Afterpay rebuilds marketing ops with new CDP and data stack

Afterpay has overhauled its marketing technology stack, deploying Hightouch to turn its Snowflake environment into a customer data platform (CDP).

From left: Kashish Gupta (Hightouch), Bree Graham and Matthew Lanki (Afterpay)

The buy now, pay later provider is using Hightouch and Snowflake to streamline customer segmentation, automate campaign delivery and enable cross-channel experimentation at scale.

It previously relied on manually built customer lists, static data export and siloed campaign execution across platforms like Facebook Ads, Google Ads and Mailchimp.

However, audiences are now centrally defined in Snowflake and automatically delivered to marketing platforms via Hightouch, eliminating repetitive work for the 22-person marketing team.

The team can also self-serve access and activate customer segments without needing data teams.

Speaking at Snowflake Summit in San Francisco, Afterpay’s marketing technology lead Bree Graham said the uplift has enabled her team to make “immediate business decisions” without relying on the company’s data scientists.

“[Now we are able] able to create one [customer] list, [and] send that into multiple different platforms based on the experiences that our consumers are doing; whether they engage in outcomes, whether they make a purchase, and whether they don't.”

An acquisition and a reset

Afterpay, which popularised the “pay in four” model, was acquired by Block for $40 billion in 2022 in what is Australia's biggest-ever acquisition.

Before then, according to data science manager Matthew Lanki, the company’s data and marketing processes were “pretty scrappy”.

However, its acquisition by Block and the ensuing adoption of Snowflake as its data warehouse spurred a wider rebuild of Afterpay's data infrastructure and martech approach.

Subsequently, Hightouch, a third-party reverse ETL platform that integrates with Snowflake, was selected as Afterpay’s customer data platform.

Following its deployment at the end of last year, Afterpay has improved efficiency and reduced several manual processes for its online marketing activities.

This includes having to create separate lists of customers for each marketing platform, such as Facebook Ads and Google Ads.

Lists had to be exported as Excel files, manually formatted to meet each platform’s individual requirements, such as column headers or user ID formats, and then uploaded one by one.

However, the Hightouch platform synchronises these audience lists directly into the various platforms automatically while also allowing Afterpay to track and analyse user behaviour across marketing platforms using a central audience list.

“We can create those experiences,” said Graham. “We can optimise and change that experience [and] also change the criteria behind that, all with one list, all in one place.”

Experimentation and emerging tech

The other key benefit for the Afterpay marketing team was to boost experimentation with different strategies.

Previously, Lanki’s team was able to run a standard A/B test – comparing different campaigns in one media channel – relatively easily using customer lists in Snowflake and uploading them manually to the marketing platform.

However, they struggled to run tests across multiple platforms and were only able to rely on “static” customer lists -- datasets that require manual updates.

Lanki said the adoption of Hightouch now allows Afterpay to easily run cross-channel experiments using data that is automatically updated and synced to all platforms daily.

The platform also synchronises exclusion lists to Snowflake and Afterpay’s SMS/email marketing platform Iterable, which makes it easier for Lanki to “jump in” and analyse the results of the experiment.

Looking ahead, Afterpay is now working with Hightouch’s AI Decisioning platform to improve campaign delivery.

“Being a buy now, pay later [operator]... we're always trying to connect our consumers with merchants, and trying to surface which merchant a customer should go and shop at based on their person [and] based on their previous purchasing behaviour,” Lanki said.

“We’re trying to understand when we should send certain customers information about which merchants would be relevant to them to shop at. This is obviously a very hard task, and we want to make sure we're not just bombarding them.”

The company has two AI-driven use cases currently in development to help determine the right message for the right person at the right time.

According to Graham, the marketing uplift has “really been able to expedite and scale” Afterpay’s AI development in just a few months.

“We’re really excited for what the future looks like,” she said.

“[We’re] excited that we took some learnings, [built] new opportunities and really have those to help us bring it to life."

Eleanor Dickinson attended Snowflake Summit in San Francisco as a guest of Snowflake.