Most Australian B2B businesses don’t have a lead problem. They have a targeting problem. There are plenty of companies out there you could email; the difficulty is knowing which ones are worth the effort, and being honest enough to leave the rest alone. That’s what an ideal customer profile is for — and done properly, it’s the single cheapest improvement you can make to your outbound.

This is a practical guide to building one for an Australian business: what actually belongs in it, how to work it out from the customers you already have, and how to keep it honest once outbound starts running.

What an ICP is (and what it isn’t)

An ideal customer profile describes the type of business you serve best — the firmographics, the situation, the trigger. It is not a buyer persona. A persona is a person: “Sarah, the practice manager, cares about admin load.” Both are useful, but they answer different questions. The ICP tells you which companies to put on the list. The persona tells you what to say once you’re there.

Getting this wrong is the most common mistake we see. Businesses write a lovely persona document, then build a prospect list off a generic industry filter that has nothing to do with it, and wonder why reply rates are flat. The list comes first.

Start with the customers you’ve already got

Don’t start from aspiration. Start from evidence. Pull your last twenty or thirty closed deals — or however many you have — and sort them into three piles: the ones that were a pleasure and profitable, the ones that were fine, and the ones you would quietly decline if they came back.

Then look for what the top pile has in common. Usually it isn’t the industry. It’s something more specific: they were at a particular size, they’d just hired someone, they were already paying for a worse version of what you sell, they had a compliance deadline, they had one person doing three jobs. The patterns you find here are far more reliable than anything you’d guess in a workshop.

Pay just as much attention to the bottom pile. A good ICP is defined as much by exclusions as by inclusions, and the businesses that drained you were probably drained by something structural — too small to afford the outcome, too big to move quickly, or buying for a reason you can’t reliably deliver on.

The criteria that actually matter in Australia

Once you’ve got the patterns, write them down as filters you could genuinely apply to a list. Vague criteria are useless: “growing businesses that value quality” cannot be searched for. These can:

  • Headcount band. The most reliable proxy for budget and for how decisions get made. A five-person firm and a fifty-person firm buy completely differently.
  • Industry, but narrower than you think. Not “healthcare” — “multi-site allied health clinics”. Not “trades” — “plumbing firms running more than three vans”.
  • Geography. Metro versus regional genuinely changes the pitch, and if you deliver on-site it changes whether you can serve them at all. Australia is large and thinly populated; a Perth prospect is not a Sydney prospect with a different postcode.
  • Structure. Single owner-operator, multi-site, franchise, or part of a group. This decides who signs and how long it takes.
  • Observable triggers. Hiring for a role that signals the problem, a new location, a website that’s obviously neglected, a job ad mentioning the thing you fix. Triggers are what turn a decent list into a timely one.
  • Disqualifiers. Write these explicitly. Government tenders you can’t service, industries with approval chains you can’t survive, businesses below a size where your price can’t work.

One useful test: could a stranger take your ICP document and build a list from it without asking you a single question? If not, it isn’t finished.

Tier it, don’t binary it

An ICP that only says yes or no throws away information. In practice you’ll find three groups: the businesses that match everything and should get real effort; the ones that match most of it and are worth a lighter touch; and the ones that match on the surface but fail a criterion that matters.

Tiering matters because effort is finite. If personalised research goes into every prospect equally, you either burn time on weak matches or dilute the strong ones. A simple A/B/C split, with a written reason attached to each grade, lets you spend the most on the prospects most likely to reply — and gives you something to argue with later when the results come in. Our ICP Agent works this way: it sources real Australian businesses against your criteria, enriches the contact details, checks each one for compliance, and scores every match into tiers with a reason, rather than handing back an undifferentiated list.

Compliance is part of the profile, not an afterthought

In Australia, who you can contact is a targeting question, not just a legal one. The Spam Act 2003 governs commercial electronic messages, and its requirements — a lawful basis for contact, clear identification of your business, and a working unsubscribe — shape the list before it shapes the email. Addresses harvested indiscriminately, or contacts at businesses with no plausible connection to what you sell, are a compliance risk and a deliverability risk at the same time.

The practical upshot: build compliance checks into the list-building step rather than at send time. We’ve written about this in more detail in how to automate cold outreach without breaking the Spam Act, and it’s worth reading before you scale anything.

Where this feeds into outbound

An ICP earns its keep at the moment it becomes a list someone actually works. The handover looks like this: the profile defines the criteria, the list is built and tiered against those criteria, and then outreach personalises against the reason each prospect scored well. That reason is the raw material for a first line that doesn’t sound generated.

This is why we treat targeting and outreach as one pipeline rather than two products. The Cold Outreach Agent takes the strongest matches from the ICP Agent and researches, drafts, sends and follows up — and because its pricing is based on prospects worked per month, a tighter ICP directly reduces what you spend to get the same number of conversations. You can see how the two fit together on the pricing page; bundling them saves 15%.

The honest limits

A few things worth saying plainly, because they’re usually left out.

  • Your first ICP will be wrong in places. It’s a hypothesis built on a small sample. Treat it as something to disprove over the first few hundred prospects, not a document to laminate.
  • A narrow ICP can be too narrow. If your criteria describe forty businesses in Australia, outbound isn’t your channel — that’s a relationship-selling problem. Check the size of the addressable pool before you commit.
  • Australian data is patchier than US data. Enrichment coverage for smaller local businesses is genuinely thinner, and no tool fixes that entirely. Expect to verify more than the vendor demos suggest.
  • Fit doesn’t equal timing. A perfect-fit business that renewed a contract last month is a bad prospect this month. Triggers help, but some of what looks like poor targeting is just bad luck on timing.

How to keep it honest

Review the profile against reality on a regular cadence — quarterly is plenty for most SMBs. The question isn’t “did we get replies”, it’s “which criteria predicted the good conversations”. If your A-tier and C-tier reply at the same rate, your scoring isn’t measuring anything and needs rebuilding. If a segment you excluded keeps appearing in inbound, that’s evidence too.

The businesses that get outbound working aren’t the ones with the cleverest email templates. They’re the ones who worked out precisely who they’re for, wrote it down in terms a list can be built from, and were willing to change it when the data disagreed. Start with the ICP Agent if you’d rather not build the list by hand, or with a free readiness audit on pricing if you’re not sure outbound is your channel at all.