It is the first Monday of the month and the operations manager opens the same five spreadsheets she has opened every Monday for years. Revenue from the accounting file. Job hours from the timesheet system. Stock on hand from the warehouse app. Margins from a pivot table someone built in 2019 and nobody else fully trusts. By the time the numbers are side by side, it is Tuesday afternoon and the meeting is tomorrow.
That scene plays out in construction offices, wholesale warehouses, transport depots and manufacturing shops across Australia, with different spreadsheets and different headaches but the same underlying problem: the numbers that run the business live in separate systems and nobody can see them together until someone copies them into a workbook by hand.
A KPI dashboard is just a screen that pulls those numbers together automatically. The hard part is not the dashboard itself. It is getting the data from where it already lives into one place, reliably, every day. This post walks through what that looks like in six industries we work across, with the five to eight numbers each one actually watches and which system holds each one.
What makes a useful operations dashboard
Before the industry examples, a few things that separate dashboards people use from dashboards that sit untouched after the first week.
The numbers should come straight from the source system, not from a spreadsheet someone updates. If a figure needs a person to paste it in, it will be stale by Wednesday.
Every number on the screen should tie back to the ledger or the operational record it came from. If the margin on the dashboard does not match the profit and loss, people stop trusting the dashboard within a month.
Start with five or six numbers. Not twenty. The owner or GM should be able to glance at the screen in thirty seconds and know whether the business had a good week. Detail goes on the second page, or on a drill-down.
And the numbers need to be current. A financial dashboard that shows last month's results is a report, not a dashboard. The useful ones update daily or close to it.
Construction: progress, cash and cost to complete
A builder running three to fifteen jobs at any time watches numbers that sit across the estimating system, the job management tool (Procore, Databuild, simPRO or a spreadsheet), the accounting file and the bank.
The numbers that matter:
- Job margin to date. Revenue claimed against costs booked, job by job. The estimating figure lives in the tender, the claimed revenue in the progress claim schedule, and the costs in the accounting file or job system. When these are in different tools, the margin is always a week or more behind.
- Cost to complete. What is left to spend on each job, compared to the original estimate. This needs the committed costs (purchase orders and subcontract agreements) as well as the invoices already paid.
- WIP position. Over-claimed and under-claimed amounts across all jobs. If you have not read it, there is a detailed walkthrough in Construction KPIs: 12 Numbers Builders Should Track.
- Cash flow forecast. Expected claim dates, retention release dates, supplier payment terms and committed costs forward. The data comes from the claim schedule, the retention register, aged payables and the purchase order register. More on this in Construction Cash Flow Forecast.
- Variations pending. Approved, submitted but unapproved, and not yet submitted. The total value of pending variations is money the business has earned or expects to earn but cannot claim yet.
- Defect close-out rate. For builders in the defects liability period, the number of open defects and how quickly they are closing. This usually lives in the project management tool.
Where it breaks: the estimating system does not talk to the accounting file, so the QS ends up maintaining a spreadsheet called something like Job_Costs_FINAL_v3.xlsx. The construction KPI dashboards we build pull the estimate, the progress claims, the ledger costs, the PO register and the retention schedule into one view.
Wholesale and distribution: stock turn, margin and fill rate
A wholesale business lives and dies by stock. Buy too much and you tie up cash. Buy too little and you lose the sale. The numbers that matter all sit across the inventory system (Cin7, Unleashed, Dear, MYOB Acumatica or an ERP), the accounting file and the sales system.
- Gross margin by customer and product line. Revenue less cost of goods, with rebates, freight and credits allocated. The challenge is that rebates sit in a spreadsheet, freight costs sit in the transport invoices and the sales data sits in the inventory or ERP system. There is a detailed look at why this number is hard to get in Why You Can't See Margin by Customer.
- Stock turn by category. How many times you sell through your stock in a year. The data is stock on hand (inventory system) and cost of goods sold (accounts). If the inventory system and the accounting file disagree on stock value, the number is wrong before you start.
- Fill rate. The percentage of order lines shipped complete on the first pick. This comes from the warehouse or inventory system and is often not tracked at all until someone asks.
- Aged stock. Items sitting in the warehouse past a threshold, say 90 or 180 days, that are tying up cash and warehouse space. The inventory system holds the receipt dates.
- Debtor days by customer. How long customers take to pay, tracked against their terms. The aged receivables report from the accounting file.
- Landed cost accuracy. Whether the estimated landed cost on each purchase (freight, duty, foreign exchange) matched the actual cost once all the supplier invoices arrived. This involves the purchase order, the supplier invoices and the exchange rate at the time of payment, which usually live in three different places.
A business intelligence dashboard over these systems puts margin, stock turn and debtors on one screen instead of a monthly spreadsheet exercise. The tricky part is getting the stock value in the inventory system to agree with the balance sheet, which is a reconciliation problem we see in almost every wholesale business.
Manufacturing: throughput, yield and job cost
A small manufacturer, whether it is a fabrication shop, a food producer or a plastics moulder, watches different numbers depending on whether it runs jobs or production runs. But the data problem is the same: production sits in one place, costs sit in another and the accounting file only sees invoices.
- OEE (overall equipment effectiveness). The share of planned production time that is actually productive. It combines availability, performance and quality. The data comes from machine run logs, downtime records and quality inspection notes, which in most small manufacturers live on a whiteboard, a paper log or a spreadsheet. For a worked OEE calculation and how to start reporting without a shop-floor system, see OEE for Small Manufacturers.
- Job cost versus quote. What a job actually cost compared to what was quoted. Materials come from purchase orders or BOM issues, labour from timesheets, and overheads from the accounting file. Comparing quote to actual usually means someone pulling from three exports.
- Yield or scrap rate. Finished output against raw material input. Critical in food manufacturing where ingredient costs are a large share of the total. The data sits in production records and stock movements.
- On-time delivery. Whether jobs shipped by the promised date. The promise lives in the order or job system, the actual ship date in dispatch or freight records.
- WIP value. Jobs started but not yet finished and invoiced. In a make-to-order shop, this is often the largest balance sheet item and the hardest to get right. Labour hours, material issues and overheads all need to land on the right job.
The gap: production data lives on the shop floor and cost data lives in the office. Connecting them is what production data integration and a sensible production planning setup are about.
Transport and logistics: revenue per run, utilisation and compliance
A transport operator watches a different set of numbers, but the data problem is familiar. Revenue sits in the TMS or billing system, costs sit in the accounts and fuel system, and driver hours sit in the roster or telematics.
- Revenue and margin per run, truck or customer. The revenue for each delivery or route, less the direct costs: fuel, tolls, driver wages or subcontractor payments. Why this is hard to get is covered in Why You Can't See Profit per Run.
- Fleet utilisation. The share of available hours or kilometres that earn revenue. Idle time, maintenance downtime and empty runs. Telematics gives the running data, but matching it to revenue means joining it to the billing system.
- Fuel cost per kilometre or per tonne-kilometre. Fuel cards give litres and cost by vehicle. The TMS or dispatch gives the route and tonnage. The two rarely line up automatically.
- Driver compliance hours. Fatigue hours against the limits. Telematics and scheduling systems hold the data, but the operations dashboard needs to show breaches or near-breaches before they happen, not after.
- Debtor days by customer. Some transport customers pay in 14 days, some in 60. The billing system knows what was invoiced, the accounting file knows what was paid, and the gap between them is cash flow.
A good transport dashboard puts revenue per truck beside fuel cost per kilometre and debtor days, so the operations manager sees whether a customer or a route is making money after all costs, not just on the rate card.
Labour hire: margin per placement, fill rate and compliance
Labour hire businesses generate a lot of transactions for each placement: timesheets, award calculations, pay runs, client invoices and compliance records. The numbers that tell you whether the business is healthy sit across the recruitment or CRM system, the rostering platform, the payroll and the accounting file.
- Margin per placement. The charge rate less the pay rate, on-costs (super, workers comp, payroll tax, leave provisions) and overheads. Simple in concept, hard when the charge rate sits in the client contract, the pay rate sits in payroll, and on-costs are calculated differently depending on the award and the state. There is more detail in Pay Rate vs Charge Rate.
- Fill rate. Shifts requested versus shifts filled. The rostering system holds this, but many businesses only know it anecdotally.
- Time to fill. From client request to candidate starting. Lives in the CRM or recruitment platform.
- Timesheet turnaround. How many hours or days between the shift ending and the timesheet being approved. Slow timesheets delay everything downstream: payroll, client invoicing and cash flow.
- Compliance currency. How many active workers have current tickets, licences, medicals and inductions. The onboarding or compliance system holds the expiry dates, but when this is a spreadsheet it goes stale quickly.
The pattern is the same: each number comes from a different system, and nobody sees them together without a manual exercise. A dashboard that pulls the roster, payroll, compliance and billing data into one place turns a fortnightly scramble into a daily glance.
Facility services: profit by site, labour cost ratio and defects
Cleaning, security and maintenance contractors manage dozens or hundreds of sites, each with its own contract value, roster, staff and service schedule. The numbers are simple in principle but scattered in practice.
- Profit by site or contract. Contract revenue less the labour, materials and overheads for that site. Revenue lives in the billing system, labour in payroll (allocated by roster), and materials and equipment in the accounting file. Why this is hard to see is covered in Profit by Site for Facility Services.
- Labour cost as a percentage of contract value. The single most important ratio in a cleaning or security business. Needs the roster (planned hours), timesheets (actual hours), payroll (cost including award loadings) and the contract value.
- Roster versus actual hours. Planned hours against hours actually worked. Over-servicing eats margin silently, shift by shift. The rostering system and the timesheet system hold the two sides.
- Defect or service call close-out time. How quickly reported defects or service requests are resolved. The job or helpdesk system holds the data, and contract KPIs often require a report.
- Award compliance. Whether pay matches the correct award rates for the shift, the classification and the day. Getting this wrong is an underpayment risk. The Cleaning and Security Award Pay Errors post covers where the roster and payroll disagree.
What ties all of these together
Every industry example above follows the same pattern. The business already has the data. It sits in the accounting file, the job or inventory system, the roster, the fleet system or the CRM. The problem is that nobody can see the numbers together without someone spending half a day copying them into a spreadsheet.
A dashboard tool like Power BI, Tableau or even a well-built spreadsheet with live connections can pull the numbers together. The real work is not the dashboard. It is the data underneath: making sure the inventory system agrees with the balance sheet, making sure timesheet hours match payroll, making sure job costs include committed costs and not just invoices paid.
That is what we spend most of our time on. For a broader look at reporting and analytics, see Business Intelligence and Analytics.
If a spreadsheet is working and you trust the numbers, keep it. A dashboard is worth building when the spreadsheet takes someone a day to update, when the numbers are stale by the time anyone sees them, or when a mistake in the copy-paste has cost real money.
Frequently asked questions
What is a KPI dashboard?
A KPI dashboard is a screen that shows the key numbers for your business, updated automatically from the systems where the data already lives. Instead of someone pulling reports from three or four systems and pasting them into Excel every week, the dashboard does it daily or in close to real time. The numbers might be job margins, stock turn, fleet utilisation or margin per placement, depending on your industry.
How many KPIs should be on a dashboard?
Start with five or six. The ones the owner or GM actually looks at when deciding whether the week was good or bad. Detail can go on a second screen or a drill-down. A dashboard with thirty numbers on it is a report nobody reads.
Do I need Power BI or Tableau for a business dashboard?
Not necessarily. Power BI and Tableau are common because they connect to most accounting, ERP and operational systems. But a well-built Google Sheet or Excel workbook with live data connections can work for a smaller business. The tool matters less than the data: if the numbers underneath are wrong or stale, the dashboard will be too.
What does building a KPI dashboard cost?
It depends on how many source systems are involved, how clean the data is and how much reconciliation work is needed to make the numbers agree. We scope every project first and give a fixed quote after a free initial chat. The only published price is our $5,000 Process and Systems Audit, which maps how your data flows and where the gaps are, and often produces the spec for the dashboard as a by-product.
Where should I start if my data is in spreadsheets?
Start with one number you care about and one source system. For example, get your stock turn updating automatically from your inventory system, then add margin from the accounting file. Build trust in one figure before adding more. If the spreadsheet is doing the job and you trust it, leave it alone until it breaks or takes too long to maintain.
If your numbers come from several systems and you are not sure what a dashboard would involve, get in touch for a free initial chat and we will tell you honestly whether it is worth the effort or whether a simpler fix will do.