The most common email marketing KPIs in ecommerce are open rate, click rate, conversion rate, and revenue per send. Every ESP dashboard puts these front and center. Every email team reports on them. And every one of them can be improving while your business gets worse.

That’s not an exaggeration. It’s arithmetic.

The problem with revenue per send

Revenue per send (or revenue per email) divides the revenue attributed to a campaign by the number of emails sent. It’s the metric most email teams optimize for, and it’s the metric most likely to mislead you.

Here’s why. Imagine two scenarios:

Scenario A: You send a promotional email to your most engaged 50,000 subscribers. Revenue per send is $0.80. The team celebrates.

Scenario B: You send a broader campaign to 200,000 subscribers, including less engaged customers. Revenue per send drops to $0.35. The team worries.

But Scenario B generated $70,000 in total revenue versus Scenario A’s $40,000. And if even 5% of those less engaged customers in Scenario B made a purchase they otherwise wouldn’t have, Scenario B created more incremental value and potentially reactivated customers who will generate future LTV.

Revenue per send punishes you for reaching further into your list. It rewards you for cherry-picking your most engaged audience. In other words, it optimizes for efficiency at the expense of total value, which is the opposite of what drives LTV.

The metrics that actually predict LTV

If customer lifetime value is the metric that determines whether your business model works (and it is), then your email program should be measured by its contribution to LTV, not by per-send efficiency metrics. HubSpot’s 2026 State of Marketing Report shows email is the #1 ROI-driving channel for B2C brands, and Litmus data confirms $36–$42 return per dollar spent. But ROI and revenue per send are different things, and optimizing for one can undermine the other.

Incremental revenue per customer

How much additional revenue does each customer generate because of your email program, compared to what they would have spent without it? This requires a holdout test to measure, but it’s the only metric that tells you the true value of your email program. Most brands who run holdout tests for the first time discover their incremental contribution is meaningful but significantly lower than their ESP’s attributed revenue number.

Second purchase rate

After a first purchase, customers are 27% likely to buy again. After the second, 49%. The second purchase is the inflection point in the customer lifecycle. Your email program’s ability to drive first-to-second-purchase conversion is the single best predictor of its LTV impact. If your second purchase rate is below 30%, fixing your post-purchase email sequence will produce more LTV than optimizing any campaign metric.

Repeat purchase rate over time

Not just “did they buy again” but “how many times and over what period?” Track purchase frequency by cohort over 6, 12, and 24 months. If customers who receive your emails buy 20% more frequently than they would otherwise (holdout-tested), that frequency gain compounds across the entire lifespan.

Email list engagement decay rate

How quickly does engagement drop off after acquisition? If 80% of your subscribers are disengaged within 6 months, your email program is failing to maintain relevance over time. The average DTC brand retains just 28.2% of customers for a second purchase, which means most email lists are dominated by one-time buyers who gradually disengage. The best programs show a gradual decay curve, not a cliff. AI-native platforms with customer memory reverse this curve by making messages more relevant over time, not less.

Unsubscribe rate by lifecycle stage

A 0.2% unsubscribe rate sounds healthy until you realize it’s 0.05% for your first 90 days and 0.5% after month 6. That late-stage unsubscribe rate is your email program actively destroying future LTV by burning out customers who would have remained purchasable through a more relevant, lower-frequency approach. Per-customer frequency optimization addresses this directly.

Why email teams optimize the wrong metrics

The reason is structural, not intellectual. Email teams optimize revenue per send and open rates because those are the metrics their ESP surfaces, reports on, and makes easy to act on. The ESP’s business model is built around sends and contacts, so the metrics it emphasizes are send-level metrics.

LTV is a cross-functional metric that spans marketing, product, finance, and customer experience. No single tool owns it. The email team can influence it but can’t measure it without data from the ecommerce platform, the finance team, and (ideally) a holdout testing framework.

This is also why AI-native email platforms represent a structural shift. Traditional ESPs optimize for send-level performance because that’s their unit of measurement. An AI-native platform optimized for LTV makes different decisions: it might send fewer emails to a customer who’s at risk of fatigue (lower revenue per send, higher LTV). It might send a non-promotional, relationship-building message that has zero immediate revenue but extends lifespan (invisible on send-level metrics, visible on LTV). It might reach deeper into the list to reactivate lapsed customers (lower revenue per send, higher total incremental value).

These are all decisions that look “worse” on traditional email metrics and “better” on LTV.

The practical shift

You don’t need to abandon open rates and revenue per send entirely. They’re useful operational signals for debugging campaign issues (deliverability problems show up in open rates, bad creative shows up in click rates). But they should not be the success metrics for your email program.

The shift looks like this:

Report to leadership on: incremental revenue contribution (holdout-tested), second purchase rate, repeat purchase rate by cohort, and email-driven LTV per subscriber. These are the metrics that connect your email program to business outcomes.

Optimize day-to-day on: open rate (as a deliverability health signal), click rate (as a creative quality signal), and unsubscribe rate by lifecycle stage (as a relevance signal). These are diagnostic metrics, not outcome metrics.

Stop optimizing for: revenue per send as a primary KPI. It’s the metric most likely to lead you toward decisions that improve short-term efficiency at the expense of long-term customer value.

The brands with the highest email-driven LTV aren’t the ones with the highest revenue per send. They’re the ones whose email program makes customers more valuable over years, not just campaigns.

LTV.ai measures what matters: incremental customer lifetime value. Our AI-native platform optimizes every send for long-term customer value, not short-term campaign metrics. Book a demo →