Why Most Attribution Is Still Flawed

Let’s call it out, most attribution models are outdated.

Many businesses still rely heavily on last click attribution, even though customers often interact with a brand nine to eleven times before purchasing. Social ads, emails, Google searches, referrals and SMS all play a role, but last click ignores that nuance.

Platform reported metrics from Meta, Google, TikTok and others are designed to show their own value. They tend to over credit themselves, leading to skewed decision making and inflated campaign ROI.

Cookie windows are shrinking. Some platforms only retain data for seven days or less. Longer consideration cycle customers in furniture, homewares or apparel can go untracked for large portions of their journey.

If you are making big media spend decisions using this kind of data, you are likely leaving revenue on the table or spending more than you need to.

The Hidden Cost of Poor Visibility

In our Spotlight Fridays discussion with attribution expert Adam Saunders, he shared an example of a retail brand spending aggressively on campaigns that looked profitable in Meta’s dashboard. After layering in multi touch attribution and product level margin data, those same campaigns were actually losing money. With better visibility, the brand reduced spend by twenty five percent while maintaining the same revenue.

It was not the platform that changed. It was the clarity. You can hear more in the full video replay linked below.

What You Should Be Measuring (But Probably Are Not)

Attribution is not just about channel performance. It is about connecting your campaigns, customers and products.

• Which campaigns drive first purchases versus repeat purchases
• Which product categories attract high margin customers
• How new versus returning customer behaviour changes over time
• Where potential loyalists are dropping off and why

Without this level of insight, you are guessing which levers to pull and often leaning too hard on price promotions for short term gains.

Do Not Just Collect Data. Do Something With It

Data overwhelm is real. The issue is rarely volume. It is structure and action.

Many retailers are turning to AI agents as a practical solution.

• Automate weekly performance reviews
• Surface outliers or anomalies in campaign results
• Flag underperforming products or high risk audiences
• Detect issues like affiliate fraud that would otherwise go unnoticed

This saves teams hours of reporting and shifts focus from reactive firefighting to proactive planning.

What Good Attribution Looks Like

Done right, attribution helps you make faster, clearer and more profitable decisions. It should help you:

• Spend more confidently across paid media
• Reduce over reliance on discounting to drive volume
• Invest in the right products, categories and customer groups
• Build loyalty by reinforcing what already works

It also creates stronger internal alignment. When marketing, finance and digital teams look at the same source of truth, there is less debate and more action.

How to Strengthen Your Marketing Clarity

Here is what to consider:

1. Move beyond last click. Test multiple attribution models across different campaign types. Use new customer ROAS for acquisition, and MER or retention metrics for CRM initiatives.
2. Tie attribution to margin, not just sales. If a top performing campaign drives low margin products, it may be hurting your bottom line.
3. Bring customer level data into the picture. Who buys, how often, and what do they return for. Segment around behaviour, not just demographics.
4. Build reporting that works for you, not just for the platform. Automate where possible, but focus on insights you can act on. Speed matters, clarity wins.

Tools That Support Smarter Attribution

Strategy should always lead, but the right tools help you act faster and with more confidence. Whether you use GA4, Power BI or a dedicated attribution platform like Triple Whale, the goal is to make data meaningful by connecting campaigns, customers and products into one clear view.

We have worked across a wide mix of technologies and always come back to the same principle: the tech should support the strategy, not define it.

If you are not sure your current setup is giving you the right visibility, it may be time to reassess how it is being used, not just what is being used.

Want to hear the full conversation that inspired these insights?
Watch the Spotlight Fridays episode with Adam Saunders: https://youtu.be/13pCqQyTQpo