Performance marketing has become one of those terms that’s lost most of its meaning through overuse. Every agency in Melbourne uses it. Most are describing different things when they do.
If you’re paying for digital advertising right now, the test is whether you can explain what you’re actually buying without using the agency’s words for it. Most business owners can’t. This is the guide worth reading before the next agency meeting.
A payment model, before anything else.
Traditional advertising charges for placement. You pay for the billboard whether anyone looks at it. Performance marketing flips that. You only get charged when a measurable event happens, typically a click, a lead form submission, or a sale, and every charge on the invoice should match an event you can find in your own analytics.
That’s the simple version. The complicated version is that the model only works if the measurement underneath is sound. Conversion tracking has to fire correctly. Pixels need to be installed properly. Attribution has to figure out who gets credit when a customer clicks a Google ad on Tuesday, sees an Instagram retargeting ad on Wednesday, and converts from an email link on Friday.
Plenty of agencies sell campaigns labelled performance marketing, where this scaffolding is broken or missing. The bill arrives the same regardless.
Most Melbourne businesses run performance campaigns across some combination of the following.
Google Ads is the usual starting point, with Search, Performance Max, Shopping, and YouTube. The intent signal is clearer here than anywhere else. Someone typing “emergency plumber Brunswick” at 11 pm is closer to buying than almost any audience you can reach on another channel.
Meta covers Facebook and Instagram. It works on behavioural signals and creative discovery rather than search intent. Strongest for products that benefit from being shown to people who weren’t actively looking but should have been.
LinkedIn is the B2B option. CPCs run high, often $8 to $15 in Australian B2B markets, but the maths usually works once your average client is worth $10,000 or more.
TikTok works for under-35 audiences and product launches, where ads that don’t look like ads outperform polished ones.
Connected TV (CTV) used to sit firmly in the brand bucket. Smart TVs and streaming devices now allow household-level targeting and view-through measurement, which has pulled it into the performance category for advertisers spending around $10k a month or more on CTV alone. Below that floor, the data isn’t dense enough to optimise against.
Retail Media Networks (RMNs) are where attributable budget tends to migrate if you sell physical products. Amazon was first. Coles and Woolworths in Australia are catching up. The retailer can see the actual purchase happen on their own platform, which closes the attribution gap that other channels are still arguing about.
Affiliate marketing belongs in this list technically, but it’s a structural sub-category rather than a channel to build a strategy around in isolation.
Six acronyms cover most of what shows up on a report.
CPM is Cost Per Mille, is what one thousand ad impressions cost. Mostly a context number. Not something to optimise against directly.
CPC, Cost Per Click, is the cost of getting one person from an ad to your website. Useful for diagnosing whether the creative is earning attention.
CPL is the cost of one form submission, demo request or phone call. Most service businesses run their accounts against it.
CPA, Cost Per Acquisition, is the cost of one paying customer. Usually, the number that decides whether a campaign keeps running or gets killed.
ROAS divides revenue by ad spend. An eCommerce retailer making $4 in revenue for every $1 spent has a ROAS of 4. Whether that’s profitable depends on margin; a 4x ROAS on a 20% margin product loses money.
MER, Marketing Efficiency Ratio, divides total revenue by total marketing spend. It’s increasingly the number senior marketers report against, because channel-level attribution has become unreliable enough that adding the channels up doesn’t reflect what’s actually driving sales.
A useful piece of arithmetic before signing any agency contract: work out average gross margin per sale, then customer lifetime value if there’s repeat purchase data. Gross margin is the absolute ceiling on CPA. LTV is the realistic ceiling. Spending above either means paying to lose money on growth.
Worth being precise here because the terms get blurred.
Digital marketing is everything done online, including performance, brand, content, SEO, email, and organic social. Performance marketing is the subset measured against direct outcomes. Brand marketing is the subset focused on perception and recall, which doesn’t fit cleanly into a CPA report.
Be wary of agencies that pretend brand and performance are the same activity. They aren’t. They do support each other: stronger brand awareness brings CPAs down, because ads convert better when people already recognise the name. But trying to optimise for both with one set of metrics tends to do both badly.
Rough guide: businesses under about $50k monthly revenue usually get more out of starting with performance, because the feedback loop is fast and the maths is provable. Past that point, brand investment starts paying back, but the returns show up in places like assisted conversions, branded search volume, and direct traffic, not in your Google Ads dashboard.
Three things, mostly.
First, the platforms have absorbed most of the manual levers. Google Performance Max and Meta Advantage+ now handle bidding, audience selection, placement and creative rotation algorithmically. Work that used to involve building 30 ad sets and writing custom bid rules now involves giving the system a clean conversion signal, a varied creative library, and getting out of the way. Agencies still pitching granular keyword-level bid management as their main value-add are usually behind where the platforms have moved.
Second, creative volume has changed. Generative AI has dropped the cost of producing an ad variation close to zero. A campaign that used to ship four variations per quarter now ships forty. That makes creative testing the highest-leverage thing in most accounts, ahead of audience targeting, placement, or bid strategy. If your agency is producing the same three ads they made in January, the account is being underserved.
Third, measurement has changed, and not in a way anyone has fully recovered from. iOS 14.5 dropped four years ago. The fallout is still working through the system. Browser cookie restrictions, Apple’s signal loss, and the slow death of third-party cookies have together made the attribution dashboards from 2019 a fiction. The way around it is server-side tracking (CAPI for Meta, Enhanced Conversions for Google), some form of multi-touch or media-mix modelling, and reporting against MER rather than just channel-level ROAS. If nobody on your account has talked to you about server-side tracking, that’s the question to ask on the next call.
There’s also a fourth change worth a quick mention. Shoppable video and social commerce have largely killed the discovery-to-purchase gap inside TikTok and Instagram. You can buy a product without leaving the video that showed it to you. The implication for performance campaigns is that the creative does most of the conversion work now, because there often isn’t a separate landing page in the funnel.
Worth being clear about the limits, because the agency pitches usually aren’t.
It won’t generate demand that doesn’t exist. If the product is in a category nobody’s looking for yet, performance channels can prove that for $20k of media spend, but they can’t solve it. That’s a content, PR and brand investment problem.
It won’t compensate for an underpriced or overpriced product. Pricing problems show up as conversion rate problems, and no amount of bid optimisation fixes them.
It won’t run on broken tracking. A business with conversion tracking firing on the wrong events is guessing about its CPA, scaling whatever’s making the dashboard look good, and usually realising the mistake six figures of spend later.
It won’t grow forever on its own. Performance budgets that scale year-on-year without parallel brand investment hit a ceiling. Every available high-intent searcher in the category has effectively been bought, so each additional dollar starts trying to convince people who weren’t going to buy yet. That’s the most common reason a campaign that worked for two years suddenly stops working in year three.
Hiring help, these are the signals worth weighing, in roughly this order.
Are they specific in the discovery call? An agency that asks for current CPA, average order value, gross margin and repeat purchase rate is operating at the right level. One that talks about awards, team size and “growth” without referencing the actual numbers is selling something else.
Do they have a real answer on tracking? Ask how they handle iOS attribution loss and whether they implement server-side conversion APIs. The quality of the answer is one of the better signals you’ll get in a 30-minute pitch.
Are they Google Premier Partner certified? It’s not everything; some excellent independent operators don’t bother with the certification, but it filters out the weaker end. Premier status puts an agency in roughly the top 3% of Google Ads partners in Australia.
Can they explain the reporting in plain language? If the dashboard needs an agency translator, the relationship is locked in by design. The reports worth respecting show what was tested last week, what won, what’s being changed this week, and how each number compares to the goal.
What’s their position on brand? An agency that says “we don’t do brand, only performance” is fine for businesses with brand investment running elsewhere. Without that, they’ll optimise the account toward the same ceiling described earlier.
Is performance marketing right for a B2B service business?
Usually yes, with one caveat. B2B sales cycles distort the feedback loop. A campaign generating leads in March that close in August needs to be evaluated on lead quality and pipeline value, not immediate conversions. A lot of agencies aren’t set up to report that way, which is something to test in the pitch.
What’s a sensible starting budget?
Below about $3,000 a month in media spend, performance channels generally don’t generate enough data to optimise against. For e-commerce, the floor is closer to $5,000. Anything below those numbers is research spend, not a performance program.
How long until the results show?
Eight to twelve weeks before there’s meaningful data. Three to six months before the campaign is properly optimised. Anyone offering guaranteed results in 30 days is either lucky, mistaken, or being economical with the facts.
Can a business owner run this themselves?
Some do it well. The honest constraint is time. A competent owner-operator running their own campaigns spends six to ten hours a week on it, once you account for creative iteration, reporting and platform changes. Whether it’s worth doing comes down to what those hours are otherwise worth.
Performance marketing in 2026 rewards businesses that treat it as an operational discipline rather than an outsourced function. That means real KPIs, agencies that answer questions in numbers, and enough fluency in your own data to push back when something looks off.
That’s the work Performancemarketer does, building accountable performance programs for Melbourne businesses that want measurable returns rather than reassuring reports. If a sharper conversation about your current spend would be useful, get in touch.
