Last-click attribution is one of the most expensive myths in Australian marketing — it tells you the last ad someone clicked before converting deserves all the credit, which is roughly as logical as giving the final salesperson all the commission on a deal that took six months to close. If you’re making budget decisions based on last-click data alone, you’re almost certainly underfunding the channels that actually drive growth and over-rewarding the ones that just happen to be nearby at the finish line. Here’s how to build a measurement framework that reflects reality.
Why Last-Click Attribution Distorts Your Data
In a typical Australian B2C purchase journey, a customer might see a YouTube pre-roll, click a Meta prospecting ad, read a blog post via organic search, receive a retargeting display ad, then finally convert via a branded Google Search ad. Last-click gives 100% of the credit to branded search — which means your Google Search campaign looks like a gold mine and your YouTube and Meta campaigns look like dead weight. You cut those “underperforming” channels, branded search volume drops six weeks later, and you’re left wondering why revenue collapsed.
This pattern plays out constantly. The problem isn’t the channels — it’s the measurement. Google’s own data shows that multi-step journeys involving three or more touchpoints are the norm for purchases above roughly $150 AUD, and that figure is even higher for B2B services and considered purchases like home improvements, financial products or education.
Multi-Touch Attribution Models: A Practical Comparison
Moving beyond last-click doesn’t mean adopting a single “correct” model — it means choosing the model that best fits your sales cycle and testing its assumptions. Here are the main options Australian marketers should understand:
- Linear attribution: Splits credit equally across every touchpoint. Simple, and better than last-click, but it treats a three-second display impression the same as a deliberate branded search click.
- Time-decay attribution: Gives more credit to touchpoints closer to conversion. Sensible for short sales cycles (under two weeks), but punishes upper-funnel activity in longer journeys.
- Position-based (U-shaped): Assigns 40% to first touch, 40% to last touch, and splits the remaining 20% across middle interactions. Works well when you care equally about acquisition and close.
- Data-driven attribution (DDA): Uses machine learning to assign credit based on actual conversion patterns in your account. Available in Google Ads and GA4 once you have sufficient conversion volume (typically 300+ conversions per month). This is the most accurate option when you have enough data.
For most Australian SMEs running Google and Meta campaigns simultaneously, a position-based or data-driven model will give a dramatically more honest picture than last-click. Switch your GA4 attribution model under Admin → Attribution Settings and compare the channel performance reports before and after — the shift is often eye-opening.
Incrementality Testing: The Gold Standard
Attribution models, even good ones, still rely on correlation. The only way to truly measure whether a channel is causing revenue — rather than just appearing in the path — is incrementality testing. The core idea: hold back a percentage of your audience from seeing an ad, then compare conversion rates between the exposed group and the holdout group. The difference is your incremental lift.
Meta’s Conversion Lift tool and Google’s Conversion Lift experiments both support this methodology, and they’re accessible without an enterprise budget. A practical approach for an Australian ecommerce brand spending $5,000–$20,000 per month on Meta: run a 10–15% holdout for four weeks, ensure your holdout group is statistically significant (Meta will flag this), and measure the difference in purchase rate. If your exposed audience converts at 3.2% and your holdout converts at 2.6%, your incremental lift is approximately 23% — meaning Meta is genuinely driving roughly one in four purchases, not just appearing in the path of people who would have bought anyway.
Incrementality testing is particularly valuable in Australia’s competitive retail and services markets where branded search often captures demand that was originally created by social or video activity. Running lift experiments quarterly gives you a defensible basis for channel investment rather than gut feel.
Customer Lifetime Value Changes Everything
Most Australian businesses measure ROI on the first transaction. That’s a mistake the moment you have any meaningful repeat purchase rate or subscription component. A customer acquired via Meta Ads for $85 who makes a single $200 purchase looks like a marginal win. If that same customer makes four purchases over 18 months at an average of $180, their lifetime value is $720 and your acquisition cost looks like a bargain.
To properly measure marketing ROI, you need to track cohort-level CLV by acquisition channel. Connect your CRM or ecommerce platform (Shopify, WooCommerce, or your ERP) to GA4 via the User ID feature, then segment lifetime revenue by the first acquisition source. This tells you which channels are acquiring high-value customers — not just first-time buyers. In practice, many Australian businesses discover that their highest-CLV customers came through organic search or referral, while paid social acquired higher volumes at lower initial value. Neither insight is visible if you’re only measuring cost-per-first-conversion.
A Practical Marketing ROI Dashboard for Australian Businesses
Rather than chasing a single “true ROI” number, build a measurement stack that triangulates from multiple angles:
- In-platform ROAS: Use it for day-to-day optimisation within each channel, but never for cross-channel budget decisions. Each platform overcounts its own contribution.
- GA4 data-driven attribution: Use this for channel-level budget allocation decisions on a monthly basis.
- Revenue cohort analysis by acquisition channel: Review quarterly to understand CLV by source.
- Incrementality tests: Run on your top two channels every six months to validate attribution model assumptions.
- Blended MER (Marketing Efficiency Ratio): Total revenue divided by total marketing spend, tracked weekly. This is your north-star metric — it’s immune to attribution games because it looks at the whole business, not individual channels.
MER is increasingly used by sophisticated Australian direct-to-consumer brands precisely because it’s unfakeable. If your total revenue is $400,000 this month and your combined digital marketing spend is $40,000, your MER is 10x. Track this number over time alongside each channel’s spend share and you can spot when a channel is genuinely contributing versus when you’re just shifting budget around and watching the blended number stay flat.
The team at D1 Marketing uses a combination of GA4 data-driven attribution, MER tracking and periodic incrementality testing across all managed accounts — it’s the measurement foundation that makes every other optimisation decision more reliable. If you want to explore how these approaches apply to your specific channel mix, their digital marketing services include full attribution setup and reporting as part of campaign management.
Frequently Asked Questions
What is the difference between ROAS and marketing ROI?
ROAS (Return on Ad Spend) measures revenue generated per dollar spent on a specific ad platform, typically using that platform’s own attribution. Marketing ROI is broader — it accounts for all marketing costs (including agency fees, creative production and tools) against total attributable revenue or profit. ROAS is useful for in-channel optimisation; ROI is the metric that actually tells you whether your marketing budget is growing the business.
How do I measure ROI across both Google Ads and Meta Ads without double-counting?
The most practical approach for Australian businesses is to use GA4’s data-driven attribution model as your cross-channel source of truth, while tracking your blended Marketing Efficiency Ratio (total revenue ÷ total ad spend) as a sanity check. Avoid adding up the attributed revenue figures from both platforms’ dashboards — they will overlap significantly because both platforms claim credit for the same conversions. A third-party tool like Northbeam or Triple Whale can also deduplicate attribution across channels if your spend justifies the investment.
How many conversions do I need before data-driven attribution is reliable?
Google’s data-driven attribution in GA4 and Google Ads generally requires a minimum of 300 conversions per month across the relevant conversion events to produce statistically meaningful results. Below that threshold, position-based (U-shaped) attribution is a more reliable alternative. Meta’s Advantage attribution tools have similar volume requirements for lift testing — Meta will flag if your holdout group is too small to generate reliable results.
Is last-click attribution ever acceptable to use?
Last-click is acceptable for very short, single-session purchase journeys — for example, an emergency service (plumber, locksmith) where a customer searches, clicks one ad and books immediately. In those cases, there genuinely is only one meaningful touchpoint. For any business with consideration periods longer than a day, repeat customers, or multiple active ad channels, last-click will systematically mislead your budget decisions.
If your current reporting is built on last-click data and you’re not sure what you’re missing, book a free strategy call with the D1 Marketing team on (03) 7048 8863. We’ll audit your attribution setup, show you where your measurement gaps are and recommend a practical framework tailored to your channels and sales cycle — no obligation, no jargon.