Business Process Improvement Guide: Map, Document, Connect, Automate

A plain-English guide to improving how work flows through a growing Australian business.

This guide covers how to improve the way work flows through your business: how to spot the processes that are holding you back, map them, write them down, simplify them, and then connect and automate what is left. It is written for owners, finance managers and operations managers in Australian businesses with roughly 10 to 200 staff, running on a mix of accounting software, other systems and spreadsheets.

The order matters. Most businesses jump straight to new software or automation and end up automating a mess. The steps that pay off come first and cost the least: see the process clearly, cut out what is not needed, and get everyone doing it the same way. Connecting systems, replacing spreadsheets and adding automation or AI come after.

Each section gives you the owner-level answer, then links to our more detailed articles and pages if you want to go deeper. Nothing here needs a technical background.

1. Spotting a Broken Process

Broken processes rarely announce themselves. They show up as small, familiar frustrations that everyone has learnt to live with. The business still works, so nobody questions it, but the cost keeps growing quietly as you take on more customers, jobs or staff.

The most useful test is simple: ask where work waits, where it gets retyped, and where it depends on one person. If month-end takes a week because someone pulls numbers out of four places, that is a process problem, not a people problem. If a job stalls whenever one staff member is on leave, the knowledge lives in their head instead of in the business. If new starters take months to learn how things really get done, the process was never written down.

Common signs we see in 10-200 staff businesses:

  • Month-end, reporting or invoicing takes longer every time the business grows
  • The same details are typed into two or three systems
  • Only one person knows how a key process or system works
  • Nobody has written down how the work is done, so everyone does it slightly differently
  • Approvals sit in inboxes and someone spends their day chasing them
  • Mistakes are found late, often by a customer or supplier
  • The answer to growth is always "hire another admin person"

Go deeper

2. Mapping How Work Really Flows

You cannot improve a process you cannot see. A process map is a drawing of how the work actually happens today: every step, who does it, which system or spreadsheet it lives in, and every hand-off between people and software. It is not the version in the policy manual; it is the version your team really follows, workarounds included.

Start with one process that hurts, such as quote to cash, procure to pay or month-end. Get the people who do the work in a room (or on a screen-share) and walk through a few real examples from start to finish. Draw it with one lane per role or system, often called a swimlane diagram, so the hand-offs stand out. A whiteboard or sticky notes is enough to begin.

Then read the map for problems. Mark every place the work waits, gets retyped, gets checked twice, or relies on a single person. Those marks become your fix list. Most businesses find that the biggest delays are not inside any one system but in the gaps between them, where someone exports, copies and pastes to keep things moving.

Go deeper

3. Writing SOPs Your Team Will Use

A standard operating procedure (SOP) is a short, written description of how a task is done in your business. Good SOPs make training faster, cut down on mistakes, and stop the business depending on one person's memory. Bad SOPs sit in a binder or a shared drive and nobody opens them.

The difference is mostly in how they are written. Write SOPs from the process map, not from scratch, so they describe the real work. Keep each one to a single task, with numbered steps, the system or form used at each step, and what to do when something goes wrong. Have the person who does the job write or check it, and have someone new try to follow it. Where they get stuck is where the SOP needs work.

Keep them where the work happens, link them from the system people already use, and give each one an owner who updates it when the process changes. If a process is too complicated to write down clearly, that is usually a sign it needs simplifying before anyone documents it.

Go deeper

4. Simplify Before You Automate

The most common mistake in process improvement is automating a process exactly as it is. If a step is pointless, automating it just makes a pointless step run faster, and now it is harder to remove because it is built into software.

Work through the map in this order. First, delete: which steps, reports, approvals or copies exist only because they always have? Second, simplify: can two checks become one, can the form ask for less, can the approval limit go up for routine purchases? Third, standardise: get everyone doing the task the same way. Only then automate what is left, and only the parts that are repetitive and rule-based.

The same thinking applies to buying software. New software over a broken process tends to automate the problem and add a monthly fee. Once the process is clean, you often find your existing systems can do the job, or that two of them just need to talk to each other.

Go deeper

5. Manual Data Entry and Double Entry

Double entry, in the process sense, is when the same information is typed into more than one place: a timesheet written on site, keyed into payroll, then keyed again into job costing. An order taken in the CRM, retyped into the inventory system, then retyped into Xero or MYOB. Each re-keying takes time, and each one is a chance for an error that someone has to find and fix later.

The cost is not only the typing. It is the delay while information waits to be entered, the decisions made on numbers that are a week old, and the time spent working out why two systems disagree. These costs are spread across many people, so they rarely show up as a line item.

The fix is to enter information once, at the point where it is created, and let it flow to every system that needs it. Sometimes that is a built-in connection or a marketplace app; sometimes it needs a proper integration between systems. Either way, map where the retyping happens first, so you fix the hand-offs that matter most.

Go deeper

6. Replacing Spreadsheets and Access Databases

Spreadsheets are a great way to start something. They become a problem when a core process, like job costing, stock, rosters or approvals, runs on them for years. You end up with several versions and no certainty about which is right, formulas only one person understands, no record of who changed what, and numbers that have to be copied in and out of your other systems by hand. An old Microsoft Access database built years ago by someone who has since left raises the same risks.

Not every spreadsheet needs replacing. A good rule: if a spreadsheet is shared by several people, feeds your accounts or reports, or would stop the business if it broke, it deserves a proper home. That home might be a feature your accounting software already has, an existing app built for the job, or a custom system connected to Xero or MYOB when nothing off the shelf fits.

Move one spreadsheet or process at a time. List what you have, decide which ones matter, clean the data before it moves, and run old and new side by side for a short period so nothing is lost.

Go deeper

7. Approvals and Invoice Processing

Approvals and supplier invoices are where many businesses feel the pain first, because they touch several people and end up in the accounts. A typical manual version: a purchase is requested by email, approved by a reply (or a hallway chat), the invoice arrives in a shared inbox, someone matches it to the order by hand, chases the approver, then keys the bill into Xero or MYOB.

An automated version follows the same steps without the chasing. Requests go to the right approver based on rules you set, such as amount, cost centre or job. Reminders go out on their own. Every decision is recorded, so you can see who approved what and when. Supplier invoices are read automatically, matched to the purchase order and receipt, and only the ones that do not match go to a person. Approved bills are posted to your accounting software ready to pay.

The same pattern suits other paperwork-heavy processes, such as recipient-created tax invoices (RCTIs) calculated from dockets or tickets, contracts and forms that need sign-off, and compliance records that need a clear trail.

Go deeper

8. Workflow Automation, RPA, Zapier and Power Automate

Once a process is simplified, automation takes the repetitive, rule-based steps off people's plates. There are a few broad ways to do it, and the right one depends on the job, not on which tool is fashionable.

Built-in features and marketplace apps come first: your accounting, job or inventory software may already do what you need. Connector tools such as Zapier and Microsoft Power Automate link cloud apps with triggers and actions ("when a form is submitted, create a contact and send an email"); they are quick to set up and suit simple, low-volume tasks. Robotic process automation (RPA) uses a software robot to click through screens the way a person would; it is useful for old systems with no other way in, but it can break when a screen changes. A custom integration connects systems directly through their APIs (the official way software talks to other software); it costs more to build but suits work that touches your accounts, runs at volume or needs proper error handling.

Whatever you use, pick one process first, decide what should happen when something fails, and make sure someone gets told. A quiet automation that stopped working weeks ago is worse than no automation.

Go deeper

9. Practical AI in Finance and Operations

AI is now genuinely useful for some everyday finance and operations work, as long as you are clear about its job. The approach that works: AI does the first pass, and a person approves. It reads a supplier invoice or a delivery docket and pulls out the details, suggests a code, flags a line that does not match the purchase order, or drafts a monthly commentary from your reports. Your team checks and approves before anything is posted, paid or sent.

Good starting points are tasks that are high in volume, follow a pattern and are easy to check: invoice and receipt capture, matching documents across systems, sorting incoming emails and requests, and first drafts of routine reports. Poor starting points are anything where a wrong answer is expensive and hard to spot, or where the judgement is the whole job.

AI also depends on the process and data underneath it. If your invoices arrive in five different inboxes and your job codes are inconsistent, fix that first. AI works best inside a clean, mapped process, not as a patch over a messy one.

Go deeper

10. Measuring the Cost of Manual Work

Before you spend money changing a process, work out roughly what it costs today. You do not need a consultant's figures or an industry average; your own numbers are more useful and more honest.

For each process, estimate how often it happens (per week or month), how long it takes each time, and who does it. Multiply by what that person's time costs the business. Then add the less obvious costs: how often errors happen and how long they take to fix, what a late invoice or a wrong stock count costs you, and the risk if the one person who knows the process leaves. For example, if your team keys supplier invoices by hand, count a typical week's invoices, time a handful of them, and note how many needed fixing later.

Compare that with the cost of the change, whether it is simplifying the process, connecting two systems or a larger build. Some processes will clearly be worth automating. Others will be small enough that simplifying them and leaving them manual is the sensible answer, and that is a fine outcome too.

Go deeper

Where to Start

You don't need to do everything in this guide. Pick the line that sounds most like your business and start there.

How Kenomont Helps

Kenomont is a small, founder-led systems consultancy. The work is done by Aaron Chambers, who has 20+ years building business systems and fixing the processes around them. We're based on the Gold Coast and work with businesses across Australia remotely, and travel when a project needs it.

We work on the processes that sit between your people and your systems: mapping them, simplifying them, then connecting and automating the parts that are worth it, around Xero, MYOB and the software you already use. That includes a purchase order approval system for a large retirement living provider, in daily use since 2014, and a forestry export and port management system with real-time weighbridge integration that we have run for Australian forestry companies since 2012.

Every project is scoped and given a fixed quote after a free initial chat. There's no obligation, and if a cheaper route like a marketplace app will do the job, we'll say so.

Frequently Asked Questions

Start with the one process that causes the most pain right now: the month-end that drags, the invoices that get chased, the orders that get retyped. Walk through a few real examples with the people who do the work and draw every step, including the workarounds. That map tells you what to fix first, and most of the early fixes cost nothing but a decision.

Fix the process first. New software laid over a broken process usually just makes the same mess faster, and you pay for it every month. Once the process is mapped and simplified, you will know whether the software you already have can do the job, whether two systems simply need connecting, or whether you genuinely need something new.

If you do the mapping and simplifying yourselves, it costs your team's time. When you want help, our Process & Systems Audit is a fixed $5,000: we map your key processes and systems and give you a written report with a prioritised plan. Any follow-on work, such as connecting systems or automating steps, is scoped and given a fixed quote after a free initial chat.

No. Most process improvement is about how people do the work, so you need the people who do it, not a technical team. When a fix means connecting systems or building an automation, we can do that part. If you have someone who looks after your systems, we bring them in.

For simple, low-volume jobs between popular cloud apps, a tool like Zapier or Power Automate is often enough and is the cheaper route. It gets shaky when the automation touches your accounts, handles lots of records, needs careful error handling or keeps breaking quietly. That is when a proper integration is worth the extra cost.

Not on its own, and you should not want it to. AI is useful for the first pass: reading invoices and dockets, suggesting codes, spotting things that do not match and drafting reports. A person still reviews and approves before anything is posted or paid, and sign-offs such as BAS stay with you and your accountant.

Use your own figures. Count how often the task happens, how long it takes each time, who does it and what an error costs to fix, then compare that with the cost of changing it. Our automation ROI calculator walks you through it. If the numbers are small, the honest answer may be to simplify the process and leave it manual.

Yes. For processes like month-end, supplier invoices and procure to pay, your accountant or bookkeeper often knows exactly where the numbers go wrong. We are happy to include them in mapping sessions and to design any changes so they keep working the way they need to in Xero or MYOB.

Which Process Would You Fix First?

Tell us which process gives you the most trouble. We'll tell you honestly whether it's one you can map and fix yourselves, or where we can help. Or call 1300 470 091.

Free initial chat. No sales pressure - if we're not the right fit, we'll say so.

Prefer to talk? Call 1300 470 091

Got a spec or files? Email [email protected]

Name, email and message are required. We'll respond within 24 hours.