SourceMedium home pagelight logodark logo
  • Customer Case Studies
  • Blog
  • Request a Demo
  • Request a Demo
What Is SourceMedium
Onboarding
Data Integrations & Inputs
Data Transformation
Data Activation
MTA
MTA
  • MTA Overview
  • MTA Models Reference
  • MTA Features
  • MTA Dash Provisioning
  • MTA FAQs
  • MTA Advanced Documentation
  • Release Notes
MTA

Source Medium Multi-Touch Attribution FAQs

Find the answers to commonly asked questions about Source Medium Multi-Touch Attribution

Source Medium MTA takes multiple data sources reporting many customer journeys, unifies them into a single schema and assesses their quality, then combines the best of those purchase journeys with Marketing Data, Orders Data, Customer Data, and User Inputs to create a Unified Purchase Journey Dataset. View this process in the figure below.

Purchase Journey Data Sources include: Google Analytics, Elevar, Blotout, Littledata, Heap, Amplitude, PostHog, Snowplow, Tapcart
Marketing Data Sources include: Meta / Facebook, Google Ads, Bing Ads, TikTok, Klaviyo
Orders and Customers Data Sources include: Shopify, Chargebee, Stripe
Learn more about the basics of Source Medium MTA in our MTA Overview
  • The attributable value is true when there is at least one touch point, for a given model dimension, on a given customer purchase
  • The attributable value is false (non-attributable) when there are no touch points for a given model dimension, on a given customer purchase
  • Purchases and Revenue are both attributable metrics, for each of the three model dimensions
Example: A pie chart showing Marketing Channel Attributable Revenue is displaying the percentage of Revenue for which at least one touch point exists, when viewing touch points of the Marketing Channel (mapped UTM) type
Learn more about the basics of Source Medium MTA in our MTA Overview

A touch point occurs any time a customer interacts with a Marketing Channel, Ad Creative, or Landing Page during a purchase journey. Touch points are tracked by the time the event occurred, and the interaction that took place. The dimension value of a touch point is the specific Marketing Channel, Ad Creative, or Landing Page that the customer interacted with at that touch point. The Distinct Dimension Values then, of a purchase journey, is the collection of unique Marketing Channels, Ad Creatives, or Landing Pages that a customer interacted with throughout their purchase journey—no dimension value is counted more than once even if a customer interacted with it multiple times.

Example: A customer visits the same landing page 3 separate times on 3 separate dates before making a purchase. This purchase journey contains 3 landing page touch points, but only 1 Distinct Dimension Value for the landing page model dimension, as the landing pages were not unique.
Learn more about the basics of Source Medium MTA in our MTA Overview
  • A purchase journey includes all recorded touch points leading up to a purchase made by a customer, an example is displayed in the figure below
  • Source Medium MTA standardizes purchase journeys to the GA4 E-Commerce Event Schema that you may already be familiar with
  • There are often many data sources reporting many purchase journeys for each purchase, Source Medium MTA selects the highest quality purchase journey available by number of valid touch points—read more on this in the modeling section below
Learn more about the basics of Source Medium MTA in our MTA Overview

Multi-Touch Attribution is a type of marketing data model which accounts for the many marketing interactions each customer has with your business before a purchase, rather than only accounting for a single interaction as many typical marketing models do.

Marketing interactions happen with customers in the form of touch points—many touch points make up a purchase journey, and many purchase journeys make up a multi-touch attribution data set. Click the tabs below to read more about each of these concepts.

Learn more about the basics of Source Medium MTA in our MTA Overview

Source Medium MTA Built-in Reports are part of a standalone dashboard separate from your main Source Medium dash. After MTA is enabled for your account, the dashboard link will be pinned to your shared Slack channel or sent via email/gchat if you do not use Slack.

For a description of each of the default modules and their functionality, see the Built-in Reporting dropdown from section 3 of our MTA Overview
  • Source Medium MTA modeling enables three different attribution types:
    • First Touch: Assigns all credit to the first valid touch point in the purchase journey
    • Last Touch: Assigns all credit to the last valid touch point before the purchase
    • Linear: Distributes credit equally among all valid touch points
Example: A purchase journey in which the customer first interacted with a Google ad, second a TikTok ad, and third a Meta ad before then making a purchase would give credit solely to Google via First Touch Attribution, and would give credit solely to Meta via Last Touch Attribution—but would distribute one third of the credit each to Google, Meta, and TikTok via Linear Attribution.
Learn more about the basics of Source Medium MTA in our MTA Overview

Email and SMS channels are excluded from first touch and linear attribution for several important reasons:

  1. Existing Relationship Required: By definition, Email and SMS require a customer to have already interacted with your brand (to provide their contact information), so they’re rarely the true “first touch” in a customer’s journey.

  2. Marketing Budget Allocation: Including Email/SMS in all attribution models can lead to undervaluing true acquisition channels, causing misallocation of marketing budgets.

  3. Frequent Touches: Email and SMS often have multiple touches in a journey, which can disproportionately influence linear attribution.

Instead, Source Medium provides a dedicated Email/SMS dimension that allows you to analyze the impact of these channels separately, while maintaining more accurate attribution for acquisition channels.

For more details, see our Email & SMS Attribution documentation.

Brand campaigns receive zero attribution credit in Source Medium’s MTA system because:

  1. Incremental Value: Brand campaigns typically target users who would have found you anyway (searching specifically for your brand name).

  2. Last-Click Bias: In traditional models, brand campaigns often receive disproportionate credit simply because they occur late in the purchase journey.

  3. Marketing Budget Optimization: Attributing conversions to brand campaigns can lead to overinvestment in non-incremental marketing and underinvestment in true acquisition channels.

However, brand campaign metrics (spend, impressions, clicks) are still fully visible in the system. This approach gives you complete transparency into your marketing activities while preventing attribution bias.

For more details, see our Brand Campaign Attribution documentation.

Source Medium’s MTA system organizes marketing data at two distinct levels:

Ad-Level Data:

  • Identified by specific ad_id values
  • Contains complete ad metadata and performance metrics
  • Receives attribution credit across all models
  • Represents marketing activities that can be tied to specific creative assets

Channel-Level Data:

  • Identified by sm_marketing_channel (e.g., Facebook, Google)
  • Contains only unattributed metrics not already counted at the ad level
  • Represents marketing activities that cannot be tied to specific ads
  • Prevents double-counting while maintaining complete spend visibility

This dual approach ensures you have both granular ad-level insights and complete channel-level spend accountability.

For more details, see our Channel-Level Attribution & Unattributed Metrics documentation.

The Email/SMS dimension is a specialized attribution dimension specifically designed for tracking and analyzing the impact of email and SMS messages on purchase journeys.

Key features include:

  • Message ID Extraction: Identifies specific messages using utm_id, utm_content, or utm_term parameters
  • Display Format: Uses a standardized [channel]message_id format (e.g., “[email]123456”)
  • Performance Connection: Links to engagement metrics from outbound message performance data
  • Attribution Rules: Has its own specific attribution logic, separate from the marketing channel dimension

This dedicated dimension allows you to analyze email and SMS performance in detail, even while these channels are excluded from first touch and linear attribution in the marketing channel dimension.

For more details, see our Email & SMS Attribution documentation.

To improve your attribution rate (the percentage of purchases with attributable touchpoints), consider these approaches:

  1. Implement Proper UTM Parameters:

    • Ensure all marketing campaigns use consistent UTM parameters
    • Add UTM parameters to email and SMS links
    • Use unique campaign and content identifiers

  2. Connect Additional Data Sources:

    • Integrate all your marketing platforms with Source Medium
    • Enable event tracking on your website and app
    • Consider implementing a Customer Data Platform (CDP)

  3. Optimize Tracking Setup:

    • Verify your Google Analytics configuration
    • Implement server-side tracking where possible
    • Ensure cross-domain tracking is properly configured

  4. Review Attribution Windows:

    • Standard attribution window is 90 days
    • Extending this window may capture more touchpoints
    • Work with your SourceMedium account manager to adjust if needed
For technical assistance with improving attribution rates, contact your SourceMedium account manager.

Yes, Source Medium can implement custom attribution rules tailored to your business model. Common customizations include:

  1. Email/SMS Attribution Rules:

    • Enabling Email/SMS in last-touch attribution
    • Customizing which message types receive attribution
    • Special handling for referral or affiliate emails

  2. Brand Campaign Configuration:

    • Customizing what qualifies as a “brand” campaign
    • Implementing partial attribution for brand campaigns
    • Creating brand-specific attribution models

  3. Attribution Windows:

    • Adjusting lookback periods for specific product categories
    • Setting different windows for different attribution models
    • Configuring business-specific conversion cycles

  4. Custom Dimensions:

    • Creating industry-specific attribution dimensions
    • Implementing custom weighting factors
    • Building customer segment-specific models

To discuss custom attribution configurations, contact your SourceMedium account manager.

Unattributed spend represents marketing activities that cannot be tied to specific ads but still contributes to overall marketing costs. To analyze this spend:

  1. Identify Channel-Level Rows:

    • Look for rows where sm_marketing_channel is present but ad_id is NULL
    • These rows contain only the unattributed portion of each channel’s metrics

  2. Run Analysis Queries:

    Copy
    Ask AI
    -- Unattributed spend by channel
SELECT
  sm_marketing_channel as channel,
  SUM(ad_spend) as unattributed_spend
FROM `sm-{{account_id}}.sm_transformed_v2_ad_attribution_performance_daily`
WHERE 
  sm_store_id = 'your-sm_store_id'
  AND date BETWEEN DATE_SUB(CURRENT_DATE(), INTERVAL 30 DAY) AND CURRENT_DATE()
  AND sm_marketing_channel IS NOT NULL
  AND ad_id IS NULL -- Only channel-level data
GROUP BY 1
ORDER BY 2 DESC

  3. Calculate Unattributed Percentage:

    Copy
    Ask AI
    -- Percentage of spend that is unattributed by channel
WITH channel_totals AS (
  SELECT
    COALESCE(sm_marketing_channel, 'Unknown') as channel,
    SUM(ad_spend) as total_spend
  FROM `sm-{{account_id}}.sm_transformed_v2_ad_attribution_performance_daily`
  WHERE sm_store_id = 'your-sm_store_id'
    AND date BETWEEN DATE_SUB(CURRENT_DATE(), INTERVAL 30 DAY) AND CURRENT_DATE()
  GROUP BY 1
),
unattributed AS (
  SELECT
    sm_marketing_channel as channel,
    SUM(ad_spend) as unattributed_spend
  FROM `sm-{{account_id}}.sm_transformed_v2_ad_attribution_performance_daily`
  WHERE sm_store_id = 'your-sm_store_id'
    AND date BETWEEN DATE_SUB(CURRENT_DATE(), INTERVAL 30 DAY) AND CURRENT_DATE()
    AND sm_marketing_channel IS NOT NULL
    AND ad_id IS NULL
  GROUP BY 1
)
SELECT 
  t.channel,
  t.total_spend,
  COALESCE(u.unattributed_spend, 0) as unattributed_spend,
  SAFE_DIVIDE(COALESCE(u.unattributed_spend, 0), t.total_spend) * 100 as unattributed_percent
FROM channel_totals t
LEFT JOIN unattributed u ON t.sm_channel u.channel
ORDER BY 4 DESC
For more details, see our Channel-Level Attribution & Unattributed Metrics documentation.

Was this page helpful?

Suggest editsRaise issue
MTA Dash ProvisioningMTA Advanced Documentation
twitterlinkedin
Powered by Mintlify
Assistant
Responses are generated using AI and may contain mistakes.