Last-click attribution was always an incomplete picture. AI has made it a dangerously misleading one. This chart breaks down where conversions actually originate across 100 businesses, and the largest single slice belongs to a category most attribution dashboards do not track at all: AI-influenced conversions. Understanding this gap is the first step toward building a measurement framework that reflects how customers actually make decisions today.

Essential Statistics

• 41 percent of conversions across 100 businesses are AI-influenced but go untracked by last-click attribution models.
• 36 percent of conversions are attributed to direct traffic under last-click models, likely absorbing credit from AI-influenced touchpoints that lack tracking parameters.
• 23 percent of conversions are classified as assisted conversions, capturing some but not all multi-touch influence.
• The AI-influenced, untracked segment is the single largest conversion category, exceeding both direct last-click and assisted conversion shares.
• Combined, AI-influenced and assisted conversions account for 64 percent of total conversions, meaning last-click direct attribution represents only 36 percent of the full picture.

Key Takeaways

• Last-click attribution is actively misrepresenting performance for the majority of conversions. With 41 percent of conversions untracked and another 23 percent only partially tracked, most attribution reports are missing or distorting more than half the customer journey.
• The inflated direct traffic share in last-click models is not organic brand strength. A significant portion of that 36 percent is likely AI-influenced traffic arriving without UTM parameters or referral data, making direct the catch-all bucket for untracked AI touchpoints.
• Assisted conversions at 23 percent confirm that multi-touch influence is real and measurable when the right tracking infrastructure is in place. The gap between 23 percent assisted and 41 percent AI-influenced untracked represents the measurement work still left to do.
• Brands using last-click attribution to make channel budget decisions are systematically underfunding the channels that influence the largest share of conversions.
• AI-influenced conversions are not a future problem to solve. They are the current majority, and every day spent on last-click-only reporting is a day of misallocated budget.

Actionable Insights

• Audit what percentage of your direct traffic arrives without referral data or UTM parameters. A spike in untagged direct sessions, especially from new visitors who convert quickly, is a signal that AI referrals are being absorbed into direct attribution. Cross-reference session sources with your AI citation tracking to estimate the true share.
• Implement UTM parameters on every linkable AI surface where your brand appears. GPT-5.4 already sends nearly half of its citation traffic with trackable parameters. Make sure your URLs are set up to receive and report that data so AI-influenced sessions stop flowing into the direct bucket.
• Move at least one reporting view to a multi-touch attribution model immediately. Data-driven attribution or linear attribution will redistribute credit more accurately across the customer journey. Use it alongside last-click for 90 days to quantify the gap for your specific business before making budget decisions based solely on either model.
• Reframe how you present AI-influenced conversions to leadership. The 41 percent figure in this data represents real revenue that existing reports are not crediting to any channel. Building internal alignment around that gap is what unlocks budget for GEO and AI optimization work.
• Add a post-conversion survey question asking how customers first heard about your brand. Even a simple one-question survey asking whether AI tools played a role in their discovery process will give you qualitative data that fills the gap your attribution stack cannot close on its own.

“Forty-one percent of conversions are AI-influenced and untracked. That is not a rounding error. That is the majority of your marketing story missing from your dashboard. Any business still running last-click attribution as its primary model is making budget decisions on half the data.” – Neil Patel