At some point, most Australian businesses with 5–50 staff hit the same wall. The CRM doesn't know what the job book knows. The spreadsheet that tracks quotes has quietly become the real system of record. Someone spends every Friday afternoon copying data from one tool into another so the invoices go out right.
When you reach that wall, you have three honest options — not two. You can integrate the tools you already have. You can replace them with one system. Or you can keep everything exactly as it is and change nothing. That third option gets left off the list by almost everyone with something to sell you, which is a shame, because it's often the correct answer.
This article is a decision framework, not a pitch. We'll argue each option properly, including the ones that don't involve buying anything.
Start With the Actual Problem, Not the Software
Before comparing options, write down — in plain words, no product names — what is actually going wrong. Not “our systems don't talk to each other” but the specific, observable cost:
- Double entry: someone types the same customer, job, or invoice details into two or more places
- Things falling through gaps: a quote gets accepted but never becomes a job, or a job finishes but never becomes an invoice
- No single answer: two systems disagree about a customer's details or a job's status, and nobody knows which is right
- Key-person risk: only one staff member understands how the pieces connect
- Reporting takes days: answering “how did we do last month?” means exporting from four tools and reconciling in Excel
If you can't name a cost from that list — or put a rough hours-per-week figure on it — stop here. You don't have a systems problem, you have a curiosity, and curiosities are cheap to leave alone. We wrote more about how businesses end up in this position in why your business runs on six systems that don't talk to each other, and the short version is: it happens gradually, sensibly, one reasonable purchase at a time. Nobody did anything wrong. The question is only whether it's now costing more than it saves.
Option 1: Keep Exactly What You Have
This deserves a genuine hearing, because the status quo has advantages that get systematically undersold.
Your staff already know it. Every system change carries a retraining cost, and in a 15-person business there's no training department to absorb it — it comes straight out of billable time. A slightly clumsy process that everyone executes flawlessly usually beats an elegant process that everyone fumbles for six months.
The risk is known. Your current setup's failure modes are familiar. You know the spreadsheet occasionally gets a stale copy saved over it, and you've learned to check. A new system or new integration introduces failure modes you haven't discovered yet, and you'll discover them at the worst possible time.
Manual steps can be a feature. That Friday-afternoon reconciliation everyone complains about? It might also be the moment errors get caught. Automating a handoff removes the human who was quietly fixing bad data on the way through. Plenty of businesses automate a process and only then discover how much silent correction was happening.
Keeping what you have is the right call when the total measurable cost is small (say, under two or three hours a week across the whole business), when the tools involved are stable and paid for, and when the business isn't growing fast enough to multiply the pain. If that's you, close this tab with a clear conscience. Revisit in a year, or when headcount changes materially. Doing nothing on purpose is a legitimate strategy — it's doing nothing by default that gets expensive.
Option 2: Integrate What You Have
Integration means keeping your existing tools and wiring them together — with native connectors, middleware like Zapier or Make, or custom-built links — so data flows between them instead of being retyped.
Integration is clearly the right answer when:
- A core system works well and is deeply embedded. If your accounting package, practice management system, or industry-specific tool does its main job properly and your team lives in it all day, ripping it out to fix a data-flow problem is surgery to cure a headache. Connect around it instead.
- The gap is narrow and well-defined. “When a deal closes in the CRM, create a job in the job book” is one clean, one-directional flow. Narrow integrations are cheap to build and easy to test.
- Both vendors maintain proper, supported connectors. A first-party integration maintained by the vendors themselves is a very different proposition from a chain of third-party workarounds.
- The tools are keepers. Only integrate systems you intend to still be running in three years. Integrating a tool you half-plan to replace is buying furniture for a house you're selling.
Done under those conditions, integration is usually the cheapest and fastest fix, with the least disruption to staff. It should be the default answer, and replacement should have to earn its place over it.
When Integration Only Delays the Problem
The trouble is that integration has a failure mode that looks like success for the first year. Watch for these signs that you're building a house of cards rather than a fix:
- Integrations break on vendor updates. If one of your tools changes its data structures or authentication every few months and your connections snap each time, you don't have an integration — you have a recurring repair bill and a business process that fails silently until someone notices the invoices stopped.
- Nobody owns it. If the whole web of automations lives in one staff member's head (or worse, their personal Zapier account), you've converted a software problem into key-person risk.
- There's no source of truth. Two-way sync between systems that each think they own the customer record is where data quality goes to die. If you can't say in one sentence which system is authoritative for each type of record, integration will amplify the confusion, not resolve it.
- The middleware bill keeps climbing. Per-task automation pricing is trivially cheap at low volume and quietly substantial at scale. If the connective tissue now costs as much as a system, the maths has changed.
- Every new hire needs a diagram. When onboarding involves explaining seven tools and the invisible glue between them, complexity has become a tax on growth.
The system-of-record test: for each type of data your business runs on — customers, jobs, quotes, invoices, documents — can you name the single system that holds the authoritative copy? If yes, integration can probably work. If several answers are “well, sort of both, plus the spreadsheet”, integration will paper over a structural problem, and you'll be back here in eighteen months having spent money to stand still.
Option 3: Replace With One System
Replacement means retiring several tools and moving their jobs into a single platform. It's the most disruptive option and the most expensive up front — so it needs to clear a higher bar. It clears that bar when:
- The systems being replaced are themselves the problem — not just the gaps between them. If the CRM is disliked, half-populated, and duplicated by a spreadsheet anyway, integrating it just plumbs bad data around faster.
- You're paying for four tools and using a fifth of each. Subscription sprawl is real money. Add up the annual spend across the tools in question before assuming replacement is the expensive option.
- The integration maths has failed. If maintaining the glue costs more per year than a consolidated system would, the “cheap” option isn't.
- Growth is about to multiply the pain. A manual workaround that costs two hours a week at 8 staff costs a great deal more at 25. Replacing before the growth spurt is far easier than during it.
Be equally honest about the costs: data migration is never as clean as promised, there's a productivity dip while staff relearn muscle memory, and you're concentrating risk in one vendor. And beware the off-the-shelf all-in-one trap — a generic platform that does everything adequately can force your business to work the way the software works, which is the original problem wearing a new outfit.
A Decision Framework
| Your Situation | Likely Answer | Why |
|---|---|---|
| Friction costs under ~2–3 hours/week total, tools are stable, headcount steady | Keep what you have | Any change costs more than the problem does |
| Core systems work well and are deeply embedded; the gap is one or two narrow data flows | Integrate | Cheapest fix, least disruption, keeps what works |
| Clear system of record for each data type; vendors maintain supported connectors | Integrate | Conditions exist for integrations that hold |
| Integrations break on vendor updates, or live in one person's head | Lean towards replace | You're paying rent on a problem, not fixing it |
| Multiple systems disagree on the truth; spreadsheets have become the real system of record | Lean towards replace | Integration would amplify the confusion |
| Combined subscriptions + middleware + manual hours exceed the cost of one system | Replace | The maths has already decided for you |
Whichever direction the table points, run the numbers over three years, not one — replacement front-loads its costs while integration spreads them out, so a one-year view flatters integration and a three-year view is honest to both. Our guide to building an IT budget for a small business covers how to put those figures side by side properly. And if the exercise surfaces bigger questions than data flows, that's often one of the signs your business needs a broader IT strategy review rather than a point fix.
If You Do Land on Replacement
The main risk with replacement is buying a generic all-in-one and spending a year bending your business to fit it. The alternative is a system shaped around how you already work. That's the gap we built iTec HQ, our operations platform for Australian SMBs, to fill: it replaces the separate CRM, job book, ticketing inbox, e-sign tool and spreadsheets with one system, built around your existing process rather than the other way round. There are thirteen modules; you run only the ones you need and aren't billed for the rest, it isn't priced per seat, it's hosted in Sydney, and sign-in runs through your own Microsoft 365 tenant — so the MFA and conditional access rules you already enforce apply automatically. Most builds are live in two to four weeks.
Wondering what one system built around your existing process would look like? The details, modules and pricing model are all on the iTec HQ page.
See iTec HQ →To be clear about where we sit: managed IT support is still our core business, and in plenty of conversations our advice is “integrate those two tools and keep the rest” — or “change nothing for now”. We'd rather give you the boring correct answer than the exciting profitable one, because you'll still be a client in five years either way.
The Bottom Line
There is no universally right answer — only a right answer for the specific costs your business is carrying. Keep what you have when the friction is small and the tools are stable. Integrate when your core systems earn their keep and the gaps between them are narrow, well-defined, and properly supported. Replace when the systems themselves are the problem, when the glue costs more than it saves, or when nobody can say which system holds the truth.
And whatever you choose, choose it deliberately, write down why, and put a date in the calendar to check whether the reasoning still holds. The expensive outcome isn't picking the wrong option — it's drifting between all three for years without ever deciding.