The single most important concept for replenishment flows is that you are not creating demand, you are simply catching it. Nik Sharma made the point on Limited Supply that these emails outperform almost every other campaign because the customer's intent to repurchase is already there. The entire game is to get your reminder in their inbox at the exact moment they need it. The consensus is that moving away from simple, static time delays and toward predictive, behavior-based triggers is the key to perfecting this timing.
Historically, brands relied on basic automations, like the classic 90-day win-back flow. As Sonja Grasser pointed out on Honest Ecommerce, this is often far too late. By the time 90 days have passed, a customer who needed a refill at day 45 has already found an alternative and moved on. The fundamental flaw with a one-size-fits-all rule is that it ignores a customer's personal purchase cadence, the product's specific consumption cycle, and individual usage habits. Sending a reminder too early feels pushy, while sending it too late results in a lost sale. This is why a simple "wait X days" trigger is the wrong way to think about replenishment.
Modern email platforms provide a much more sophisticated toolkit. On an episode of Honest Ecommerce, Klaviyo's own data science team discussed moving past arbitrary rules you might set up in a few seconds. Instead, they recommend using predictive properties like a customer's "expected date of next order." This feature uses a customer's past purchase history to calculate a surprisingly accurate forecast of their next purchase date. Using this date as the trigger for your flow is what enables true 1:1 personalization. The email isn't sent 30 days after the last purchase, it's sent when that specific customer is algorithmically determined to be ready, whether that's 28 days for one person or 52 for another.
Beyond predictive timing, effective replenishment requires product-level segmentation. As Aaron Orendorff mentioned on the Up Arrow Podcast, you have to get savvy about what the person actually ordered. A customer who bought a 15-day trial-size starter kit has a completely different replenishment need than a customer who bought a 90-day supply of the same product. Your flows should be forked with different timing and messaging based on the SKU or product category. A customer's first reorder is also a critical moment that has compounding effects on their lifetime value, so tailoring the follow-up to their specific first purchase is crucial.
Ultimately, a perfectly timed replenishment flow is a core pillar of Customer Lifecycle Marketing. Building on the predictive timing from Klaviyo, you can also incorporate other behavioral signals that Sonja Grasser mentioned, such as a customer service ticket or a significant lag in their typical visit frequency. These can be inputs into an even smarter model that also considers churn risk. If a customer is highly engaged and has a low churn risk, you can probably be less aggressive. If their next expected order date is approaching and they are flagged as being at high risk of churning, that's the perfect time to trigger a flow with a helpful reminder and perhaps a small incentive to secure the order.
So, where should you start? First, if you don't have access to predictive analytics, calculate the average time between orders for your top 3-5 products and use that as a starting point to create separate flows for each. But if you're using a platform with more advanced Klaviyo Strategies, you should immediately transition your triggers to use the "expected date of next order" property. This single change will dramatically improve your timing. From there, you can layer on product-specific logic and more advanced churn prediction. The goal is to make your email feel like a timely, helpful service, not just another marketing message. When done right, you are simply making it easy for customers to buy something they already want.
