Be ‘data-driven’ said, well, everybody. So why is ‘being data-driven’ still an unachievable ambition to so many online businesses?

One challenge is the abstract nature of the term: ‘data-driven’. The barrier, simply, is knowing what being data-driven looks like. When you know where you’re going, it’s much easier to get there.

This post outlines top-level techniques to leveraging data to increase acquisition and retention for ecommerce businesses. Spoiler: we’ll equate ‘being data-driven’ to ‘leveraging first-party data’ (purists might baulk, but that’s the 80/20 of it).


1. Quicker Primer: 1st vs 3rd party data

2. Barriers To Leveraging 1st Party Data

3. Applications Of 1st Party Data:

a) Increase acquisition

b) Increase retention

c) Website experience

d) Cross-channel application

4. How To Activate 1st Party Data


Third-party data refers to information collected about users from another company. This type of customer data is typically aggregated from various sources and pooled together. Recent privacy laws make this data harder to collect for advertising/targeting purposes.

First-party data is collected by brands directly from their audience and customers. Common datapoints are name, email address, products purchased and price. Common data sources are CRM and sales data, in-store purchase data, and customer survey data.

Third-party data is readily access on platforms like Facebook Ads. The skilled media buyer can leverage these data signals (click-through rates, watch rates, audience types, etc.) to optimise campaigns that increase sales. The problem is that in reality such optimisation is typically done by the same algorithm that’s managing everybody else’s Facebook campaigns.

Companies seeking to be data-driven should leverage first-party data because:

a) it’s a competitive advantage (unique datapoints);

b) it’s a more powerful (read: profitable!) targeting method compared to third-party targeting;

c) beyond paid traffic spend (the most profitable use case), understanding first-party data has repercussions that extend to all marketing channels and indeed across the entire business.


1. Data silos. Ecommerce performance is hidden behind dozens of logins (Facebook Ads, Google Ads, Shopify, Google Analytics, etc.).

2. Lack of expertise (data engineers). The modern data stack (a set of tools for moving and storing data) is extremely fragmented, requiring data collection, ingestion, storage, modeling and reporting tools – to name a few. Each of them have different best-in-class vendors.

3. Data engineering costs. Due market economics of supply and demand, it’s never been more expensive to hire a data engineer.

4. Lack of resource (data consumers). Once you have the data, someone needs to analyse it. That’s often someone internal to the organisation with deep subject matter expertise who is very close to ecommerce initiatives and decision-making processes. Unfortunately, such individuals are also largely preoccupied with launching this year’s Black Friday campaign.

5. SaaS tools pretending they can do it. There’s no one-size-fits-all approach for marketing optimization. But that doesn’t stop a cluster of vendor tools selling plug-and-play reporting templates that simply surface data and offer zero value.


1. Increase Acquisition

Users are targeted based on propensity to purchase, when they last purchased, product viewed, and a dozen other parameters. In turn, this increases click-through rates and conversion rates for prospecting and retargeting activities.

2. Increase Retention

Drive repeat purchases with personalised email marketing efforts and customer retargeting campaigns. Marketing is more relevant when supplemented with the additional customer data that’s collected at the point of purchase.

3. Website Experience

Brands can personalise the website experience based on customer cohorts. Cohorts can be based on frequency, recency, monetary or category parameters, for example.

More obviously, user experience data collected from the website can be used to increase conversion rates (i.e. reducing bounce rates and funnel abandonment rates), and optimise specific device/browser experiences. To achieve this, common data sources include web analytics (e.g. Google Analytics) and user testing (e.g.

4. Cross Channel Application

First-party data isn’t limited to a specific channel or platform. Learnings from the most profitable audiences can be applied cross-channel; DTC ad spend can target similar audience parameters across Facebook. Google, TikTok et al. This can be a huge head-start when diversifying into new channels, cutting the learning curve by weeks or months.


Companies that seek to become data-driven can implement a modern data stack – a collection of tools needed to centralize and organize data for efficient use.

For ecommerce teams, the modern data stack solves for the ‘data silo issue’ by centralising data from multiple sources in a data warehouse, then pushing that data to reports for upstream analysis, or to go-to-marketing tools for real-time optimisation.

The path to maturity typically looks like this:

1. Automated reporting. Real-time reports, fully-customised and automated.

2. Business intelligence. Enabling brands to ‘self-serve’ data for themselves with a visual query builder e.g. deep-dive into customer cohorts, deepen audience understanding and create new data-driven initiatives.

3. Data Operationalisation. Pushing automated audiences to platforms like Google, Facebook and Klaviyo.

4. Customer Data Platform (CDP). Turning the data warehouse that enables 1-2-1 or one-too-many hyper-personalised communications.

Not every company has the expertise to navigate these infrastructure decisions alone. Our managed data stack services help to set up modern data stacks, create automated reports and operationalize data. So brands can 10x the value of their data without 10x’ing their time.