There are two problems in using CRM segmentation to build your lookalikes. First, applying multiple filters often results in overengineered audiences. Second, there’ll be limited to no feedback on your selection criteria (input) versus audience content (output). The output might have a size, but no predicted customer lifetime value (pCLV) or propensity to purchase (pConv).

Replace this with Custom Audiences: A beautifully simple way to build Predictive Audiences with a combination of historic and forward-looking data. 

Traditionally, marketers have only ever been able to look back. Now, you can look forward and keep an eye on the rear mirror to drive your advertising forward.

Earlier this year, Programmai launched Acquisition and Retention audiences for specific workflows. We’ve designed and iterated these to help ecommerce brands acquire new and existing customers at scale. 

With Programmai Custom Audiences, the floor’s yours.

  • Experiment with predictive customer lifetime value (pCLV) and propensity to buy (pConv).
  • Layer your audiences with additional data points, such as products or countries.
  • Refine messaging based on your segmentation.

The best chess player thinks three steps ahead.

And listens to the psychology of their opponent. The best marketer does the same: Strategising around long-term gains and short-term tactics, splitting the board and directing each piece accordingly.

The board will never offer the answers. Rather, it offers a playground to ask better questions and find the hypotheses that you can test for meaningful results.

Here’s an example of the Custom Audience in action.

Early on, we built this chart into a dashboard so that you can meet your customers where they are.

With Custom Audiences, you can strategise around each section. For instance:

  • High pConv, low pCLV: These customers are likely to buy in the short term but will need nurturing to become loyal customers (Low pConv, high pCLV).
  • Low pConv, low PCLV: You could ignore these customers altogether. They’re unlikely to buy or continue purchasing your products.
  • High pConv, low pCLV: Leave these ideal customers be. Advertising to this set may result in the ad platforms over attributing sales that would have happened anyway. Excluding this group should result in high incrementality, lower ad spend and a better customer experience. Conversely, you could target them with an ad campaign specifically for high-spending, loyal customers. 
  • Low pConv, high pCLV: These loyal customers may not be ready to buy again. Leave them alone or design a thoughtful nurture.

Once you have your segment, you can target with specific ads, whether to nurture or convert, create Predictive lookalikes from Predictive segmentation, set budgets or simply exclude, forget or ignore.

For the Programmai marketer, Custom Audiences enabled them to focus on today’s trends rather than long-term signals of value.

The future of Activation is Predictive!

Rather than just moving your first-party data from A to B, Programmai makes your data intelligent in transit. Build Predictive Audiences within any ad platform to hit any marketing outcome.

Invest in the future of a customer to drive real bottom-line results.

  • Build and manage predictive audiences to hit any marketing objective.
  • Only reward good behaviour. Build up to 20% lookalikes around positive first-party signals.
  • Report on real-world results with confidence. No gaming metrics.