Staff Attendance, Leave and Claims: Spreadsheet, Ready-Made App or Custom System?

Somewhere between the tenth and the hundredth employee, staff admin quietly becomes a job. Someone is chasing clock-in records, someone else is answering "how many days do I have left?" on WhatsApp, and at month end a pile of receipts arrives that has to be checked, approved and reimbursed. Most small businesses start with a spreadsheet, and for a while that is the right answer. The question this guide helps with is when it stops being the right answer, and what to move to.

It is a decision guide for owners and office managers, not a product review: it names no software, because the right choice depends on your rules rather than on a brand. If you are not sure that spreadsheets are the problem at all, start with signs your business has outgrown spreadsheets; for the general buy-or-build question across every kind of software, see off-the-shelf vs custom software. This page stays on one use case: keeping track of who worked, who is away and who is owed money.

Attendance, leave and claims are three different jobs

They are usually bundled together as "HR admin", but each behaves differently, and a tool that is good at one can be weak at another.

• Attendance is a record of fact. Who arrived, who left, who worked where. It is mostly data capture, plus the exceptions: late arrivals, missed clock-ins, work done off-site. The hard part is making the record trustworthy, not clever.
• Leave is a request, an approval and a balance. A person asks, a manager decides, and a running balance must update correctly. Rules pile up fast: different leave types, half days, carry-forward, blackout periods, a limit on how many people from one team can be away at once.
• Claims are a request with evidence and a payout. A person submits an amount, a category and a receipt; one or two people approve it; finance reimburses it. The hard part is the paper trail and the approval limits.

That difference matters when you choose. A business with simple attendance but complicated claims needs a different tool from one with the reverse problem, and many of the "we need an HR system" conversations are really about only one of the three.

Write your rules down before you choose any tool

The single most useful preparation is also the cheapest: put your actual policy on one or two pages. Software, whether bought or built, can only follow rules that someone has stated. Ambiguity that people quietly resolve by asking the boss today becomes a design question the moment you automate it.

• Leave types and who gets what. Annual, sick, unpaid, compassionate, replacement leave, and any extra types you grant yourself. Note which are tied to length of service, job grade or contract type.
• How balances behave. Do they reset on a fixed date or on each person's anniversary? Does unused leave carry forward, expire or get paid out? Can balances go negative?
• The approval chain. Who approves what, who covers when the approver is away, and whether a second approval is needed above a certain length of leave or amount of money.
• Claim categories and limits. Mileage, meals, parking, client entertainment, equipment. What needs a receipt, what is capped, and how long after the expense a claim can still be submitted.
• Working patterns. Fixed hours, shifts, part-time, remote or field staff, and any branches whose public holidays differ.
• Reporting. What does the owner, the manager and the person doing payroll actually need to see each month?

Statutory minimums for leave and working hours are set by law and by employment contracts, not by software. This guide deliberately does not state them; confirm the current rules that apply to your staff with the relevant authority or a qualified adviser, and then write your own policy on top of them.

Option one: a spreadsheet

For a team of ten with one or two leave types and a single approver, a well-kept shared spreadsheet or form-and-sheet combination is hard to beat. It costs nothing extra, everyone already understands it, and changing a rule means editing a cell.

Where it works:

• One person owns the file and everyone else only asks them questions.
• Leave rules are simple enough to explain in two sentences.
• Approvals happen in conversation and nobody ever needs to prove later who agreed to what.
• Claims are few and the receipts are collected in a folder anyway.

Where it starts to break:

• Balances drift. A formula is overwritten, a row is inserted in the wrong place, or two copies of the file exist. Staff stop trusting the number, and every query becomes a manual check.
• Approvals live in chat. "Yes, fine" on WhatsApp is not a record. When a dispute arises, someone scrolls back through months of messages.
• Everyone can see everything. Sharing one file with the whole team also shares colleagues' medical leave and claim amounts, which is rarely what you want. Splitting it into private copies then creates the version problem.
• No history of changes. You cannot easily tell who changed a balance, or when.
• Month end is manual. Attendance totals and approved claims have to be collected, reconciled and re-typed for whoever runs payroll.

If three or more of these sound familiar, the spreadsheet is no longer saving money; it is moving the cost into somebody's evenings.

Option two: a ready-made HR app

Many subscription HR and attendance apps are aimed at small and mid-sized companies. They typically bring a staff mobile app, leave requests with an approval flow, balances that update by themselves, a claims form with receipt upload and a set of standard reports, all working from day one. For a business with a conventional policy, this is often the best value: you rent something that has already been tested by many other companies, and you do not carry a build project.

What to check before you commit:

• Does it follow your rules or its own? Test your three oddest policies, not the easy ones. If you have to change your policy to fit the product, ask whether you are happy to.
• How is it priced as you grow? Many products charge per employee per month, and some charge per module, so attendance, leave and claims can each be a separate line. Price it at your headcount today and at your headcount in two years, and check the current figures with the vendor yourself, since they change.
• Where does the staff data sit, and can you export it? You want to be able to leave with your records in a usable file.
• Does it fit the way your staff actually work? Field teams, shift workers and branches with different holidays are where generic products often need workarounds.
• What happens at month end? Ask to see the exact report or file you would hand to payroll, and check that it matches what the person doing payroll needs.

Cost shape is the main difference from the other options: a ready-made app is a running subscription with no build fee, while a custom system is a one-off build you then own and run.

Option three: a custom system

A custom attendance, leave and claims system is a web application built around your rules, with staff logging in on their phone or laptop. It earns its cost when the product you would otherwise use cannot follow your policy, or when the HR process is tied into how the business runs.

• Your approval chain is unusual. Project-based approvals, branch managers who approve only their own site, or a second sign-off above a set amount.
• Your rules are specific. Special leave types, shift-based entitlements or claim limits that vary by role, which a generic product forces you to fake.
• You need one place for several things. Attendance feeding job costing, claims tied to a client or project code, or leave visible on a team's scheduling board.
• You want to own the data and the workflow. No per-employee fee that grows with headcount, and the freedom to change a rule when your policy changes.

Gotka's App & System Development service lists indicative starting prices of RM8,000 for web apps and dashboards and RM12,000 for business systems and integrations. Which one a staff attendance, leave and claims system falls under depends on its scope, for example how many roles it has and how many other systems it must connect to, and the figure is always confirmed by a written quote rather than assumed. For how long a build of this kind typically takes, see how long a custom business system takes.

The three options side by side

None of the columns wins every row, which is the point. Read down the rows that matter most to you and see which column keeps coming up.

Staff data is personal data: privacy and access

Attendance times, medical leave, claim receipts and bank details are personal data about your employees. Whichever option you pick, treat it with the same care you would want for your own records.

• Collect only what you need. If you only need to know whether someone was at work, you do not need a continuous location trail.
• Be careful with location and biometrics. Phone check-ins that record where someone was, or fingerprint and face terminals, collect more sensitive data than a sign-in sheet. Decide whether that is necessary, and tell staff plainly what is recorded and why.
• Limit who sees what. A manager usually needs their own team's leave, not everyone's medical certificates. Claims receipts and bank details should reach finance, not the whole office.
• Keep a record of changes. Being able to answer "who changed this balance and when" protects both you and the employee.
• Take advice on your obligations. Malaysia's Personal Data Protection Act 2010 is the starting point for how personal data is handled; ask a qualified adviser how it applies to your staff records.

A worked example: three hypothetical businesses

These are illustrations, not real clients, to show how the same three jobs lead to different answers.

• A 12-person design office, one location. Two leave types, the owner approves everything, claims are a handful a month. A shared spreadsheet and a receipts folder are fine. The sensible step is to write the policy down, lock the formulas and agree that approvals are recorded by email rather than chat. Moving to a tool now would add cost without removing a real problem.
• A 45-person services company with two branches and rotating shifts. Public holidays differ by branch, supervisors approve only their own site, and attendance is needed by shift. A ready-made HR app is the first thing to trial here, using the three hardest shift rules as the test. If it handles them, rent it; if it needs workarounds for all three, that is a signal to look at a custom build.
• An 80-person contractor with project-coded claims. Staff work on client sites, claims must be charged to a project code, and a project manager and then finance both approve. Attendance hours also feed job costing. This is the profile where a custom system is most likely to pay off, because the process is the business's own and no generic product models it cleanly.

Test with one real month, and write the brief if you go custom

Do not decide from a feature list. Take one recent month of real attendance, leave and claims, and run it through the candidate: your spreadsheet, a trial of a ready-made app, or a written description for a custom quote. Count how many cases it handles without a workaround.

1. Pick the hardest cases. Half-day leave across a public holiday, a claim over the approval limit, a missed clock-in corrected a day later.
2. Run them end to end. From the staff member's request to the report payroll receives.
3. Ask the people who will use it. One manager, one regular employee, and whoever does the month-end work.
4. Count the workarounds. Zero or one is a good fit. Several means the tool is fighting your process.

If the answer points to a custom build, bring the material from what to prepare before your first conversation with a software developer and add these items to the brief:

• Roles and what each can see: employee, manager, finance, owner, and who may see medical or salary-related detail.
• The approval chain and who covers when an approver is away.
• Your written leave and claim rules, including the awkward exceptions.
• Existing balances and history that must be carried in; see moving your data into a new system.
• The month-end output, as the exact report or file layout the person doing payroll can use. Anything beyond an export, such as a direct connection to other software, should be scoped and quoted separately, never assumed.
• How you will accept it: a test month with real cases, as in a user-acceptance testing checklist.

When quotes arrive, how to compare two custom software quotes helps you line them up fairly.

Common mistakes

• Choosing a tool before writing the policy. You end up arguing about software when the real disagreement is about rules.
• Testing only the easy cases. Every product handles a full day of annual leave. The decision lives in the exceptions.
• Sharing one spreadsheet with everyone. It exposes colleagues' leave and claim details, and invites accidental edits.
• Treating attendance as surveillance by default. More data is not better; staff trust and your obligations both argue for collecting what you need.
• Ignoring the per-employee price curve. A subscription that looks small at 15 people can look different at 80.
• Building custom for a standard problem. If a ready-made app handles your policy, a build is cost with no gain.
• Assuming a custom system will connect to everything. Every connection to another tool is a separate piece of scope.

Where Gotka Technologies fits

If your test month shows that a spreadsheet and a ready-made app both fight your process, Gotka's App & System Development service builds web apps and business systems. Indicative starting prices are RM8,000 for web apps and dashboards and RM12,000 for business systems and integrations, each confirmed by a written quote once your rules, roles and connections are understood. If a ready-made app fits, that is a fine outcome too; the aim of this guide is to help you tell the difference before you spend.

Key terms used in this guide

• Approval chain: the sequence of people who must approve a leave or claim request, and who covers when one of them is away.
• Leave balance: the days of a given leave type a person still has available, after approved requests are deducted.
• Carry-forward: unused leave moved into the next period instead of expiring.
• Role-based access: showing each person only the records their job requires.
• Audit trail: a record of who changed what and when.
• Off-the-shelf (ready-made) software: a product built for many customers, configured rather than built for you.
• Custom system: software built to your own rules and workflow.
• Month-end export: the file or report of approved attendance, leave and claims handed to whoever runs payroll.