How Mixsoon Outperformed Default Flows by Up to 22.9x

Moving From “Recent Visitors” to “Today’s Buyers”

In the world of high-growth skincare, relying on “current interest” triggers for price drops and low inventory alerts is a recipe for leaving money on the table. We worked with mixsoon to move beyond the narrow window of recent site activity and toward a predictive model that identifies demand before the customer even returns to the store.

By replacing standard event-based triggers with Retentics AI, mixsoon didn’t just optimize their existing flows; they enabled a new system that generates up to 22.9x more revenue by reaching the “invisible” audience that standard flows miss.

By targeting forecasted intent rather than just recent clicks and website engagement, they turned simple alerts into a massive revenue multiplier.

The A/B Testing Methodology

We implemented a methodology for mixsoon that bypassed surface-level metrics to isolate the true driver of incremental revenue:

• Defining the Scope:

  We didn’t just test one email against another. We compared two distinct strategies: targeting users based on recent clicks and website engagement (Current Interest) vs. targeting users statistically likely to buy (Forecasted Interest).

  
  

• The 90-Day Window:

  Skincare has a specific rhythm. A three-month test allowed us to account for product shelf cycles and natural replenishment rhythms.

Running the test:

• 90-day testing window:

  This allowed us to account for varying product shelf lives—from daily SPF to targeted hair treatments—and the natural restock behavior of Cardon’s customers.

  
  

• Behavioral Accuracy:

  Throughout the run, we ensured alerts reached the right customers at the precise moment of need. We prioritized high-intent orders over mere clicks.

Analysis and Decision Making

For mixsoon, we focused on the “multiplier of success” rather than simple percentage growth. We compared the absolute revenue generated by the new triggers against the legacy logic to see exactly how many times more effective the new system was.

The 2 KPIs That Actually Predict Bottom-Line Revenue

These KPIs formed the foundation of the mixsoon test when evaluating Retentics AI against their previous default flow setup.

KPI 1: Low Inventory Revenue Multiple

• The Result: 22.9x Multiple.

• Why it matters:

  Most default flows only hit people who recently engaged with a product. This KPI proves that if you target the “forecasted” audience — people who need a restock but haven't visited the site yet — you can dramatically scale your results.

KPI 2: Price Drop Revenue Multiple

• The Result: 9.4x Multiple.

• Why it matters:

  A price drop is only effective if it reaches the right eyes. By expanding reach to forecasted buyers, we proved that a price change can drive 9.4x more revenue than standard triggers.

The Key Health Metrics That Support Performance

• Audience Expansion:

  Moving beyond “yesterday’s visitors” allowed mixsoon to reach customers who were ready for their next bottle of Bean Essence but hadn’t clicked a link recently. This is the most efficient way to grow without increasing top-of-funnel spend.

  
  

• Speed to Value:

  mixsoon didn’t need a massive technical overhaul. They paired their existing creative with our smart triggers and were live in two days.

Final Thoughts

Success in repeat purchasing isn’t about volume; it’s about landing at the peak of intent. Relying on “current interest” alone leaves significant revenue on the table. When you switch to “forecasted interest,” you stop chasing customers and start meeting them exactly where they are.