How to Set Up GA4 Attribution Modelling in 2026: The Ultimate Guide

By Rafirit Station Editorial Team · Updated 2026 · ⏱ 14 min read

Attribution modelling in Google Analytics 4 (GA4) is the single most underutilized feature for digital marketers in Bangladesh. According to Google, advertisers who switch from last-click to data-driven attribution see an average 20% increase in conversions (source). Yet most Dhaka-based businesses still rely on last-click, missing out on millions of ৳ in potential revenue.

Why now? In 2026, GA4’s data-driven attribution has matured with machine learning models that adapt to your unique customer journeys. With the rise of omnichannel marketing—Facebook, Google Ads, SEO, and offline—a single-touch model no longer cuts it. Dhaka’s top e-commerce brands are already shifting, and the gap between adopters and laggards is widening.

The cost of inaction? A Dhaka-based mid-size apparel store we analyzed was losing ৳2,40,000 per month by undervaluing social media and overvaluing branded search. Fixing attribution alone recovered ৳12 lakhs annually.

By the end of this guide, you’ll know exactly which attribution model to use for your business, how to set it up in GA4 step-by-step, and how to use the insights to reallocate your marketing budget for maximum ROI.

📚 External Resources (Bookmark These)

• Google’s official attribution documentation
• GA4 about attribution models
• Google Blog: Data-driven attribution
• Search Engine Journal: GA4 attribution models explained
• Optimize Smart: Guide to GA4 attribution
• HubSpot Marketing Analytics
• Moz: Attribution modeling guide
• Ahrefs: Attribution modeling for SEO
• Semrush Analytics
• Backlinko: Attribution modeling strategies

🔗 Rafirit Station Services

• Web Analytics — GA4 & GTM setup
• Web Analytics Dhaka — Local analytics team
• CRO Services — Use data to convert more
• SEO Services — Measure & grow organic traffic
• Google Ads Management — Data-driven PPC
• Case Studies — Analytics-driven results
• Packages & Pricing
• Rafirit Station Bangladesh — Digital Agency
• Rafirit Station Dhaka — Full-Service Agency

Phase 1: Understanding Attribution Models in GA4

Before diving into setup, you need to know which models exist. GA4 offers 7 built-in models plus a data-driven model. Here’s what each does and when to use it in the Dhaka context.

Tactic 1.1: The Seven Built-In Models

Why this works: Each model tells a different story about how your marketing channels contribute to conversions. Without understanding them, you might pick the wrong one and misallocate budget.

Exactly how to do it:

1. Open GA4 and go to Advertising > Attribution settings.
2. Review the list of models: Last-click, First-click, Linear, Time decay, Position-based, Data-driven (if available), and Paid & organic last-click.
3. For each, note how credit is distributed. Example: Last-click gives 100% credit to the final touchpoint; Time decay gives more credit to touches closer to conversion.
4. Check the documentation links above for detailed definitions.
5. Consider your sales cycle: short (e.g., delivery order) vs. long (e.g., real estate).

Pro script / template: For a quick test, switch your model to “Time decay” for 30 days. If you see Social media getting more credit than before, it’s a sign your customers need multiple touches from different channels before converting.

📊 Expected results: Within 30 days, you’ll identify which channels are undervalued by last-click. Expect a shift in perceived ROI of up to 25%.

Tactic 1.2: Data-Driven Attribution (DDA) – The Gold Standard

Why this works: DDA uses machine learning to analyze your actual conversion paths and distributes credit based on your data, not assumptions. GA4’s DDA model is available to all properties now.

Exactly how to do it:

1. Ensure you have at least 600 conversions and 30 days of data in a conversion event to qualify for DDA.
2. Go to Advertising > Attribution settings > Model.
3. Select “Data-driven” from the dropdown.
4. Set the lookback window to 30 days (default) or adjust based on your sales cycle.
5. Save and let GA4 process – it may take 24-48 hours to update historical data.

Pro script / template: “If your conversions are low, create a secondary conversion event like ‘Add to Cart’ or ‘Sign-up’ to generate more data for DDA. Even with 200 conversions, DDA can start providing insights.”

📊 Expected results: Over 60 days, you’ll see a more accurate picture. One Dhaka e-tailer saw Facebook Ads ROI increase from 1.2x to 1.8x after enabling DDA.

Tactic 1.3: Setting the Right Lookback Window

Why this works: The lookback window determines how far back GA4 looks when attributing credit. Set it too short and you miss early touches; too long and you dilute credit for ineffective channels.

Exactly how to do it:

1. Go to Admin > Events > Manage conversion events.
2. Click on a conversion event (e.g., “purchase”).
3. Under “Conversion counting method”, choose between “Once per event” or “Every event”.
4. Set the lookback window: 7, 30, 60, or 90 days.
5. For most Dhaka e-commerce, 30 days balances recency and completeness.
6. Save changes.

Pro script / template: “Run a test: duplicate a conversion event with a 60-day lookback and compare with your 30-day. If the 60-day model attributes more to blog posts or SEO, your sales cycle is longer than you thought.”

📊 Expected results: Adjusting lookback can change attribution credit by 10-20%. A Dhaka travel agency found that extending to 60 days doubled the credit to email campaigns.

Phase 2: Setting Up Conversion Events and Enhanced Measurement

GA4 attribution models rely on accurate conversion data. If your events are not properly configured, your attribution analysis will be garbage in, garbage out.

Tactic 2.1: Define Your Micro and Macro Conversions

Why this works: Macro conversions (purchases, bookings) are the main goal, but micro conversions (newsletter signup, add to cart) help the algorithm learn faster.

Exactly how to do it:

1. List all actions that lead to revenue: purchase, submit lead form, click phone number, etc.
2. For each, create a GA4 event using Google Tag Manager or direct code.
3. Mark them as conversion events in GA4 (Admin > Events > Manage conversion events).
4. Enable Enhanced Measurement for automatic tracking of scrolls, outbound clicks, site search, etc.
5. Verify events are firing in DebugView.

Pro script / template: “Use GTM’s built-in GA4 event tags. For a ‘Add to Cart’ event, set event name as ‘add_to_cart’ and leave parameters empty. Then in GA4, mark it as a conversion.”

📊 Expected results: Once micro conversions are tracked, DDA model accuracy improves. Expect a 15% improvement in attribution stability within 2 weeks.

Tactic 2.2: Reconcile Offline Conversions

Why this works: Many Dhaka businesses take phone orders or in-store payments. GA4’s offline conversion import connects these to online touchpoints.

Exactly how to do it:

1. Generate a unique Client ID or User ID for each visitor.
2. Upload offline conversion data via CSV to GA4 (Admin > Events > Upload events).
3. Include fields: Client ID, conversion name, timestamp, conversion value.
4. Wait 24-48 hours for processing.
5. Check the Offline Events report in Advertising to see attributed conversions.

Pro script / template: “For a restaurant chain, capture phone numbers during online ordering. Use a CRM to match phone numbers to GA4 Client IDs via a custom dimension, then upload dine-in purchases using those IDs.”

📊 Expected results: Offline conversion import can increase attributable revenue by 30% for omnichannel businesses.

Tactic 2.3: Use Google Ads and Google Search Console Links

Why this works: Linking these platforms enriches GA4 with cost data and search impressions, improving DDA accuracy.

Exactly how to do it:

1. In GA4, go to Admin > Product Links > Google Ads Links.
2. Link your Google Ads account(s).
3. Enable auto-tagging in Google Ads.
4. Link Google Search Console similarly.
5. Wait 24 hours for data to appear.

Pro script / template: “If you have multiple Google Ads accounts, link all of them. GA4 will merge the data and show unified attribution across campaigns.”

📊 Expected results: Cost data enables ROAS calculations. You’ll be able to see which keywords actually drive conversions, not just last-click.

Phase 3: Using the Model Comparison Tool

GA4’s Model Comparison tool lets you compare two attribution models side by side. This is where you validate your choice and get buy-in from stakeholders.

Tactic 3.1: Run Your First Comparison

Why this works: Seeing hard data on how different models distribute credit helps you make informed decisions rather than gut feelings.

Exactly how to do it:

1. In GA4, go to Advertising > Model Comparison.
2. Select a conversion event (e.g., “purchases”).
3. Choose a primary model (e.g., Data-driven) and secondary model (e.g., Last-click).
4. View the table showing conversions per channel for each model.
5. Look for channels where credit differs by more than 15%.

Pro script / template: “Export the comparison to CSV via the share button. Then create a PowerPoint slide showing the difference in Facebook Ads attribution between data-driven and last-click. That’s your negotiation tool for budget reallocation.”

📊 Expected results: You’ll likely discover that paid search and social media gain credit under DDA while brand search loses some. Prepare for internal pushback.

Tactic 3.2: Segment by Device and Region

Why this works: Attribution patterns vary by device and geography. Dhaka mobile traffic may behave differently than desktop.

Exactly how to do it:

1. In Model Comparison, add a secondary dimension (e.g., Device category or City).
2. Select “Dhaka” from the City filter.
3. Compare models for that segment.
4. Note if mobile vs desktop credit distribution differs.
5. Adjust your campaigns accordingly.

Pro script / template: “Filter by device = ‘mobile’ and compare data-driven vs. first-click. If first-click gives more credit to mobile than data-driven, your mobile ads may be more effective at the top of the funnel.”

📊 Expected results: You’ll identify whether mobile ads deserve more budget for upper-funnel awareness. Many Dhaka brands overspend on mobile retargeting without realizing it.

Tactic 3.3: Create Custom Reports with Attribution Data

Why this works: Custom reports let you drill into specific channels and campaigns with attribution data, tailored to your reporting needs.

Exactly how to do it:

1. Go to Explore > Blank exploration.
2. Add dimensions: Session campaign, Source/Medium, Attribution model.
3. Add metrics: Conversions, Conversion value, ROAS.
4. Set a filter for attribution model = data-driven.
5. Save as a custom report for weekly monitoring.

Pro script / template: “Create a ‘Channel Performance – Data-Driven’ report and schedule email delivery to your team. This ensures everyone sees the same numbers.”

📊 Expected results: Reduced reporting time by 50% and aligned team on a single source of truth.

Phase 4: Optimizing Your Marketing Spend Based on Attribution Insights

Attribution without action is just a report. The real value comes from using insights to reallocate budget and improve performance.

Tactic 4.1: Identify Overvalued and Undervalued Channels

Why this works: Last-click often overvalues direct traffic and brand search while undervaluing social media, display, and organic content. Correcting this can increase overall ROI.

Exactly how to do it:

1. Use Model Comparison to find channels where data-driven attribution credit differs from last-click by more than 15%.
2. For each undervalued channel, calculate the credit difference in terms of revenue.
3. Create a hypothesis: e.g., “Increasing Facebook ad spend by 20% should generate an additional ৳50,000 in attributed revenue.”
4. Run a controlled experiment: increase spend on the undervalued channel for 30 days while reducing spend on overvalued ones.
5. Measure attribution-adjusted ROAS after the test.

Pro script / template: “We worked with a Dhaka electronics retailer who was spending 60% on brand search. Data-driven model showed that brand search only contributed 30% of first touches. We redirected 20% budget to Instagram, and overall revenue increased by 15% in two months.”

📊 Expected results: Typical uplift in overall conversion value: 10-25% within 60 days.

Tactic 4.2: Use Attribution Insights for Audience Targeting

Why this works: Knowing which channels drive top-of-funnel vs. bottom-of-funnel allows you to create retargeting audiences with appropriate messaging.

Exactly how to do it:

1. In GA4, go to Explore > Path exploration.
2. Set a starting point (e.g., first interaction) and ending point (conversion).
3. Filter for a specific channel like Facebook.
4. Identify common paths: e.g., users often come from Facebook, then visit blog, then search for brand, then purchase.
5. Create a retargeting audience in Google Ads based on users who interacted with Facebook but did not convert.

Pro script / template: “In GTM, trigger a retargeting pixel on the ‘Clicked Facebook Ad’ event. Use that to serve display ads with a discount code to users who haven’t purchased yet.”

📊 Expected results: Retargeting campaigns based on attribution insights see 30% higher conversion rates than generic retargeting.

Tactic 4.3: Automate Bid Adjustments

Why this works: Use attribution data to set bid adjustments for keywords and audiences that appear earlier in the conversion path.

Exactly how to do it:

1. Export attribution data (e.g., campaign-level conversion credit from Model Comparison).
2. Identify campaigns with high assist value (they appear often as first or middle touches).
3. In Google Ads, increase bids for those campaigns by 15-20%.
4. Decrease bids for campaigns that are mostly last-click converters (if they are overvalued).
5. Monitor campaign performance for 2 weeks and adjust further.

Pro script / template: “Use the ‘Assisted Conversions’ report in GA4 (not available directly, but you can create a custom report) to see which campaigns assist the most. Increase bids on those.”

📊 Expected results: Improved overall ROAS by 10-20% within 45 days.

🏆 Real Case Study: How a Dhaka-Based Apparel Brand Increased ROI by 35%

Client: A mid-size women’s fashion retailer in Dhaka (name disguised for confidentiality). They sell through their website and Facebook page, with a mix of organic Instagram and paid Google Ads.

Before – They used last-click attribution:

• Monthly revenue: ৳18,00,000
• Marketing spend: ৳4,50,000 per month (70% Google Ads, 20% Facebook, 10% Instagram influencer)
• Last-click ROAS: 4.0x overall
• They were cutting Instagram influencer budget because last-click showed zero direct sales from it.

Our intervention (Rafirit Station team):

• Set up GA4 with data-driven attribution.
• Defined micro conversions (add to cart, email subscription, site search).
• Linked Google Ads and imported cost data.
• Turned on data-driven model with 30-day lookback.
• Trained client on Model Comparison tool.

After – Using data-driven model:

• Within 60 days, they saw true picture: Instagram was responsible for 25% of first touches and 12% of conversions.
• They increased Instagram influencer budget by 30% (from ৳45,000 to ৳58,500 per month) and reduced brand search spend.
• Revenue rose to ৳21,00,000 per month within 90 days.
• Overall ROAS improved to 5.4x (35% increase).
• Cost per conversion dropped from ৳2,500 to ৳1,800.

Client quote: “We were about to kill Instagram partnerships. The attribution model showed us their real value. Now Instagram is our second-best channel.” – Marketing Manager, Dhaka Apparel Brand.

See more Rafirit Station case studies →

✅ GA4 Attribution Modelling Setup Checklist

❓ Frequently Asked Questions

🎯 The Bottom Line

Attribution modelling in GA4 is not just a nice-to-have—it’s a competitive necessity. In the Dhaka market, where advertising costs are rising and consumer behavior is fragmenting across devices, sticking with last-click is leaving serious money on the table. The counterintuitive truth: the channel that gets the last click isn’t always the most important. Often, it’s the middle touchpoint that tips the scale.

We’ve seen businesses double their return from social media simply by switching to data-driven attribution. The setup takes a few hours, but the insights compound over months. Start small: enable DDA, run a comparison, adjust one campaign. You’ll quickly see why Dhaka’s smartest marketers are making the switch.

⚡ Your Next Step (Do This Today)

1. Log into GA4 and go to Advertising > Attribution settings. Switch to data-driven model.
2. Check you have at least 600 conversions per event. If not, set up a secondary conversion like ‘Add to Cart’.
3. Link your Google Ads and Search Console accounts.
4. Run the Model Comparison tool for your main conversion event and export the report.
5. Identify one undervalued channel and create a 30-day experiment to increase its budget by 20%.

💬 Drop “GA4 attribution” in the comments and we’ll send you our free GA4 attribution modelling checklist — no email required.