Attribution Reporting

Attribution reporting answers the question every merchant asks: “Which marketing channels actually drive my revenue?” By analyzing the full customer journey from your first-party data, Upstack Analytics assigns conversion credit across channels using multiple attribution models — giving you an independent, unbiased view of performance.

​

Available Attribution Models

Upstack Analytics supports four attribution models. Each model uses the same underlying data — identity-resolved customer journeys — but distributes conversion credit differently.

​

Comparing Models

The attribution dashboard lets you view the same data through multiple models simultaneously. This reveals channels that look different depending on how you measure: Example comparison for a DTC apparel store: In this example, Meta Ads is the strongest acquisition channel (first-touch), Google Search and Klaviyo are strong closers (last-touch), and Organic Social is an underappreciated top-of-funnel contributor that last-touch nearly zeroes out.

​

Understanding Channel Credit

Channel credit is the aggregated attribution for a marketing channel across all conversions in your selected time period.

​

What Counts as a Touchpoint

A touchpoint is a session where the visitor arrived through a trackable marketing channel:

• Paid Social — Session with utm_source = facebook, instagram, etc.
• Paid Search — Session with utm_source = google, bing and utm_medium = cpc
• Email — Session with utm_source = klaviyo, mailchimp, etc.
• Organic Search — Session with referrer from a search engine and no paid UTM parameters
• Organic Social — Session with referrer from a social platform and no paid UTM parameters
• Direct — Session with no referrer and no UTM parameters

​

How Credit Sums Work

Within a single attribution model, total attributed revenue equals your actual total revenue. There is no double-counting — this is the fundamental advantage over ad platform self-reporting. If a customer journey has three touchpoints and results in a $90 purchase, the Linear model assigns$30 to each channel. The total attributed is still $90.

​

Lookback Windows

The lookback window defines how far back Upstack searches for qualifying touchpoints before a conversion. A shorter window gives credit to channels that drive quick conversions. A longer window captures channels that influence early in the journey but convert later. Test different windows to see how your channel mix shifts.

​

Reading the Attribution Dashboard

The attribution dashboard organizes data into four sections:

​

Channel Overview

A summary table showing each channel’s attributed revenue, conversion count, and ROAS across your selected model and date range. This is your starting point for budget allocation decisions.

​

Model Comparison

Side-by-side view of all four models for the same time period. Use this to identify channels that vary significantly across models — those are the ones worth investigating further.

​

Journey Explorer

Drill into individual customer journeys to see the exact sequence of touchpoints. Filter by channel, conversion value, or number of touchpoints. Useful for understanding how your customers actually discover and buy from your store.

​

Trend View

Attribution credit by channel over time. See how channel performance evolves week-over-week. Identify seasonal patterns, campaign impacts, and long-term shifts in your marketing mix.

​

Tips for Actionable Attribution

Start with comparison, not a single model. Look at where models agree (strong signal) and where they disagree (needs investigation). Match the lookback window to your product. A 28-day window for a $15 impulse buy dilutes signal. A 7-day window for a$500 product misses the real journey. Use first-touch for acquisition budgets. If you’re trying to grow your customer base, first-touch tells you which channels bring net-new visitors. Use last-touch to benchmark against platforms. Compare Upstack’s last-touch to Meta’s self-reported numbers. The gap reveals how much each platform over-claims. Watch the Linear model for hidden contributors. Channels that score low on both first-touch and last-touch but appear in Linear are mid-funnel influencers. Cutting them may quietly reduce overall conversion rates.