another kept by the project managers, and subcontractor invoices in a third, at accounting. Once a month someone gathers them into a profitability report, and by the time the margin is visible, the project is already over budget.
The automation brings costs from the ERP or purchasing, hours from timesheets, invoices allocated to projects and revenue from accounting into one place. The margin is calculated daily, and the projects trending over budget are visible while you can still intervene.
Stock levels are updated after a count, movements are kept in separate files per warehouse, and the value in accounting does not match what is on the shelf. Discrepancies come to light months later, when too much or too little has been ordered.
The system keeps stock across all locations on one screen, classifies movements (sales, transfers, waste, returns) and reconciles automatically with accounting. The alert comes when stock drops below the reorder point, not when the customer is waiting for goods.
Issued invoices sit in a spreadsheet, payment status is typed in by hand when the money arrives, and the cash forecast starts from optimistic assumptions that are never updated. By the time management sees the report, the big delays are already problems.
The automation links invoicing to the bank statement: a screen of receivables, alerts at overdue thresholds and a forecast based on customers' real payment behaviour. It is the fastest of the eight and the only one that applies to any company.
Who worked where and on which project is written on timesheets and then typed into Excel. Overtime shows up only in the payroll report, too late to adjust, and labour cost per project does not match the real hours.
Digital time tracking, linked to projects, allocates cost automatically by project, department or cost centre and warns when someone approaches the hours ceiling. The same data feeds the labour law compliance check: breaks, maximum hours, overtime limits.
When you sell through direct sales, distributors, retail and online, the data comes from different systems: the CRM (Customer Relationship Management, the customer database), the e-commerce platform, distributor reports. Someone combines them in Excel, product analysis takes hours of pivot tables, and sales team performance is visible weeks after month end.
The system collects data from all channels daily, calculates margin per product and keeps customer segmentation current, based on what they actually buy. The team sees its results daily, not at month end.
Electronic invoicing through RO e-Factura, reports to the Labour Inspectorate, site safety records and reporting to ANAF (the Romanian tax authority) are often kept in separate files. The result is a scramble before an inspection and the risk of a fine for a missing document.
The automation integrates invoice submission, checks working hours against the legal limits, tracks safety incidents and keeps documents with versions and renewal reminders. The audit trail builds itself, transaction by transaction.
A growing company works with dozens of suppliers, but orders are kept in Excel or in email, and delivery performance is remembered (“I think they are usually late”). Quality problems are discovered after the fact, without data to back up the complaint, and prices are compared by hand at renewal.
A purchase order system tracks every delivery, calculates on-time delivery per supplier, links quality problems to goods receipt and produces the scorecard before the negotiation. The data decides who stays a supplier, not memory.
The managing director assembles the picture of the company from five sources: finance, the project managers, sales, human resources and treasury. By the time the management report is compiled, the data is two or three weeks old.
The executive dashboard consolidates the key indicators on one screen: revenue, costs, margins and cash position, the health of active projects, the sales pipeline, headcount and turnover. Custom alerts flag what needs immediate attention. It is the last automation on the list precisely because it feeds on the others.
The first automation is one of project profitability, stock and receivables: their source data already lives in systems and the result shows within weeks. The order is not by importance but by the ratio between effort and what becomes visible. Project profitability is the biggest pain in construction and manufacturing, stock in distribution, and receivables matter to everyone and are the fastest to implement.
The second phase adds timesheets, sales across channels and compliance: high impact, but they need other teams involved and cleaner data. Suppliers and the executive dashboard come when needed or at the end. The dashboard, in particular, only makes sense once there is something to link.
The practical rule is one automation at a time, taken all the way, with the matching Excel file closed at the end. Two started in parallel means two teams in transition and no report you can rely on. What happens to the team in the first weeks after the file is closed is described in the Valley of Despair.
The recommended order
In month 1 you implement one quick win, in months 2–3 one or two more automations, and in months 4–6 the executive dashboard links them. Month 1 is for a single automation from the first three, implemented completely: data connected, automated report, the team used to opening the screen instead of the file. It ends when the matching Excel file is closed, not when the dashboard is ready.
In months 2–3 you add one or two more, chosen by what hurt most in the first month. Each one starts faster than the first, because the links to accounting and timesheets already exist.
In months 4–6 the automations are linked to each other, project costs with cash, timesheets with profitability, and the executive dashboard shows them on one screen. After six months comes continuous improvement: new data sources, refined reports, new areas as the company changes. The pace is the company's, not the vendor's: a peak-season month is skipped, not forced.
Six months, four stages
One quick win from 01–03, implemented completely, with the matching Excel file closed.
One or two more automations, on the links already built.
The automations are linked to each other; the executive dashboard shows them on one screen.
Continuous improvement: new sources, refined reports, new areas.
A construction company with 80 employees and six active sites replaces the monthly profitability report with a margin calculated daily; the company is fictitious. Before, material costs came from the ERP, labour hours from paper timesheets typed into Excel by the site managers, and subcontractor invoices from accounting, two weeks after goods receipt. The profitability report came out in the third week of the following month.
At the Ploiești warehouse, the 12% margin from the bid had fallen to 4% without anyone noticing until the third month. The cause was mundane: two subcontractor invoices booked to another site and a month of overtime that did not appear in the file. Nobody had made a mistake; the data was in three places.
The implementation took three weeks. Material costs stay in the ERP, timesheets moved to the phone and are allocated per site, subcontractor invoices are linked to the project when booked, and revenue from accounting. The margin is calculated every night, for every site.
In week 36, Ana M., the finance director, opens the warehouse's profitability file and sees a single problem: two subcontractor invoices not allocated to the project. She allocates them in five minutes and the margin corrects the same day. Next month's report no longer exists; the screen does.
Automating reporting assumes a volume that justifies the effort and source data you can rely on; in three situations it is not the next step.
Pick one idea from the first three and count, this week, the hours the matching Excel file eats today: who opens it, how long they stay, who waits for it. That number is the argument, not the percentages in presentations. Once you have it, the business process automation guide shows the steps, the costs and what happens in the first four weeks.
With a single one from the first three: project profitability if you are in construction or manufacturing, stock if you are in distribution, receivables if you want the fastest result. The test is simple: the source data already lives in a system, and the Excel file it replaces eats hours every week. The rest come once the first is closed.
Not Excel, but Excel as a system of record. One-off analyses, a bid simulation or a quick calculation stay in spreadsheets. What disappears is the file into which data is copied from other systems to produce a report; that one is closed once the matching automation works, otherwise the data lives in two places.
A single automation takes between one and six weeks, by our 2026 estimates: receivables are the fastest, compliance and the executive dashboard the longest. A complete system, with three or four linked automations and the executive dashboard on top, is built in about six months. Dirty source data lengthens every stage.
An internal owner, usually the finance director or the operations manager, who owns the reports and decides what changes. They do not need an IT department: modern platforms run in the cloud, update themselves, and the technical side stays with the vendor. The one thing they cannot delegate is the rule: the replaced Excel file is never reopened.
The first conversation is about your process: who does what, on which documents, where time is lost.
One email a month, only when we publish. Nothing else.
Corning, Volvo and Cox Automotive automated their data flows to decide on today's numbers. A 50–100 employee company has the same problem, at a different scale.
Read →The four chapters (A–D) of the Romanian as-built file, who is responsible for each, what is usually missing at handover and how to build it as you go.
Read →Material approval (MAF / submittals): what the packet contains, the technical office → supervisor → designer → client circuit and where time is lost.
Read →