Three platforms claim the same order. That can happen because each applies its own attribution rules, window and identity logic. Attribution describes how a report allocates credit among eligible interactions. It does not tell you whether the ads caused the purchase. This page explains common models, click and view reporting, lookback windows and cross-device matching. Use the definitions to reconcile reports first, then bring a separate incrementality method to the causal question.
🧭 Jump to a term
What attribution assigns: Attribution · Attribution model · First-touch attribution · Last-touch attribution
How models divide credit: Linear attribution · Time-decay attribution · Position-based attribution · Multi-touch attribution (MTA) · Data-driven attribution (DDA)
Eligibility and conversion paths: Click-through conversion · View-through conversion · Attribution window (lookback window) · Conversion lag · Cross-device attribution · Self-attributing network (SAN) · Assisted conversion
⭐ Know these first
Start with Attribution, Last-touch attribution, Multi-touch attribution (MTA), View-through conversion, Attribution window (lookback window). Then follow the grouped learning order below.
📎 How to read this page
What it means gives the precise meaning. Operator translation gives the version you might hear in a real ecommerce meeting. In real life shows an illustrative example. Watch out and the confusion boxes show where a familiar term can mislead.
📈 Read the relationships first
These combinations are diagnostic hypotheses, not proof of causality. Compare the same period and scope, then investigate the mechanism.
📣 Platform attributed orders ↑ + store orders flat
Usually means: Reports may use different windows, interaction rules or overlapping paths. Check next: Reconcile the definitions before comparing totals.
🖱️ Click-through credit ↑ + view-through credit flat
Usually means: A campaign may be gaining direct response, or click reporting may have changed. Check next: click eligibility and conversion lag.
🔍 Lookback window ↑ + attributed conversions ↑
Usually means: A longer window includes more prior interactions. Check next: Compare the setting and avoid treating the increase as incremental lift.
📱 Cross-device paths ↑ + match coverage ↑
Usually means: More signed-in or consented activity may be connected. Check next: identity coverage and regional privacy settings.
What attribution assigns
01 · 🟢 Core
Attribution = Allocation of conversion credit across eligible interactions
🧠 What it means
Allocation of conversion credit across eligible interactions
💬 OPERATOR TRANSLATION
“Attribution assigns reporting credit to selected marketing interactions under defined rules. It does not establish that an interaction caused the conversion.”
🛍️ In real life
A shopper sees social creative, searches the brand, then orders; a report assigns credit according to its chosen rules.
🔗 Related: Attribution model · First-touch attribution · Last-touch attribution · ↑ all terms
02 · 🔵 Operations
Attribution model = Rule or method used to allocate conversion credit
🧠 What it means
Rule or method used to allocate conversion credit
💬 OPERATOR TRANSLATION
“An attribution model defines how eligible interactions share reported conversion value. Models describe credit allocation and can produce different answers from the same path.”
🛍️ In real life
A team compares a last-touch report with a model that distributes credit across several interactions.
🔗 Related: Attribution · First-touch attribution · Last-touch attribution · ↑ all terms
03 · 🔵 Operations
First-touch attribution = Credit assigned to the earliest eligible interaction
🧠 What it means
Credit assigned to the earliest eligible interaction
💬 OPERATOR TRANSLATION
“First-touch attribution assigns all or a specified share of credit to the earliest qualifying touchpoint in the measured journey. Eligibility and identity rules still matter.”
🛍️ In real life
A shopper first discovers the brand through a search ad, then later purchases after email; the search ad receives first-touch credit.
🔗 Related: Attribution · Attribution model · Last-touch attribution · ↑ all terms
04 · 🟢 Core
Last-touch attribution = Credit assigned to the latest eligible interaction before conversion
🧠 What it means
Credit assigned to the latest eligible interaction before conversion
💬 OPERATOR TRANSLATION
“Last-touch attribution assigns conversion credit to the final qualifying touchpoint under the report’s rules. Direct traffic handling and channel eligibility can change what counts as last.”
🛍️ In real life
A shopper clicks a branded search result immediately before purchase; search receives last-touch credit under the selected setup.
🔗 Related: Attribution · Attribution model · First-touch attribution · ↑ all terms
How models divide credit
05 · 🔵 Operations
Linear attribution = Credit distributed evenly across eligible interactions
🧠 What it means
Credit distributed evenly across eligible interactions
💬 OPERATOR TRANSLATION
“Linear attribution divides a conversion’s credit evenly among the eligible interactions included in the path. It does not estimate each touchpoint’s causal contribution.”
🛍️ In real life
Four qualifying touchpoints receive one quarter of the reported credit each.
🔗 Related: Time-decay attribution · Position-based attribution · Multi-touch attribution (MTA) · ↑ all terms
06 · 🔵 Operations
Time-decay attribution = More credit assigned to interactions closer to conversion
🧠 What it means
More credit assigned to interactions closer to conversion
💬 OPERATOR TRANSLATION
“Time-decay attribution applies a rule that weights later eligible interactions more heavily than earlier ones. The decay curve and included touchpoints depend on implementation.”
🛍️ In real life
A model gives the email clicked one day before purchase more credit than an ad seen two weeks earlier.
🔗 Related: Linear attribution · Position-based attribution · Multi-touch attribution (MTA) · ↑ all terms
07 · 🔵 Operations
Position-based attribution = Credit assigned by a rule emphasizing selected journey positions
🧠 What it means
Credit assigned by a rule emphasizing selected journey positions
💬 OPERATOR TRANSLATION
“Position-based attribution gives specified journey positions, often first and last, larger shares and distributes the remainder among other eligible touches. Exact shares vary.”
🛍️ In real life
A team assigns larger credit shares to first discovery and final click, with the rest spread across middle interactions.
🔗 Related: Linear attribution · Time-decay attribution · Multi-touch attribution (MTA) · ↑ all terms
08 · 🟢 Core
Multi-touch attribution (MTA) = Attribution using multiple interactions in a conversion path
🧠 What it means
Attribution using multiple interactions in a conversion path
💬 OPERATOR TRANSLATION
“Multi-touch attribution assigns credit to more than one eligible interaction. The method may be rules-based or model-based; multiple touches do not make the result causal.”
🛍️ In real life
A retailer’s report distributes one order’s credit across paid search, email and social interactions.
🔗 Related: Linear attribution · Time-decay attribution · Position-based attribution · ↑ all terms
09 · 🔵 Operations
Data-driven attribution (DDA) = Model-based allocation of credit informed by observed data
🧠 What it means
Model-based allocation of credit informed by observed data
💬 OPERATOR TRANSLATION
“Data-driven attribution estimates credit using an analytical model and available interaction and conversion data. Inputs, eligibility, privacy limits and platform implementation affect results.”
🛍️ In real life
A platform estimates channel credit from observed paths and its own model requirements.
🔗 Related: Linear attribution · Time-decay attribution · Position-based attribution · ↑ all terms
Eligibility and conversion paths
10 · 🔵 Operations
Click-through conversion = Conversion associated with an eligible ad click
🧠 What it means
Conversion associated with an eligible ad click
💬 OPERATOR TRANSLATION
“A click-through conversion is reported after an eligible interaction with an ad, within the platform’s attribution and timing rules. It does not alone show the ad caused the order.”
🛍️ In real life
A customer clicks a paid social ad and buys later inside the configured reporting window.
🔗 Related: View-through conversion · Attribution window (lookback window) · Conversion lag · ↑ all terms
11 · 🟢 Core
View-through conversion = Conversion associated with an eligible ad impression without a recorded click
🧠 What it means
Conversion associated with an eligible ad impression without a recorded click
💬 OPERATOR TRANSLATION
“A view-through conversion is attributed after an eligible ad impression and a later conversion without a qualifying click, under platform rules. It is especially sensitive to window and eligibility choices.”
🛍️ In real life
A shopper sees a display ad, does not click, and later purchases within the platform’s view-through window.
🔗 Related: Click-through conversion · Attribution window (lookback window) · Conversion lag · ↑ all terms
12 · 🟢 Core
Attribution window (lookback window) = Period in which prior interactions can receive conversion credit
🧠 What it means
Period in which prior interactions can receive conversion credit
💬 OPERATOR TRANSLATION
“An attribution window sets how far back a system considers eligible interactions before conversion. Platforms may offer separate windows by interaction type and reporting setting.”
🛍️ In real life
A seven-day click window includes an eligible click seven days before the order but excludes an older one.
🔗 Related: Click-through conversion · View-through conversion · Conversion lag · ↑ all terms
13 · 🔵 Operations
Conversion lag = Elapsed time between a defined interaction and conversion
🧠 What it means
Elapsed time between a defined interaction and conversion
💬 OPERATOR TRANSLATION
“Conversion lag is elapsed time from a specified marketing interaction to conversion. State which interaction starts the clock and whether the measure uses click, impression or first exposure.”
🛍️ In real life
A customer purchases five days after first clicking a campaign link.
🔗 Related: Click-through conversion · View-through conversion · Attribution window (lookback window) · ↑ all terms
14 · 🔵 Operations
Cross-device attribution = Association of interactions and conversions across devices
🧠 What it means
Association of interactions and conversions across devices
💬 OPERATOR TRANSLATION
“Cross-device attribution links eligible activity on different devices using the platform’s identity and privacy methods. Coverage is incomplete and varies by consent, login and measurement setup.”
🛍️ In real life
A shopper researches on mobile and later buys on a signed-in desktop account.
🔗 Related: Click-through conversion · View-through conversion · Attribution window (lookback window) · ↑ all terms
15 · 🔵 Operations
Self-attributing network (SAN) = Ad platform that reports conversions attributed within its own system
🧠 What it means
Ad platform that reports conversions attributed within its own system
💬 OPERATOR TRANSLATION
“A self-attributing network receives conversion signals and reports its own attribution according to platform rules. Comparing SAN reports requires reconciling windows, event definitions and overlap.”
🛍️ In real life
Two ad platforms each report the same order as attributed to their campaign.
🔗 Related: Click-through conversion · View-through conversion · Attribution window (lookback window) · ↑ all terms
16 · 🔵 Operations
Assisted conversion = Conversion path interaction that preceded a conversion but was not the final credited touch
🧠 What it means
Conversion path interaction that preceded a conversion but was not the final credited touch
💬 OPERATOR TRANSLATION
“An assisted conversion describes an earlier interaction present in a path before conversion under a reporting method. It depends on path completeness and does not prove incremental effect.”
🛍️ In real life
Email introduces a shopper who later converts through a branded search click.
🔗 Related: Click-through conversion · View-through conversion · Attribution window (lookback window) · ↑ all terms
🔀 Attribution vs incrementality
Attribution allocates reported credit to interactions. Incrementality estimates whether outcomes would have changed without the intervention.
🔀 Click-through vs view-through
Click-through includes a qualifying ad click. View-through may include an impression with no recorded click. Their windows and eligibility differ.
🔀 Attribution window vs affiliate cookie duration
A general marketing lookback window is not automatically an affiliate cookie term. Affiliate program rules define their own tracking and commission duration.
🔀 Platform ROAS vs business return
A platform’s credited revenue follows its attribution scope. Reconcile order totals, returns and costs before treating it as a business-wide return.
🤔 Still confused?
Follow this thread: Attribution → Linear attribution → Click-through conversion. That sequence moves from the basic object or relationship to the decisions and checks it supports.
Sources and scope
Primary documentation checked on 27 September 2026. Platform features and eligibility can change; examples and cartoon situations are illustrative.
