Since the dawn of the internet, one question has perplexed advertisers more than any other: how do you know which of your ads are driving the most conversions?

Received marketing wisdom reminds us of the rule of seven: a customer must be exposed to your marketing messages at least seven times before they’ll click the ‘Buy’ button.

So how can you tell which ad made them finally take the plunge?

Advertisers have attempted to answer this question through the development of different attribution models. These are sets of rules that assign credit for conversions to touchpoints in conversion paths. 

First, there were mixed marketing models (MMM), which provided general conversion data but lacked specific details. Then came multi-touch attribution models (MTA) — such as linear, position-based, and time decay.

These attribution models provide different insights about the conversion path, but none paint the full picture of the customer journey and how different touchpoints interact before conversion occurs. 

Therefore, advertisers need to move beyond current attribution models and find more accurate ways to measure the outcomes of their advertising campaigns.

That’s where data-driven attribution (DDA) comes in. DDA uses machine learning and artificial intelligence to provide a more detailed understanding of how each touchpoint contributes to a conversion. 

This article will explain what the new DDA model is and how advertisers can use it to optimize their ad campaigns and drive conversions.

What is the New Data-Driven Attribution Model?

The DDA model is driven by machine learning and algorithms that analyze data across all the marketing touchpoints that eventually lead to a sale. Here is Google’s definition of data-driven attribution:

“Data-driven attribution is different from the other attribution models because it uses your conversion data to calculate the actual contribution of each ad interaction across the conversion path. Each data-driven model is specific to each advertiser.”


Two features jump out here — firstly, data-driven attribution calculates the actual contribution of each ad interaction. When people visit your website and make a purchase, they have probably interacted with several different ads before zeroing in on the one that finally makes them convert. 

By contrast, position-based attribution models would credit a single interaction or a handful of interactions for the conversion, such as the first or last ad interaction. DDA credits each interaction by determining how much it contributes to the sale.

Secondly, data-driven attribution is a highly adaptable type of attribution. It builds a different attribution model for each advertiser, using conversion tracking data to determine how much credit each unique ad should take for the eventual conversion. 

Why DDA? The Attribution Problem

As advertising continues to evolve, simplistic attribution models such as last-click attribution will continue to fall short of advertisers’ needs. The truth is humans don’t make a purchase after clicking on a single ad, so attributing a conversion to just one touchpoint isn’t an accurate way to track your campaigns.

First, they’ll do a bunch of generic searches. Then, they might watch a YouTube video on the product. Next, they may check out a social media platform — such as a Facebook group or Reddit forum — to see what other people have to say. Finally, they’ll go to your website and make their purchase there. 

Linear, time decay, and other position-based attribution models were created to address the issues with last-click attribution but leave critical data gaps that could provide advertisers with valuable information if addressed. 

Data-driven attribution seeks to credit all touchpoints for the eventual conversion. This model aims to provide accurate conversion tracking for businesses while maintaining user privacy. It’s essentially a mission to future-proof attribution as advertising and the internet continue to get more complex. 

Data-driven attribution models use machine learning and data science to gather large amounts of data and insights on the user journey. These models accurately calculate how much each touchpoint has contributed to a conversion. 

They apply tailored, custom-made weightage to each touchpoint based on the specific contexts of that particular user and the particular marketing channels, and they yield extremely precise insights. 

So what does that look like in real life? Let’s say you are running an ad campaign for a new theater production. When a customer buys a ticket, the attribution credit will be split between all the contributing touchpoints for a total of one, for example:

  • New York Times — 0.3 credit
  • Huffington Post — 0.2 credit
  • YouTube ad — 0.1 credit
  • Search ad — 0.4 credit

Armed with this information, you’ll be able to optimize your campaign to focus more on the platforms and formats that perform best. 

Since DDA solves many of the problems related to previous attribution models, we anticipate it will become the standard attribution model in digital advertising and expect to see it rolled out further in the near future.

The Benefits of Data-Driven Attribution

Data-driven attribution is the next evolution of digital advertising attribution, going beyond position-based attribution models to provide more accurate information about advertising campaigns. Here are three of the main benefits advertisers will see from implementing DDA.

1. More Holistic Conversion Data

DDA can give you a more holistic view of your conversion paths and shine a light on touchpoints that deserve credit but get none under other attribution models. 

If a user clicks on an ad but doesn’t convert, that doesn’t mean the ad plays no role in their eventual conversion. If the same user eventually converts — for example, by navigating directly to your website and completing the sale there — DDA will assign part of the credit to the original click-through.

In a first-click attribution model, the original interaction would get all the credit. In a last-click attribution model, it would get none. With DDA, it would receive its due share — this is how DDA provides a more nuanced analysis of your campaigns.

2. More Accurate Reporting

DDA makes your reporting less black-and-white. Using predictive algorithms, it identifies and analyzes the most statistically significant data from a customer’s path, then assigns credit to the most influential touchpoints in the 90 days before conversion.  

Therefore, partial conversion credit tells a more complete story of the success of your campaign and provides more accurate reports.

3. Better Optimization of Ad Campaigns

Having a clearer and more nuanced understanding of your ad campaigns can help you optimize them in real-time when you implement DDA in conjunction with automated bidding strategies, such as programmatic advertising. 

When used in this way, data-driven attribution can drive more conversions at the same cost-per-acquisition, increasing sales and improving return on ad spend (ROAS).

Data Requirements for the DDA Model

One drawback of the DDA model is that the algorithm requires large amounts of data in order to track conversions accurately. For this reason, the DDA model won’t be available for all types of conversion actions.

To be eligible for data-driven attribution, most conversion actions will need to have at least:

  • 300 conversions 
  • 3,000 ad interactions in supported networks within 30 days

Since eligibility criteria are tied to the advert rather than the advertiser, you might find that some of your ads get data-driven attribution while others use MTA models. 

Which Attribution Model Should You Use?

This is a very personal choice for a brand, and the attribution model you ultimately go with depends on the company’s context and how you want to track your conversions. 

However, advertisers who are looking for more precise attribution data would do well to use data-driven attribution for at least some of their campaigns. 

Data-Driven Attribution Will Improve Campaign Effectiveness

When combined with programmatic advertising, data-driven attribution in digital marketing can inform hyper-targeted, extremely effective ad campaigns that yield greater ROAS than campaigns using other attribution models.

While DDA presents an exciting opportunity for advertisers to make their campaigns even more effective, harnessing and implementing DDA models goes beyond the capacity of many in-house teams. 

That’s why many businesses partner with trusted programmatic specialists like Grapeseed Media to deliver expertly crafted and optimized bespoke campaigns that help them grow. 

Get in touch with us to learn more about how we can help you.