GA4 Per-Conversion Attribution Settings Guide 2026

Step 1: Verify Your Key Events

Go to Admin > Events and confirm which events are marked as key events (conversions). You can only configure per-conversion attribution for events flagged as key events. Common examples:

• • purchase — e-commerce transactions
• • generate_lead — form submissions
• • sign_up — account registrations
• • add_to_cart — shopping intent signals
• • begin_checkout — high-intent funnel events

Step 2: Navigate to Conversion Management

Go to Advertising > Conversion Management in your GA4 property. This section — not Admin > Attribution Settings — is where per-conversion settings live. You'll see a list of all your key events with their current attribution configuration.

Step 3: Select a Key Event

Click on a specific key event (e.g., purchase) to open its attribution settings. You'll see two configurable options: Attribution Model and Lookback Window.

Step 4: Choose the Attribution Model

Select from three available models:

• Data-Driven Attribution (DDA) — ML-based, distributes credit based on actual contribution. Requires 400+ conversions for the key event and 20,000 total conversions.
• Paid and Organic Last Click — All credit to the last touchpoint (paid or organic) before conversion. Ignores Direct traffic when a prior touchpoint exists.
• Google Paid Channels Last Click — All credit to the last Google Ads click. Use only if you need Google Ads-aligned reporting.

Step 5: Set the Lookback Window

Choose how far back GA4 should look for touchpoints:

• • 7 days — for acquisition events (first_visit, first_open)
• • 30 days — for quick-conversion events (signups, downloads)
• • 60 days — for mid-cycle events (leads, trials)
• • 90 days — for long-cycle events (purchases, enterprise deals)

Step 6: Save and Verify

Click Save. Note that attribution changes apply going forward — they don't retroactively change historical data. Allow 24-48 hours for the new settings to fully propagate in your reports. Use the Model Comparison report (Advertising > Model Comparison) to verify that DDA is producing different results than last-click, confirming it's actually active.

Data-Driven Attribution (DDA)

GA4's recommended default. DDA uses machine learning to analyze both converting and non-converting paths, then assigns credit based on each touchpoint's actual contribution to the conversion. According to Google's documentation, DDA considers factors like time between touchpoints, device type, number of ad interactions, and the order of touchpoint exposure.

Best for: High-volume conversions where multiple channels contribute (purchases, lead submissions, trial signups).

Paid and Organic Last Click

100% of conversion credit goes to the last non-direct touchpoint. If a user clicks a Google Ad, then visits directly and converts, the Google Ad gets full credit. This model is deterministic — no ML, no estimation, no minimum data requirements.

Best for: Low-volume conversions under 400/month, micro-conversions (add-to-cart, page views), and situations where you need simple, consistent reporting.

Google Paid Channels Last Click

Attributes 100% of credit to the last Google Ads click. All other channels (organic, social, email, direct) get zero credit. This model exists for parity with Google Ads' native reporting.

Best for: Only when you need GA4 reports to match Google Ads reporting exactly. Not recommended as a general-purpose model since it ignores non-Google channels entirely.

E-Commerce: Purchase Events

Why: Shoppers research across multiple sessions before purchasing. A typical e-commerce journey involves 4-8 touchpoints — discovery on social media, research on Google, retargeting ads, then a direct visit to buy. DDA captures this full path and distributes credit accurately. The longer lookback accounts for comparison shopping and wait-for-sale behavior.

SaaS: Free Trial Signups

Why: SaaS trial signups usually take 2-4 weeks from first touch. The user reads a blog post, sees a retargeting ad, checks a review site, then signs up. DDA gives proportional credit to each step. A 30-60 day window covers the typical SaaS evaluation period without inflating attribution from stale touchpoints.

Content / Newsletter: Email Signups

Why: Newsletter signups are low-friction, single-session conversions. The user lands on a page and signs up — or they don't. Last-click accurately captures which channel drove the session. A short lookback window prevents a random organic visit three months ago from getting credit for today's signup.

B2B: Demo Requests / Lead Forms

Why: B2B sales cycles are long. A decision-maker might first find you through a blog post, return via LinkedIn, download a whitepaper, attend a webinar, then finally request a demo. DDA with a 90-day window captures the full influence chain. For more on tracking complex B2B funnels, see our marketing attribution dashboard guide.

Channel-Level ROAS Will Diverge by Conversion Type

With global attribution, Google Ads might show 4.2x ROAS across all conversions. With per-conversion settings, that same channel might show 5.1x ROAS for purchases (DDA capturing assist value) but only 2.8x ROAS for newsletter signups (last-click showing lower influence). This is more accurate — your channels perform differently for different goals, and now your data reflects that.

Upper-Funnel Channels Get More Purchase Credit

When you switch purchases to DDA with a 90-day window, channels like organic social, display ads, and content marketing will receive more credit for assists. Last-click systematically undercredits channels that drive awareness but don't close the sale. DDA fixes this — expect paid social and organic content to show 15-30% higher attributed purchase value compared to last-click.

Micro-Conversion Attribution Gets Cleaner

With a shorter lookback window on micro-conversions (add-to-cart, video views), you eliminate attribution noise from old touchpoints. A user who visited via organic search 60 days ago and adds an item to cart today after clicking a retargeting ad — the retargeting ad correctly gets the credit with a 30-day window, not the organic visit from two months ago.

1. Assuming DDA Is Active Without Checking

The most common mistake. You select data-driven attribution for a key event that only gets 150 conversions per month. GA4 silently falls back to last-click. You think you're running sophisticated ML-based attribution, but you're actually on last-click. Always verify using Model Comparison — if both models show identical results, DDA isn't active.

2. Setting the Same Window for All Events

Using a 90-day lookback for newsletter signups inflates credit for touchpoints that had nothing to do with the conversion. Using a 7-day window for purchases cuts off touchpoints that genuinely influenced the buying decision. The whole point of per-conversion settings is differentiation — use it.

3. Changing Attribution Without Documenting Baseline

Before changing any attribution settings, export your current channel performance data. Attribution changes affect reported numbers from the change date forward. Without a baseline, you won't know whether a change in reported ROAS reflects actual performance improvement or just the attribution model shift. Take screenshots or export reports for the 30 days before switching.

4. Ignoring Google Ads Alignment

If you import GA4 conversions into Google Ads for bid optimization, your per-conversion attribution settings carry over. Switching purchases from last-click to DDA will change how Google Ads' smart bidding evaluates campaign performance. This can cause temporary bid fluctuation. Notify your paid team before making changes and allow 1-2 weeks for smart bidding to recalibrate.

5. Expecting Retroactive Changes

Per-conversion attribution settings apply from the date you save them. Historical data stays as-is under the old model. If you need to compare old vs. new, use the Model Comparison report — it lets you view historical data through different attribution lenses without changing actual settings.

What are GA4 per-conversion attribution settings?

GA4 per-conversion attribution settings let you configure different attribution models and lookback windows for each key event independently. Previously, all conversions shared one global attribution model. Now you can use data-driven attribution with a 90-day window for purchases while using last-click with a 30-day window for newsletter signups — all within the same GA4 property.

How do I set different attribution models per conversion in GA4?

Navigate to Advertising > Conversion Management in your GA4 property. Click on a specific key event to open its settings. Under Attribution Model, select from Data-Driven, Paid and Organic Last Click, or Google Paid Channels Last Click. Under Lookback Window, choose 7, 30, 60, or 90 days. Save your changes. Repeat for each key event that needs different settings.

Should I use the same attribution model for all conversions?

No. Different conversion types have different buyer journeys. High-value purchases typically involve multiple touchpoints over weeks, making data-driven attribution with a 90-day lookback window ideal. Low-friction conversions like email signups happen quickly and are better served by last-click attribution with a shorter 30-day window. Matching the model to the conversion type gives you more accurate marketing insights.

How does per-conversion attribution affect my marketing reports?

Per-conversion attribution changes how credit is distributed across your marketing channels for each conversion type. Your Google Ads campaigns might get more credit for purchases (with data-driven attribution showing assist value) but less credit for newsletter signups (with last-click attribution). This means your channel-level ROAS and CPA metrics will vary by conversion type, giving you a more nuanced view of which channels drive which outcomes.

What attribution model works best for lead generation vs e-commerce?

For lead generation, use data-driven attribution with a 30 to 60-day lookback window. Leads typically convert within a few weeks, and DDA captures the multi-touch journey from awareness to form submission. For e-commerce purchases, use data-driven attribution with a 60 to 90-day lookback window since high-value purchases involve longer consideration periods. For micro-conversions like add-to-cart, last-click with a 7 to 30-day window works well.

What is the minimum data needed for data-driven attribution in GA4?

GA4 requires at least 400 conversions for a specific key event and 20,000 total conversions across all events within your lookback window for data-driven attribution to activate. If your property falls below these thresholds, GA4 silently falls back to last-click attribution without notifying you. Check your conversion volume before relying on DDA — many smaller businesses unknowingly run on last-click.

Can I use GA4 per-conversion attribution with Google Ads?

Yes. When you link GA4 to Google Ads and import conversions, the attribution model you set per key event in GA4 carries over to your Google Ads reporting. This means Google Ads will use your per-conversion attribution settings when optimizing bids and reporting on campaign performance. For best results, align your GA4 attribution settings with your Google Ads bidding strategy.

How do I check if data-driven attribution is actually active for my conversions?

Go to Advertising > Model Comparison to see if data-driven attribution shows different results than last-click. If both models show identical credit distribution, DDA likely fell back to last-click due to insufficient conversion volume. You can also check your key event volume in the Events report — you need 400+ conversions per key event within your lookback window.