What you will learn in this article:
What the Adaptive Attribution model is and how it solves the limitations of rule-based and black-box platform models.
The 5 mathematical factors used to determine exact touchpoint weights for every conversion.
How to activate, inspect, and evaluate Adaptive Attribution inside the Admetrics Data Studio.
Practical strategies to optimize ad spend using data-driven attribution insights.
Who this article is for
This documentation is designed for e-commerce growth leaders, performance marketers, media buyers, and data analysts who manage multi-channel ad budgets and analyze campaign performance.
It directly addresses key marketing and reporting challenges, including:
The "vampire ad" trap: Retargeting and brand search campaigns claiming 100% conversion credit for revenue generated by top-of-funnel prospecting.
Platform self-reporting bias: Ad networks (such as Meta or Google) double-counting conversions and grading their own performance.
Arbitrary lookback windows: Global 7-day, 30-day, or 90-day windows that amputate long-memory channels (e.g., influencer, YouTube) while over-crediting short-lived touches (e.g., brand search).
Black-box machine learning: Closed models that assign conversion credit behind closed doors without providing auditable proof or touchpoint-level signal breakdowns.
Disclaimer: Adaptive Attribution is available by default for data from July 29, 2026 onwards. For earlier periods, Admetrics uses the Linear Attribution model. If you encounter any issues or if anything remains unclear, please reach out to our team at [email protected] or via the in-app chat.
What are the benefits
Unlike traditional rule-based or platform-native attribution, Adaptive Attribution provides a balanced, mathematical representation of how your marketing channels contribute to revenue.
100% credit allocation integrity: Every conversion distributes exactly 100% of order value across its touchpoints. No credit is invented, double-counted, or lost.
Fully inspectable and transparent: Every weight assigned to a touchpoint is mathematically derived and open to inspection.
Customer-type intelligence: Automatically adapts attribution logic based on whether a customer is New, Returning, or Reactivated.
Quality and engagement weighted: Sessions with high engagement (page depth, product views, add-to-carts) earn significantly more credit than superficial single-page bounces.
Dynamically learned multipliers: Account-specific parameters learn the true contribution of prospecting, retargeting, and brand search directly from your customer journey data.
How to use it
Step 1: Access the Data Studio
Log into your Admetrics account.
Select a view you want to report on from the left navigation menu.
Step 2: Select the Adaptive Attribution model
Click on the Attribution Model dropdown menu at the top of your dashboard.
From the list of available models, choose "Adaptive".
Step 3: Compare against legacy models
To evaluate how Adaptive Attribution shifts credit compared to traditional models:
Use the Attribution Comparison tool in Data Studio.
Select a legacy attribution model alongside Adaptive Attribution to identify over-funded or under-credited campaigns.
Requirements
To utilize the Adaptive Attribution model in your account, ensure the following prerequisites are met:
Active Admetrics pixel integration: Your store must have the Admetrics pixel installed and actively tracking user sessions and conversion events.
Cross-channel UTM standardization: Consistent UTM parameter structure across all active campaigns.
Sufficient historical data: Adaptive Attribution is available for data from July 29, 2026 onwards. Model accuracy continues to improve as more customer journey data is collected.
How Adaptive Attribution calculates touchpoint weights
Stage 1: The data preprocessing pipeline
Before calculating weights, raw journey data passes through a 5-step cleaning pipeline:
Noise Filtering - Strips bot traffic, payment gateway redirects, self-referrals, bounces
Customer Type - Classifies conversion as New, Returning, or Reactivated
Journey Gap Split - Detects activity silences and drops stale pre-gap touchpoints
Direct Resolution - Stitches direct visits back to originating parent touchpoints
Deduplication - Merges same-day/same-channel hits into a single consolidated touchpoint
Stage 2: The 5-signal weighting formula
Once cleaned, every touchpoint weight is determined by multiplying five core factors:
Factor | Description | Example Impact |
Position | Evaluates structural position in the journey based on customer type. | Discovery touches carry higher weight for New buyers; trigger touches carry higher weight for Returning buyers. |
Recency | Applies exponential decay based on channel-specific memory half-lives. | A Brand Search touch decays over hours; an Influencer video touch retains memory over several weeks. |
Quality | Factors in session engagement depth (time on site, pageviews, add-to-carts). | A 3-minute visit with 4 page views and an add-to-cart receives significantly higher weight than a 5-second bounce. |
Adaptive Multiplier | Account-learned factor representing a channel's real role. | Prospecting channels are credited fairly for intent creation rather than losing credit to organic search. |
Momentum | Rewards touchpoints that trigger subsequent actions or re-ignite dormant journeys. | An email click that prompts an return visit receives extra credit for driving momentum. |
Practical examples and use cases
Use case 1: Spotting and reallocating budget from "vampire ads"
Scenario: A retargeting campaign displays a high ROAS in Meta Ads Manager, while top-of-funnel TikTok prospecting campaigns appear underperforming.
Last touch view: Retargeting receives 100% of conversion credit because it was the final touchpoint.
Adaptive Attribution view: The model detects that the retargeting click was a 4-second session prior to checkout, whereas the initial TikTok click led to a 3-minute session with multiple product page views.
Action: Reallocate budget from over-credited retargeting ads ("vampire ads") into high-converting prospecting creatives.
Use case 2: Journey split vs. fixed lookback windows
Scenario: A customer clicks a YouTube ad in January, goes silent for 60 days, then receives a win-back email in March and purchases.
Fixed 30-Day Window: Amputates the January YouTube ad entirely because it falls outside the 30-day cutoff, attributing 100% to Email.
Adaptive Attribution:
Detects that the 60-day silence exceeds the returning customer gap threshold.
Splits the journey into two consideration cycles.
Drops the stale January touchpoint based on true journey gap detection rather than an arbitrary calendar cutoff.
Frequently asked questions
Q: Is Adaptive Attribution available for historical data?
A: Adaptive Attribution is available by default for data from July 29, 2026 onwards. For earlier periods, Admetrics uses the Linear Attribution model.
Q: How does Adaptive Attribution differ from Meta or Google Ads attribution?
A: Ad network attribution models operate in platform silos, claiming credit for conversions whenever their ad was viewed or clicked, often resulting in double-counting across networks. Adaptive Attribution uses cross-channel first-party journey data and ensures that total attributed revenue across all touchpoints equals exactly 100% of order value.
Q: Why is there no lookback window setting for Adaptive Attribution?
A: Fixed attribution windows (7, 30, or 90 days) apply a rigid cutoff regardless of customer type or channel memory. Adaptive Attribution detects natural gaps in activity to split journeys into new consideration cycles, while applying channel-specific exponential decay to older touchpoints. The structure of the journey itself determines the relevant timeframe.
Q: Are the channel multipliers fixed or dynamic?
A: They are dynamic. The adaptive channel multipliers are learned per account from your store's specific customer journey data, continually reflecting the true role and contribution of your marketing efforts.
How to share feedback
We are constantly refining our knowledge base documentation and platform capabilities.
If you have feedback, questions, or suggestions regarding this article:
Use the Was this article helpful? reaction buttons at the bottom of this page.
Reach out directly to our support team via the in-app chat widget in your Admetrics dashboard.
Send your suggestions to [email protected].