How to Test and Improve Your Email Marketing Strategy
27 May 2025 · 3 min read
Build a central tracker for monthly hypotheses
Most ecommerce brands run email marketing entirely on vibes. They send whatever campaign idea comes to mind on Tuesday morning, glance at the revenue dashboard, and move on to the next task.
That is a mistake. If you want to scale past $100,000 a month in attributable email revenue, you need Kaizen, which simply means a state of continuous improvement.
One thing I like to do is set up a dedicated testing board using Miro or a clean spreadsheet. You approach every single month with one specific hypothesis, run controlled tests in your campaigns, document the winner, and systematically feed that learning back into your broader marketing strategy.
The best place to run rapid experiments is your broadcast campaign schedule. In a recent test for an ecommerce client, we spent a month testing three distinct unique selling points across 6 separate campaigns.
We tracked opens, clicks, and revenue per recipient for each angle. The clear winner was an educational hook built around overcoming procrastination, which beat our standard product-focused angles by a noticeable margin.
Here is the thing: most businesses stop once the campaign ends. They log a win in their head and do nothing with it. That data is useless unless you operationalise it.
Feed campaign winners directly into your automated flows
This single step is what separates an average email programme from an exceptional one. When a campaign angle wins, you immediately open your Klaviyo flows and update your automated copy to reflect that finding.
In our case, we went straight into the welcome flow and abandonment sequences to rewrite key hooks around that winning procrastination angle. Flows generate revenue 24 hours a day on autopilot, so giving them higher-converting copy permanently lifts your baseline revenue.
Treating flows as a set-and-forget asset is lazy. Your automated flows should evolve every 30 days based on what your live audience is actually clicking and buying in your campaigns.
Optimise send times with structured split tests
Once your core angles are validated, you can move on to testing operational variables like send timing. In one testing cycle, we ran 11 distinct split tests comparing a 7:00 a.m. morning send against a 6:00 p.m. evening send.
The evening window between 6:00 p.m. and 9:00 p.m. consistently outperformed the morning send across engagement and conversion metrics. The difference was not subtle.
We immediately locked in that evening window for all future campaign sends. Even better, we adjusted every flow delay longer than 4 hours so automated emails landed during that high-converting evening window instead of during busy working hours.
Test layout structure and dynamic product blocks
Design and layout testing should follow the same disciplined formula. In a subsequent monthly cycle, we tested the impact of embedding dynamic product recommendation blocks against static product grids.
Dynamic product feeds won decisively across click-to-open rates and average order values. Once we confirmed the numbers, we updated every legacy flow template that lacked dynamic blocks.
After 6 to 12 months of running this continuous cycle, your retention engine looks nothing like the baseline you started with. Every flow email reflects dozens of historical split tests that have already proven to convert your specific audience.
Final Thoughts
Continuous improvement is not complicated, but it requires relentless consistency. Pick one clear hypothesis each month, test it across your campaigns, document the data, and update your automated flows with the winning variation. Those small, compounding improvements are how you build a dominant retention channel.
