Why ERP Projects Need KPIs From Day One
Six months after go-live, someone on the board asks whether the ERP project worked. The answers that come back are usually impressions: "the system is running", "people are getting used to it", "inventory is still a headache". None of that helps anyone decide whether to invest more, fix something, or move on, because none of it says how much performance changed or where the problems still are.
ERP implementation KPIs are a short list of numeric measures that you define before the project starts and track during and after it. They answer three separate questions. Is the project on plan and on budget? Are people actually using the system the way it was designed? And have the business processes that justified the purchase improved?
This guide groups the KPIs along those three questions: project health during implementation, adoption after go-live, and operational and financial value. For each one we explain how to calculate it, where the data comes from and what it tells you, and we walk through a hypothetical Saudi distribution company to show how the numbers look on a management dashboard.
Capture a Baseline Before Anything Else
The most common measurement mistake is remembering KPIs only after go-live, when there is nothing left to compare against. If you never recorded how many days month-end close took before the new system, you cannot prove it got faster. Baseline work belongs in the analysis phase, ideally at least three months before go-live.
The baseline does not need to be perfect. It needs a method you can repeat the same way afterwards. Typical baseline measures:
- Month-end close: working days from the last day of the month to an approved final trial balance, averaged over the last three months.
- Inventory accuracy: from the most recent physical count, the share of counted SKUs whose quantity matched the records.
- Order-to-cash cycle: average days from receiving a customer order to collecting payment, from a sample of at least 50 orders.
- Invoice errors: credit notes issued in a quarter because of a wrong price, quantity or customer detail.
- Manual effort: the team's estimate of hours spent on recurring reports, reconciliations and copying data between files.
Write down how each number was extracted and by whom. If the method changes after go-live, the comparison becomes meaningless.
Group One: Project Health KPIs
These belong to the project team and the steering committee and are reviewed roughly every two weeks. Their job is to show drift early, while it is still cheap to correct.
1. Schedule adherence
Milestones completed on time divided by milestones due to date. Milestones should be concrete: requirements signed off, chart of accounts configured, master data migrated, user acceptance testing passed. A slip on an early milestone such as requirements sign-off cascades through everything after it, so even a one-week delay deserves attention.
2. Budget variance
Variance = (actual cost to date ÷ planned cost to date) - 1. The total matters less than the breakdown: subscription or licences, consulting, customization, training, data migration, and internal staff time. That last line is almost always ignored, even though an accountant who spends half the week testing is not doing their normal job, and someone is paying for that.
3. Scope stability
The number of approved change requests after requirements sign-off, and their cumulative impact on time and cost. Change requests are not bad in themselves. A steady stream of them usually means the requirements were not understood, or that departments are trying to rebuild their old way of working inside the new system.
4. Data migration quality
Records passing validation as a share of records migrated, measured separately for customers, suppliers, items, opening balances and fixed assets. The decisive test is balance reconciliation: does the total of customer balances in the new system equal the receivables control account in the old one at cut-off? Does migrated stock value equal the book figure? Every difference should be explained before go-live, not after.
5. User acceptance testing results
Test scenarios passed at first attempt, and open defects by severity. A rule many teams follow: no go-live with any critical defect open, for example a VAT calculation error or e-invoices failing to submit.
Group Two: Adoption KPIs After Go-Live
A system nobody uses delivers nothing, however well it is configured. For the first three months, adoption KPIs matter more than value KPIs, because value will not show up until the new way of working settles.
6. Active usage
The share of licensed users who posted real transactions during the week, not just logged in. System logs show this directly. Low usage in one department is an early warning about training, screen design or the department head's own commitment.
7. Transactions outside the system
This tells you more than usage rates. How many purchase orders this month started as an email or paper form and were keyed in later? How many invoices were prepared in a spreadsheet? How many reports are still assembled by hand from exports? Each one means duplicated work, a chance of error and a gap in the audit trail.
8. Support tickets
Weekly ticket volume, classified as how-to questions, configuration errors or genuine faults. A spike in the first weeks is normal. What should worry you is volume that has not dropped after two months, or the same people asking the same questions, which means training did not land.
Group Three: Operational and Financial Value
These are the measures senior management cares about, and they should tie directly to the reasons the company bought the system. If inventory chaos was the main reason, inventory accuracy comes first. If slow close was the reason, close days come first.
9. Month-end close days
When sales, purchasing, inventory and payroll post to one ledger automatically, accountants stop assembling figures from scattered sources. Fewer close days is one of the clearest signs an implementation worked. Track the number of manual adjusting entries at month-end too: fewer adjustments mean data is entering correctly at the source.
10. Inventory accuracy
The share of SKUs whose physical count matches the system within a defined tolerance. The link to revenue is direct. An item the system says is available but is not on the shelf becomes an order you cannot ship, and an item on the shelf that the system cannot see becomes a duplicate purchase and idle cash. Regular cycle counts keep this KPI honest; see our guide to stocktaking.
11. Order-to-cash cycle time
From customer order to cash collected. This one number spans quoting speed, stock availability, dispatch, invoice accuracy and collections. Break it down by stage to see where time is lost; our order-to-cash guide covers each stage.
12. Invoice accuracy and e-invoicing acceptance
Invoices issued without a later credit or debit note to fix an error. In Saudi Arabia, add a local measure: the share of e-invoices accepted by the Fatoorah platform at first submission. Under Phase 2, standard B2B tax invoices are cleared by ZATCA before they reach the buyer, and simplified invoices must be reported within 24 hours of issue. Repeated rejections point to problems in customer master data or tax setup that need fixing at the root.
13. Return on investment
ROI = (measured benefits - total cost of ownership) ÷ total cost of ownership, over a defined period, typically three to five years. Benefits should be measured, not assumed: staff hours saved and redeployed, excess stock reduced, early-payment discounts captured, penalties avoided. For a fuller method, read our article on ERP ROI.
A Hypothetical Example: A Distributor in Dammam
Consider a food distributor in Dammam with 60 staff, 1,800 SKUs and three warehouses that moved to a cloud ERP. The figures below are invented for illustration, to show how a KPI view looks to management:
| KPI | Baseline | Month 3 | Month 9 | Target |
|---|---|---|---|---|
| Month-end close | 14 working days | 10 days | 5 days | 5 days |
| Inventory accuracy | 82% | 88% | 96% | 97% |
| Order-to-cash | 58 days | 55 days | 46 days | 45 days |
| Invoices needing correction | 6.5% | 4% | 1.2% | Under 1% |
| Transactions outside the system | Not measured | 18% | 3% | Zero |
| Weekly support tickets | n/a | 35 | 6 | Under 10 |
Improvement in the first three months is modest, and that is normal. The team is still learning, some data still needs cleaning, some procedures are being adjusted. The danger is management declaring failure at this point, which is why phased targets for months 3, 6 and 9 should be agreed before go-live.
Order-to-cash also improves more slowly than the rest, because it depends on customer behaviour and credit policy as much as on software. The system provides visibility and alerts, but chasing a late payer is a management decision.
Choosing KPIs Without Drowning in Them
A list of thirty KPIs is a list nobody reads. A practical rule is six to ten for senior management, with longer lists inside each department. To choose them:
- Write down why you bought the system: three to five specific reasons, such as "we cannot see branch profitability", "close takes two weeks" or "we lose orders to stockouts".
- Turn each reason into a measure: branch profitability becomes "branch P&L available within five days of month-end"; stockouts become "orders shipped complete".
- Give every KPI one owner: a named person who measures and explains it, not a whole department.
- Fix the data source: ideally extracted from the system automatically, not from a spreadsheet someone updates by hand.
- Set phased targets: expected values at 3, 6 and 12 months, agreed between the owner and management.
- Attach an action: what happens if the target is missed? Extra training, a revised procedure, a configuration review.
Department-Level KPIs
| Department | Suggested KPIs |
|---|---|
| Finance | Close days, manual adjusting entries, days sales outstanding, e-invoices rejected |
| Procurement | Requisition-to-PO time, spend under approved POs, first-pass three-way match rate |
| Inventory and warehouse | Inventory accuracy, stock turns, slow-moving items, orders shipped complete |
| Sales | Quote turnaround, quote-to-order conversion, late orders |
| HR and payroll | Payroll right first time, wage file uploaded on time, expired employee documents |
Saudi Compliance KPIs Worth Adding
Part of an ERP's value in Saudi Arabia is lower compliance risk, and that can be measured too. A few simple monthly indicators:
- VAT returns filed on time: returns are monthly if annual taxable supplies exceed SAR 40 million and quarterly otherwise, due by the last day of the following month. Late filing penalties run from 5% to 25% of the tax due, so track returns filed on time against returns due.
- Wage file uploaded within the window: since 1 March 2025 the Mudad wage protection file must be uploaded within 30 days rather than 60. Track the months it was uploaded on time.
- Output VAT difference: the gap between output VAT in the ledger and on the return. The closer to zero without manual reconciliation, the healthier the tax setup.
- Expired employee documents: iqamas or licences past expiry without renewal. The target is zero.
These rarely appear in an ROI case, but a single avoided penalty can cover several months of subscription, so management should see them as numbers rather than reassurances.
Common Measurement Mistakes
- Measuring the launch, not the outcome: going live on schedule is not success if staff run the old spreadsheets in parallel.
- No baseline: without a before figure, improvement cannot be shown.
- Judging too early or too late: assessing value in month one is unfair; waiting two years without measuring hides problems.
- KPIs that are hard to extract: if a measure takes two days of manual work each month, it will quietly stop being measured.
- Ignoring internal cost: counting only subscription and consulting makes ROI look better than it is.
- Punishing the messenger: if bad numbers bring blame, people learn to hide them.
Building the Dashboard
The aim is for the general manager to see the headline KPIs on one screen, refreshed from the system without manual work. For each KPI show the current value, the target, the six-month trend and a simple status colour. Review it monthly in a short meeting that starts with the red items and ends with named owners and actions.
Between months nine and twelve, many implementation KPIs become permanent operating KPIs. The question shifts from "did the project work?" to "are we holding and improving the level?". At that point drop measures like support tickets and off-system transactions and keep the value measures.
Tranquil's cloud platform brings accounting and finance, inventory, sales, purchasing and HR into one system, with e-invoicing that covers Phase 1 and Phase 2, and customization services can build reports around your own KPIs. If you are still planning, start with our guide to ERP implementation.
Frequently Asked Questions
When should we start measuring ERP KPIs?
Before go-live. Project health KPIs start with the project, and operational baselines should be captured during analysis, otherwise you will have nothing to compare against later.
How many KPIs do we need?
Six to ten at management level is enough for most mid-sized companies, with more detailed lists per department. Each one should map to a clear reason for buying the system.
When do ERP benefits show up in financial KPIs?
It depends on company size and implementation quality. The first months are usually a learning period; clear gains tend to appear once off-system work stops and procedures settle. Set phased targets rather than a single date.
What is the single most useful KPI right after go-live?
The share of transactions done outside the system. If people are going back to spreadsheets and email, no other KPI will improve.
Can Tranquil report on these KPIs?
Financial and operational data sit in one system, and custom reports can be built through customization services. Talk to the Tranquil team about the right KPIs for your project.