Introducing Causal Attribution: Your New Daily Incrementality Solution

If you’re investing in paid media, you should be able to easily track which channels and tactics add incremental value to your business every day, right? 

We think so, but it hasn’t always been easy. But with privacy-durable experiments, Haus helps brands uncover the channels and tactics that are actually driving online and offline conversions. Now, we’re introducing a new solution to take these learnings one step further: Causal Attribution. 

Causality is at the core of Haus’ mission – after all, there’s only so much one can accurately infer by stitching together clicks, views, and other correlative signals. With Causal Attribution, Haus solves the problem of getting a daily read on each channel’s incremental performance.

What is Causal Attribution?

Think of Causal Attribution as a way to weave incrementality into your day-to-day. By syncing with your ad platform data, we recalculate returns based on an incrementality adjustment determined from your experiments. You can visualize your daily cost per incremental acquisition (CPIA) from a given channel and compare it against platform-reported data. The best part? Data is updated daily, so you have the latest information at your fingertips to optimize ad channels – and budgets – for performance.

How is Causal Attribution different from other attribution solutions?

As its name suggests, Causal Attribution incorporates causal data (experiment results) into the attribution equation instead of solely relying on correlational data (platform metrics). Using causal data is crucial for understanding incrementality because it establishes a cause-and-effect relationship – demonstrating if an ad directly caused a conversion.

Correlation-based solutions like in-platform reporting or multi-touch attribution (MTA) can suggest a relationship between views, clicks, and conversions, but they can’t confirm incrementality. For brands with smaller marketing budgets, correlation-based solutions are a good starting point for understanding which ad channels generally work on a daily basis. But for brands investing several millions of dollars in paid ads and advertising on multiple channels, these metrics may not accurately represent each platform’s actual business impact.

We've consistently seen this across our customers' experiment results. In a recent study, we examined the incrementality of platforms such as Meta, YouTube, Search, CTV, Pinterest, and Snapchat on several brands' e-commerce sales over the last twelve months. We found that:

• Overall, platforms over-reported 65% of the time 
• 82% of incrementality experiments showed that platform reporting was either under-reporting or over-reporting conversions by more than 25% 
• 60% of experiments showed discrepancies of more than 50% 

Causal Attribution helps you understand how a channel is performing week over week without relying on the inaccuracies of platform data.

How does Causal Attribution work?

Causal Attribution uses your GeoLift experiment results to adjust your platform-reported metrics. At the completion of every Haus experiment, you’ll receive an Incrementality Factor in your analysis. Incrementality Factors ("IFs") tell you how many platform-credited conversions were actually incremental by dividing incremental conversions from a test by platform-reported conversions. 

Causal Attribution then automatically multiplies platform-reported conversions by the incrementality factor to yield an incrementality estimate. The result? An easy-to-read dashboard that shows you the daily, weekly, and monthly incrementality of your channels as well as the cost per incremental acquisition (CPIA) per tactic. Armed with this information, it’s easy to see opportunities to optimize your media buying.

Let’s walk through an example: 

This chart shows YouTube's and Brand Search's performance for a brand over the course of a month. This brand ran experiments on both platforms and saw that YouTube was under-reporting incremental conversions by 25% (i.e. IF = 1.25) and Brand Search (i.e. IF = 0.10) was over-reporting by 90%. After adjusting the platform CPA to account for incrementality, this brand found that YouTube (CPIA of $60) is a more efficient channel than Brand Search (CPIA of $200). Since the brand is spending significantly more on Brand Search than they are on YouTube, they have a variety of actionable steps they could take depending on their business needs and goals. 

A few ideas include:

• Decrease spend on Brand Search and reinvest it into other areas of the business. 
• Shift some ad spend immediately from Brand Search to YouTube to optimize for ROI. 
• Run a test to determine YouTube's optimal spend level and see if they can increase spend efficiently.

Ready to see Causal Attribution in action? Schedule a demo today. Existing Haus customers can contact your CSM to get started.