Glossary

Attribution, marketing analytics, and API terminology used throughout the documentation.

A B C F G H I J L M P R S T U W Z

A

AOV: Average Order Value — total revenue divided by the number of orders over a given period.
Attribution: The process of assigning credit for a conversion (sale, signup) to the marketing touchpoints that influenced it.

B

Bearer Token: An authentication credential sent in the HTTP Authorization header. Attribution.ai API keys are bearer tokens prefixed with "atai_".

C

Campaign: A specific marketing initiative within a channel, usually identified by UTM parameters or ad platform campaign IDs.
Causal Inference: Statistical methods that estimate the causal effect of marketing spend, as opposed to mere correlation. Includes techniques like Granger causality and holdout tests.
Channel: A marketing source that drives traffic and conversions — e.g., Google Paid, Facebook, Email, Organic, Direct, TikTok.
Conversion: A desired action completed by a customer, typically an order or purchase in e-commerce attribution.
CPA: Cost Per Acquisition — total marketing spend divided by the number of conversions attributed to that spend.

F

First-Touch Attribution: A model that assigns 100% of conversion credit to the first marketing touchpoint in the customer journey.
Fractional Attribution: An attribution approach that distributes credit across multiple touchpoints rather than assigning all credit to a single interaction.

G

Granger Causality: A statistical test that determines whether one time series (e.g., ad spend) is useful in forecasting another (e.g., revenue). Used in incrementality analysis.

H

HMAC: Hash-based Message Authentication Code — used to verify webhook payloads. Attribution.ai uses HMAC-SHA256 for Shopify webhook verification.
Holdout Test: An experiment where a subset of users is excluded from seeing ads to measure the incremental impact of advertising.

I

Incrementality: The additional conversions or revenue caused by marketing that would not have occurred otherwise. Measured via experiments and causal inference.

J

JSON-RPC: A remote procedure call protocol using JSON. Attribution.ai's MCP server uses JSON-RPC 2.0 over HTTP.

L

Last-Touch Attribution: A model that assigns 100% of conversion credit to the last marketing touchpoint before conversion.
Lift Test: An experiment that measures incremental impact by comparing exposed vs. unexposed groups.

M

MCP: Model Context Protocol — an open protocol for connecting AI assistants to external tools and data sources. Attribution.ai implements an MCP server.
MMM: Media Mix Modeling — a statistical technique that uses historical data to estimate the impact of different marketing channels on business outcomes like revenue.
Multi-Touch Attribution: Attribution models that distribute conversion credit across multiple marketing touchpoints in the customer journey, rather than giving all credit to one.

P

Pixel: A lightweight JavaScript snippet embedded on web pages that tracks user behavior (page views, sessions, UTM parameters) for attribution purposes.

R

ROAS: Return On Ad Spend — revenue generated divided by advertising cost. A ROAS of 4.0x means $4 revenue per $1 spent.

S

Session: A group of user interactions on a website within a given time frame, typically 30 minutes of inactivity. Sessions help track customer journeys.
Survey Attribution: Attribution data collected directly from customers via post-purchase surveys, such as "How did you hear about us?"

T

Touchpoint: Any interaction between a customer and a marketing channel — an ad click, email open, social media visit, etc.

U

UTM Parameters: URL parameters (utm_source, utm_medium, utm_campaign, utm_term, utm_content) used to track the effectiveness of marketing campaigns.

W

Webhook: An HTTP callback that sends real-time notifications when events occur. Attribution.ai receives Shopify webhooks for order and customer events.

Z

Z-Score: A statistical measure indicating how many standard deviations a data point is from the mean. Used in anomaly detection to identify unusual revenue or order patterns.

Was this page helpful?

Error Reference Changelog

Campaign

A specific marketing initiative within a channel, usually identified by UTM parameters or ad platform campaign IDs.

Causal Inference

Statistical methods that estimate the causal effect of marketing spend, as opposed to mere correlation. Includes techniques like Granger causality and holdout tests.

Channel

A marketing source that drives traffic and conversions — e.g., Google Paid, Facebook, Email, Organic, Direct, TikTok.

Conversion

A desired action completed by a customer, typically an order or purchase in e-commerce attribution.

CPA

Cost Per Acquisition — total marketing spend divided by the number of conversions attributed to that spend.

MCP

Model Context Protocol — an open protocol for connecting AI assistants to external tools and data sources. Attribution.ai implements an MCP server.

MMM

Media Mix Modeling — a statistical technique that uses historical data to estimate the impact of different marketing channels on business outcomes like revenue.

Multi-Touch Attribution

Attribution models that distribute conversion credit across multiple marketing touchpoints in the customer journey, rather than giving all credit to one.