Attribution FAQ & Glossary: Every Term Explained
Marketing attribution has its own dense vocabulary, and the jargon is often where people get stuck before they ever get to strategy. Here’s every term you’ll actually run into, explained the way we explain it to clients — plainly, with the practical implication attached.
This is organized as a reference, not a narrative — jump to the term you need. Where a definition has a common misconception attached to it, we’ve flagged it.
Core Attribution Concepts
Attribution is the practice of assigning credit for a conversion (a sale, lead, signup) to the marketing touchpoints that contributed to it. The word “attribution” alone usually implies “attribution modeling” — the rules used to split that credit.
Touchpoint is any interaction a prospect has with your marketing before converting — a paid ad click, an organic search visit, an email open, a social post view. Not every touchpoint is trackable; untrackable ones (a word-of-mouth mention, a billboard) are part of why attribution is always somewhat incomplete.
Conversion path (also called a customer journey or conversion journey) is the ordered sequence of touchpoints a specific user had before converting. Attribution models operate on conversion paths.
- Conversion window / lookback window: the time period before a conversion during which touchpoints are eligible for credit — commonly 30 or 90 days, configurable per platform.
- Assisted conversion: a conversion where a channel appeared somewhere in the path but wasn’t the final touch — used to spot channels that “assist” more than they “close.”
Attribution Models Explained
Last-click (last-touch) attribution gives 100% of the credit to the final touchpoint before conversion. It’s the default in many platforms because it’s simple and cheap to compute, but it systematically overvalues bottom-of-funnel channels like branded search and undervalues discovery channels like content and social.
First-click (first-touch) attribution gives 100% of the credit to the very first touchpoint in the path. Useful for evaluating which channels generate initial awareness, but it ignores everything that happened to actually close the sale.
Linear attribution splits credit evenly across every touchpoint in the path. It’s fair in a naive sense but assumes every touch mattered equally, which is rarely true.
Time-decay attribution gives more credit to touchpoints closer in time to the conversion, on a sliding scale. It’s a reasonable middle ground for longer sales cycles where recency plausibly matters more.
Position-based (U-shaped) attribution gives a fixed larger share (often 40%) to the first and last touchpoints, splitting the remainder among the middle touches. A common variant, W-shaped attribution, adds a third weighted point at lead creation, common in B2B funnels with a distinct lead-conversion moment.
Data-driven attribution (DDA) uses statistical modeling (commonly a form of Markov chain or Shapley value analysis) to assign credit based on actual observed patterns in the account’s own conversion data, rather than a fixed rule. It’s more accurate in principle but requires meaningful conversion volume to be stable — this is the single most misunderstood requirement in attribution.
Measurement and Data Terms
UTM parameters (Urchin Tracking Module) are tags appended to a URL — source, medium, campaign, term, content — that tell analytics tools where traffic came from. Inconsistent UTM tagging is the most common cause of broken attribution in the accounts we audit.
Multi-touch attribution (MTA) is the umbrella term for any model that distributes credit across more than one touchpoint, as opposed to single-touch models (first-click, last-click).
Media Mix Modeling (MMM) is a statistical approach that measures the impact of marketing spend on outcomes using aggregate, channel-level data (often at a weekly or monthly level) rather than individual user-level tracking. It doesn’t require cookies or user-level data, which makes it increasingly relevant as tracking gets harder, but it’s coarser and slower to produce insight than touchpoint-level MTA.
Incrementality testing measures the true causal lift of a channel by comparing conversion behavior with and without exposure to it (commonly via geo holdouts or PSA/control ad tests). It’s the closest thing to ground truth in marketing measurement, and increasingly used to validate or correct attribution model output.
- Server-side tracking: sending event data from your server rather than the browser, improving accuracy against ad blockers and browser privacy restrictions.
- Consent Mode: Google’s framework for modeling conversions when a user declines cookie consent, filling gaps statistically rather than leaving them blank.
Channel and Platform-Specific Terms
Cross-device tracking is the ability to recognize the same person across multiple devices (phone, laptop, tablet), usually via a logged-in user ID. Without it, a single customer’s journey can appear as several unrelated sessions.
View-through conversion credits a conversion to an ad that was shown (but not clicked) before the user later converted through another path. Common in display and video advertising reporting, and a frequent source of inflated channel performance claims if not clearly labeled.
Self-reported attribution refers to platforms (Google Ads, Meta Ads Manager) reporting conversions based on their own tracking and modeling, independent of a cross-channel view — these numbers routinely overlap and, added together across platforms, exceed 100% of actual conversions.
Business and Reporting Terms
Attributed revenue is revenue assigned to a channel or campaign by a given attribution model — always model-dependent, never an absolute fact, which is why the model used should always be stated alongside the number.
Marketing-qualified lead (MQL) attribution ties attribution to the lead-generation stage rather than the final sale, common in B2B where sales cycles are long and the marketing team’s KPI is lead handoff, not closed revenue.
Customer Acquisition Cost (CAC) by channel divides channel spend by attributed conversions for that channel — a direct downstream output of whatever attribution model is in use, which is why two teams using different models will report different CAC for the same channel.
Emerging and AI-Search-Era Terms
Dark traffic (also called “direct” or untracked traffic) is traffic that arrives without identifiable source data — often from copy-pasted links, apps, or, increasingly, from AI chat interfaces that don’t pass referrer data consistently.
Zero-click search describes a search interaction where the user gets their answer directly in the SERP or an AI Overview without visiting any website — a growing share of queries that attribution models built around clicks and sessions simply can’t see.
Answer engine visibility refers to a brand’s presence and citation within AI-generated answers (AI Overviews, chat assistants) — a form of marketing influence that currently has no standardized attribution method, though brand search lift and direct traffic growth are commonly used as proxies.
Frequently Asked Questions
What's the difference between multi-touch attribution and media mix modeling?
Multi-touch attribution tracks individual user-level touchpoints and assigns credit within a specific conversion path. Media mix modeling works at an aggregate, channel-level using statistical correlation over time, without needing to track individual users.
Is data-driven attribution the same as AI attribution?
They're often used interchangeably in marketing, but "data-driven attribution" specifically refers to statistical models (like Shapley value or Markov chains) applied to your own conversion data — it's a form of applied statistics rather than a general AI system.
What does "assisted conversion" actually tell you?
It tells you a channel showed up somewhere in a conversion path without being the final touch — useful for identifying channels that build awareness or consideration even if they rarely close the sale directly.
Why do self-reported numbers from different ad platforms never add up?
Each platform tracks and claims credit for conversions independently, often using generous attribution windows and view-through credit, so overlapping claims across platforms routinely sum to well over 100% of actual total conversions.
What is dark traffic and why does it matter more now?
Dark traffic is untracked visits with no identifiable source, and it's growing because AI chat interfaces, apps, and privacy-focused browsers often don't pass the referrer data attribution models rely on.
Do I need to know all these models to run attribution well?
No — most businesses need to understand three or four (last-click, first-click, data-driven if volume supports it, and position-based as a middle ground) plus the data-quality terms like UTM parameters and cross-device tracking.
Terry has 30+ years in software and SEO. He’s the founder of Salterra Digital Services and SEO Spring Training, host of the Roundtable SEO Mastermind, and lead instructor at SEO University — teaching the exact tactics his team uses on client work.
Continue learning
Ready to master this?
This guide is one lesson from the Analytics, Measurement & Attribution course. Get every lesson, framework and checklist — plus the full 38-course catalog — inside SEO University.
Practitioner-focused training across the full digital marketing stack — from technical SEO to conversion optimization and the AI search era. By Salterra Digital Services, since 2011.