How to win with systems

Most businesses do not have a software problem. They have a systems problem that they keep trying to solve by buying software. This guide is how to win with systems instead: fewer tools, one source of truth, processes that are written down, and automation pointed at the jobs that actually cost you money.

I work with businesses across Wollongong and the Illawarra, and the pattern is almost always the same. The business grew. Each new problem got a new app. Nobody owned the whole picture. Now there are 14 logins, three versions of the client list, and a senior person spending Friday afternoon copying numbers into a spreadsheet so the owner can see what happened that week.

None of that is a moral failing. It is what happens when you grow without a plan for your systems. The good news is that it is fixable, and the fix is usually cheaper than the waste.

What I mean by a system

A system is not a tool. A system is the agreed way a piece of work moves through your business, from trigger to done, including who does what, where the information lives, and what happens when something goes wrong.

Your quoting system, for example, might be: enquiry lands on the website, gets logged as a deal, someone qualifies it within a day, a quote goes out from a template, the quote is chased twice, and when it is signed a job is created and the team is told. The tools (WordPress, HubSpot, PandaDoc, Asana) are just where those steps happen.

When people say "our systems are a mess", they usually mean one of three things:

• The steps are not agreed, so everyone does it their own way.
• The information lives in too many places, so nobody trusts any of them.
• The tools do not talk to each other, so people are the integration layer.

Fix those three and you win. Buy another app without fixing them and you get a fourth version of the problem.

Tool sprawl: how you got here and what it costs

Tool sprawl is what happens when every team, and sometimes every person, picks their own software. Sales has a CRM. Marketing has an email platform with its own contact list. Ops runs jobs in a project tool. Finance lives in the accounting package. Someone built a very clever spreadsheet in 2021 that now runs half the business and only they understand it.

The subscription fees are the visible cost. They are rarely the big one. The real costs are:

• Double entry. The same client details typed into three tools.
• Reconciliation. Time spent working out which version is right.
• Context switching. Hunting across apps for the latest file, note or email.
• Dropped balls. Leads and tasks that fall in the gap between tools.
• Key person risk. Processes that only exist in one person's head or one person's spreadsheet.

Example: pricing the waste

Here is the kind of maths I run in an audit. These numbers are illustrative, but they are typical of what I see in a 20-person service business.

Example only. Costs use loaded hourly rates of $60 for admin and general staff, $70 for sales and ops, and $95 for a sales manager, over 52 weeks. Swap in your own numbers.

Sixteen hours a week is not a dramatic number. Nobody notices it. It is two days of someone's time, every week, forever. That is the number that pays for fixing your systems, usually several times over.

One source of truth (and why it matters more than any tool)

A single source of truth means every important piece of information has exactly one home. Other tools can display it or use it, but there is one place where it is created and corrected.

This is the single most important systems decision you will make. Get it right and integrations become simple. Get it wrong and every integration becomes a fight about which copy wins.

Here is how I usually split it for a service business:

The rule is simple: create it in its home, correct it in its home, and let everything else read from there. If someone fixes a client's phone number in the project tool, it should not stay fixed only in the project tool.

If two tools both think they own the client record, you do not have a source of truth. You have an argument that runs on autopilot.

The client record is where most businesses get this wrong. Pick the CRM as the owner of who your clients are, and make every other tool look up from it. That one decision cleans up more mess than any amount of automation.

Process before software

If you automate a bad process you get a faster bad process. If you move a messy process into a new tool, you get the same mess with a nicer interface and a bigger bill.

Before you change any software, write the process down. Not a 40-page manual. One page per process, in this format:

1. Trigger. What starts it? (A form submission, a signed quote, the first of the month.)
2. Steps. Each step as a verb: qualify, quote, chase, book, deliver, invoice.
3. Owner. One named role per step. "The team" is not an owner.
4. Tool. Where each step happens.
5. Data. What information is created or needed at that step.
6. Done. How you know it is finished, and what happens next.
7. Exceptions. The three most common ways it goes off the rails.

Do this with the people who actually do the work, not just the managers. The gap between "how we say we do it" and "how it is actually done" is where most of the waste is hiding.

Once a process is on one page, three things become obvious: steps that add nothing, steps that are done twice, and steps that are perfect for automation because they are repetitive, rule based and boring.

The systems audit method, step by step

This is the same method I use in paid audits. You can run a lighter version yourself over two or three weeks.

The tool inventory template

For step one, a simple spreadsheet is fine. These are the columns I use:

The "users who actually log in" column is usually the most revealing one. Paid seats for people who left last year are common. So are tools the business pays for that do a job another tool already does.

What to automate first

Not everything should be automated, and the things that should be are not always the obvious ones. I score every candidate on four questions:

• Volume. Does it happen often? Daily beats monthly.
• Rules. Can you write down exactly what should happen, including exceptions?
• Pain. Does it cost real hours, cause errors, or slow down revenue?
• Stability. Is the process settled, or still changing every month?

High volume, clear rules, real pain and a stable process is your first automation. Anything that fails on "rules" or "stability" goes back to the process step first.

A sensible first five for most service businesses

1. Website enquiries straight into the CRM with source tracking, owner assignment and a task to follow up. No more forwarded emails.
2. Quote generation from the deal record, so pricing, client details and scope pull through into the proposal template instead of being retyped.
3. Signed quote creates the job: deal moves to Won, a project is created from a template with the right tasks and dates, and the team is notified.
4. Automatic quote chasers at set intervals, stopped the moment the client signs or replies.
5. The weekly report built itself, from dashboards that read live data rather than a spreadsheet someone assembles on Friday.

Notice what is not on this list: AI. AI is useful, and I build plenty of it, but it sits on top of clean data and working processes. If your enquiries are still arriving by forwarded email, an AI agent will not save you. I cover that order of operations in AI agents explained.

Measuring time saved (properly)

If you do not measure before, you cannot prove anything after. Most businesses skip the baseline, then six months later nobody can say whether the project was worth it.

Take a baseline first

For each process you plan to fix, capture three numbers for two to four weeks:

• Volume: how many times it happens per week.
• Handling time: how long each one takes, on average, end to end.
• Error rate: how often it goes wrong and needs rework.

You do not need time-tracking software. A tally sheet next to the person doing the work for two weeks is enough. Ask them to be honest, and make it clear this is about the process, not their performance.

Worked example

Say your team sends 25 quotes a week. Each one takes 35 minutes to build because details are copied from emails and the CRM into a Word template, then exported and emailed.

• Before: 25 x 35 minutes = 875 minutes, about 14.6 hours per week.
• After a CRM-to-PandaDoc template with pricing tables pulling from the deal: 25 x 10 minutes = 250 minutes, about 4.2 hours per week.
• Saving: about 10.4 hours per week, or roughly 540 hours per year.
• At a loaded rate of $70 an hour, that is about $37,800 a year of capacity.

Illustrative numbers. Your volumes and rates will differ.

Be honest about what "saved" means. Hours saved are only worth money if the time goes somewhere useful: more quotes, faster follow-up, better delivery, or not hiring the extra admin person you were about to hire. Decide where the time goes before the project starts.

The measures I track after go-live

Example stack for a 10-50 person business

You do not need the same stack as everyone else, but this combination works well for a lot of service businesses in the 10 to 50 staff range: professional services, trades with an office team, agencies, allied health, education providers and B2B suppliers. It is also a combination I build and support all the time across the Illawarra.

How the data flows

1. A visitor fills in a form on the WordPress site. The form submits straight into HubSpot, carrying the page, campaign and source.
2. HubSpot creates or updates the contact and company, assigns an owner by rule, creates a deal in the right pipeline and sets a follow-up task.
3. When the deal reaches the quoting stage, the salesperson generates a PandaDoc quote from the deal. Client details, line items and pricing pull through from HubSpot.
4. PandaDoc status (sent, viewed, signed) syncs back to the deal. Chasers run until it is signed.
5. When signed, the deal moves to Won and an Asana project is created from the right template, with the client, scope and value attached and the delivery lead assigned.
6. Project milestones update a property on the HubSpot deal or company, so sales and account managers can see delivery status without asking.
7. Dashboards in HubSpot show pipeline, win rate, speed to lead and revenue by source. Nobody builds a weekly spreadsheet.

Some of these connections are native integrations, some need a light automation layer, and some are best done with a small custom integration using APIs and webhooks. The right answer depends on your plan levels and how specific your process is. That is exactly what systems integration work sorts out.

What this stack is not good at

Being blunt: this is not a full ERP. If you hold stock, run complex manufacturing, or need detailed job costing, you will likely need a specialist operations system alongside it, with the CRM still owning the client relationship. Do not force HubSpot or Asana to be something they are not.

Common mistakes I see

• Buying the tool first. Then bending the business around its defaults.
• No owner. Every system needs one person accountable for how it is set up and kept clean.
• Migrating the mess. Moving ten years of duplicates and junk fields into a shiny new CRM. Clean first. See the CRM data clean-up guide.
• Automating exceptions. Building logic for the case that happens twice a year while the daily case is still manual.
• Skipping training. A system people do not understand becomes a system people work around.
• Big bang rollouts. Changing five tools in one month. Change one thing, settle it, then the next.
• No baseline. Then no way to tell if it worked.

FAQ

Where to start this week

You do not need to fix everything. Pick one process that costs you real hours, write it on one page with the people who do it, and take a two-week baseline. That alone will tell you more than any software demo.

If you would rather have someone senior do it with you, get in touch. I will tell you straight whether it is worth fixing, what it will cost, and what you will get back.