Data-Driven Attribution
AnalyticsAlso: DDA · Algorithmic Attribution · ML Attribution
Quick definition
Data-driven attribution is an attribution model that uses machine learning to assign conversion credit across all touchpoints in a customer's journey. Instead of applying a fixed rule like last-click or linear, it analyses patterns across thousands of actual conversion paths to estimate each touchpoint's real contribution.
A customer touches all four, then converts.
Credit distributed
Last-click gives everything to brand search. The upper-funnel work that created the demand is invisible.
How it varies across Australia
Data-driven attribution requires a minimum threshold of conversion volume before the model becomes statistically reliable. Australian mid-market businesses often fall below that threshold in their direct channels, which means the model quietly reverts to a fallback rule. Most businesses using data-driven attribution in Google Ads are not aware when this happens.
See data and tracking maturity across Australian industries →How it differs from rule-based models
Assigns all credit to the final touchpoint before conversion. Fast to compute, blind to upper-funnel.
Splits credit evenly across every touch. Fairer in principle, but treats a brand impression the same as a purchase click.
Uses observed conversion patterns to estimate each touch's causal contribution. Requires volume to be reliable.
Minimum volume threshold appliesWhat it actually means
Every attribution model is answering the same question: which marketing touches actually moved the customer toward buying? Rule-based models answer that question with an opinion baked in at the start. Last-click says the final touch did all the work. Linear says every touch did equal work. Data-driven attribution answers the question with a calculation instead.
The model compares two groups: customers who converted and customers who did not. It looks at which touchpoints appeared more often in converting paths than in non-converting paths, and uses that signal to assign fractional credit. A touchpoint that consistently appears in converting paths gets more credit. One that appears equally in both gets less.
The catch is volume. Machine learning needs enough paths to find a signal. Google Ads requires a minimum of 300 to 3,000 conversions per month depending on the conversion type before data-driven attribution is considered reliable. Below that threshold, the model either silently reverts to a fallback rule or produces noisy credit assignments that look scientific but are not.
For Australian businesses running mid-volume campaigns, this is a real problem. The conversion rate is often fine. The total conversion count is not always enough to feed a stable model across every channel and conversion type you care about.
Data-driven attribution is only as good as the data it is driven by. Below a certain volume, the model is just last-click with extra confidence.
How it shows up
Data-driven attribution shows up as the default model in Google Ads and Google Analytics 4 (GA4). When you look at your channel performance reports and see fractional credit distributed across branded search, generic search, YouTube, display and performance max, you are likely looking at a data-driven attribution output.
It also shows up in disagreements. When your paid search team shows ROAS numbers from Google Ads and your attribution platform shows different numbers, a model difference is often part of the gap. Data-driven attribution in Google Ads weights Google-owned touchpoints by definition, because that is the data it has access to. Cross-channel data-driven attribution from a platform with full funnel visibility is a different and usually more honest number.
The Australian context
Australian advertisers face a structural challenge with data-driven attribution that US counterparts often do not. The population is smaller, conversion volumes are lower, and the average Australian media mix includes channels like radio and out-of-home that do not produce digital touchpoints for the model to learn from.
This means data-driven attribution in Google Ads or Meta Ads tends to see only a portion of what actually caused the conversion. It is modelling the digital path, not the full customer journey. Australian businesses investing in above-the-line media alongside performance channels should treat platform-reported data-driven attribution with particular scepticism, and consider Marketing Mix Modelling (MMM) to account for the channels the click-based model cannot see.
Where people get this wrong
Related terms
Common questions
How is data-driven attribution different from last-click?
Last-click gives all conversion credit to the final touchpoint before the sale. Data-driven attribution uses machine learning to distribute fractional credit across all touchpoints based on which combinations of touches appear more often in converting paths. The practical difference is that upper-funnel channels like display and generic search get more credit under data-driven attribution.
Does Google Ads use data-driven attribution by default?
Yes, Google Ads switched its default to data-driven attribution in 2022. New accounts and new conversion actions default to the data-driven model. If your account does not meet the minimum conversion volume threshold, the model either uses a fallback or produces less stable credit assignments. Check conversion volume in your attribution settings.
What volume do I need for data-driven attribution to be reliable?
Google Ads requires a minimum of 300 conversions in a 30-day period for the data-driven model to run. Some conversion types require more. Below that, the model is either inactive or unreliable. For most Australian mid-market campaigns, this threshold is the limiting factor, not the model itself.
Should I use data-driven attribution or Marketing Mix Modelling?
They answer different questions. Data-driven attribution assigns credit to individual digital touchpoints within a click-based path. Marketing Mix Modelling (MMM) estimates the causal contribution of spend across all channels including offline. For Australian businesses with mixed digital and non-digital media, MMM captures what data-driven attribution structurally cannot see.
Debrief
Get the next one
No spam. No fluff. Just the next article, straight to your inbox.
Keep exploring
About New Rebellion
New Rebellion is a marketing intelligence consultancy. We build tools, score Australian businesses on how their marketing actually performs, and publish Debrief every day. This dictionary is part of how we work in the open.
How we think →