Email frequency optimisation

Email frequency optimization is the practice of finding the number of emails per subscriber per period that maximises total engagement and revenue without eroding the list through unsubscribes, spam complaints and quiet disengagement. It treats send cadence as a dial to tune, not a routine to repeat. Sending three emails a week might earn more revenue than two, or it might earn less once you subtract the subscribers who tune out or leave.

The core tension is simple. More sends create more chances to convert in the short term, but each extra send also raises the risk of fatigue. A subscriber who receives one useful email a week may open every one. The same subscriber receiving five emails a week may stop opening any of them, mark one as spam, or unsubscribe. Optimization means measuring both sides of that trade and settling on the cadence where total value peaks.

Frequency is not one global number. It interacts with segment, lifecycle stage and content type. A new subscriber in an onboarding sequence tolerates a denser cadence than a customer two years in. Optimizing frequency well means tuning it per segment, then watching how engagement, deliverability and revenue respond.

There is no single formula that hands you the perfect cadence. Optimization is an experiment that compares cadences and measures net value per subscriber for each. To run it well, you need to track the inputs that move in opposite directions as frequency rises.

1. 1

   Sends per subscriber per period

   The cadence you are testing, for example two, three or four emails per subscriber per week. This is the variable you are tuning, so it must be measured per subscriber rather than as a total send volume.

2. 2

   Revenue per subscriber

   Total email-attributed revenue for the period divided by active subscribers at that cadence. This is the upside you are trying to grow and it tends to rise with frequency, then plateau or fall.

3. 3

   Unsubscribe and complaint rate

   The share of recipients who opt out or mark a send as spam. This is the direct erosion signal and it climbs as cadence increases. A small rise compounds because every lost subscriber takes all their future revenue with them.

4. 4

   Engagement decay

   The drop in email open rate and click rate as frequency rises. Falling engagement is an early warning that fatigue is setting in well before subscribers actually leave.

5. 5

   Deliverability impact

   Low engagement at high frequency tells mailbox providers your mail is unwanted, which lowers inbox placement. A cadence that quietly pushes you to the spam folder destroys value across the whole list, not just the over-mailed segment.

In practice you run a frequency experiment. Split a segment into groups, give each a different cadence, hold content and timing as constant as you can, and run it long enough for unsubscribe and deliverability effects to surface. Then compare net value per subscriber across groups. The winning cadence is rarely the highest one tested, because the erosion cost usually overtakes the revenue gain before you reach daily mail.

A metric tree turns frequency from a gut-feel decision into a diagnosis. It decomposes the net value of a cadence into the forces that grow with each extra send and the forces that erode the list as you push harder.

The first level splits into revenue per subscriber, list erosion and deliverability health. Revenue per subscriber breaks down into how many of your sends are opened, how many opens click through, and how much each click is worth. List erosion breaks down into unsubscribes and spam complaints. Deliverability health depends on engagement signals and complaint volume, which together decide whether the next send lands in the inbox at all.

This structure lets you see why a cadence works or fails. If revenue per subscriber is flat but unsubscribes are climbing, you are over-mailing and should pull back. If both revenue and engagement keep rising with no erosion, you have headroom to send more. The tree tells you which lever moved, and each lever has a different owner.

Benchmarks for cadence depend heavily on sender type and list intent. A daily newsletter and a B2B nurture programme live at opposite ends of the scale. The numbers below are starting ranges, not targets. The right cadence is whatever your own frequency experiment proves sustainable.

The most reliable benchmark is internal and directional. Plot revenue per subscriber and unsubscribe rate against cadence for your own list. The optimal frequency sits just before unsubscribes start climbing faster than revenue. A complaint rate above 0.1 percent is the line most mailbox providers treat as a deliverability risk, so treat it as a hard ceiling regardless of revenue.

Improving frequency optimization means replacing a fixed habit with a tuned, segment-aware cadence that you keep testing. The aim is to extract more value per subscriber while protecting the list and the inbox placement that makes every future send possible.

The metric tree approach starts by finding which branch limits net value. If revenue per subscriber is plateauing while erosion climbs, the cadence is too high and the fix is to pull back or segment. If engagement is strong and erosion is low, there is room to test sending more.

KPI Tree lets you model this by connecting each branch to the team that owns it. The content team owns relevance, which decides how much cadence a subscriber tolerates. The lifecycle team owns segmentation and the preference centre. The deliverability owner watches complaint volume and inbox placement. With RACI ownership on every node and a push to the accountable owner when unsubscribes or complaints move, a rising erosion signal reaches the person who can act before it costs you the list.

1. 1

   Optimising for this weeks revenue only

   Judging a cadence by a single sends revenue ignores the future value of subscribers who unsubscribe. A send that earns more today but erodes the list loses money over the year.

2. 2

   Using one global frequency

   A single cadence for the whole list over-mails the disengaged and under-mails the eager. Frequency should be tuned per segment, because tolerance for sends varies enormously across the list.

3. 3

   Ignoring deliverability feedback

   Rising frequency with falling engagement tells mailbox providers your mail is unwanted. Tracking only opens and revenue, while ignoring inbox placement and complaints, hides the most damaging cost of over-mailing.

4. 4

   Confusing send volume with frequency

   Total emails sent is not the same as emails per subscriber. A growing list can mask a rising per-subscriber cadence. Always measure frequency per recipient, not as an aggregate.

5. 5

   Never revisiting the cadence

   A cadence that was optimal a year ago may be wrong now. List composition, content and competition change. Frequency optimization is a recurring experiment, not a one-time decision.