Multi-touch attribution overclaims revenue because it distributes credit for each conversion across multiple touchpoints. When you sum the attributed revenue across all channels, the total exceeds your actual revenue. This is double counting, and it affects almost every MTA system.
If your total revenue for the month is £500,000 and you add up the attributed revenue reported by each channel in your MTA tool, you will almost certainly get a number larger than £500,000. The difference is the overclaim.
Why overclaiming happens
When a customer converts after touching three channels, an MTA model distributes credit across all three. If each channel is credited with 50 percent of the conversion value, the total attributed revenue is 150 percent of the actual conversion value. Multiply this across all your conversions and the overclaim compounds quickly.
The overclaim is not a bug or a deliberate distortion. It is a structural consequence of distributing credit fractionally across touchpoints rather than assigning the full conversion value to exactly one.
The platform attribution problem
Overclaiming gets worse when each platform runs its own attribution model independently. Meta credits its touchpoints using Meta's data. Google credits its touchpoints using Google's data. Both claim credit for the same conversion because they each saw a touchpoint they associate with that buyer.
- Meta's ad manager shows 200 purchases attributed to a campaign
- Google Ads shows 180 purchases attributed to search campaigns running simultaneously
- Your actual shop recorded 220 purchases in total
- The sum of platform reported conversions is 380, which is 73 percent more than the actual number
Why this matters for budget decisions
If you allocate budget based on attributed revenue reported by each platform or each MTA channel, you are optimising against inflated numbers. Channels that appear in many conversion paths will receive disproportionate credit, regardless of their actual causal contribution.
You may increase budget to channels with high attributed ROAS, not realising that much of their attributed revenue would have converted through other channels or organically. This is how brands overspend on retargeting while underspending on channels that generate genuine new demand.
The fix is not to find a better MTA model. Every MTA model will overclaim by design when it distributes credit fractionally. The fix is to triangulate MTA signals with incrementality tests and marketing mix modelling, which measure causal contribution rather than credit distribution.
How to identify overclaiming in your data
Sum your attributed revenue across all channels in your MTA tool and compare it to your actual total revenue for the same period. The ratio is your overclaim factor. An overclaim factor above 1.2, meaning attributed revenue is 20 percent above actual, is common. Above 1.5 is a serious signal that the model is not reliable for budget decisions.
Working with MTA data responsibly
MTA data is useful for understanding relative channel importance and customer journey patterns. It is not reliable for setting budget targets or evaluating absolute channel effectiveness. Use MTA for directional insights and journey analysis, and use experimental measurement for budget decisions.
Is data-driven attribution less likely to overclaim than linear or time-decay models?
Data-driven attribution is more accurate within its data set, but it still overclaims for the same structural reason: credit is distributed fractionally across touchpoints. The overclaim may be smaller with DDA because it weights touchpoints more carefully, but it does not eliminate the problem.
Can I normalise MTA data to remove the overclaim?
You can apply a normalisation factor to bring total attributed revenue in line with actual revenue. This preserves the relative proportions across channels while removing the aggregate overclaim. Some MTA providers do this automatically. The limitation is that normalisation does not tell you whose attribution numbers are too high and whose are too low.
Is MTA worth using at all if it overclaims?
Yes, with the right expectations. MTA provides useful signals about which channels appear most frequently in conversion paths and how customer journeys are structured. These are valuable inputs for understanding your audience. The mistake is using attributed revenue totals as if they represent causal contribution.