> ## Knowledge Base Index
> Fetch the complete knowledge base index at: https://help.seino.ai/sitemap.xml
> Use this file to discover available pages before exploring further.
> Pure-Markdown content can be obtained by appending a '.md' suffix to the content URLs listed in the sitemap (without the trailing slash).

# Choosing the right metric

A target is only as useful as the metric behind it. The best metric for a target ties to revenue or retention, can be influenced by the next campaign, has clean data feeding it, and isn't distorted by the way modern inboxes work.

## Think in three tiers

Most email programs need a small set of metrics arranged in tiers — start at the top and work back to the levers.

| Tier | Includes | Role | Availability |
| ---- |
| **Tier 1 — Leading indicators** | Click-through rate, click-to-open rate, click-to-conversion rate | Move *before* Tier 1 — give you time to act | Click-to-conversion depends on whether your ESP tracks conversion events |
| **Tier 2 — Outcomes** | Revenue per email, email-attributed revenue, total email revenue | What the business cares about | Depends on whether your ESP exposes attributed revenue |
| **Tier 3 — Guardrails** | Unsubscribe rate, bounce rate, complaint rate, deliverability | Catch failure modes (over-mailing, list fatigue, deliverability drops) | Universally available |

A plan with one Tier 1 metric and one Tier 3 guardrail is usually enough to start. Tier 2 is what you watch weekly to predict where Tier 1 is heading.



![](https://storage.crisp.chat/users/helpdesk/website/-/3/c/9/8/3c9863e1e93c4800/shapes-at-26-06-30-143839_1x5j8sz.png =824xauto)

## Pair leading with lagging

Every lagging outcome target should ship with at least one leading indicator nearby. Revenue (lagging) tells you what happened last month; click-to-conversion (leading) tells you what's about to happen. Without the pair, you only know you missed once it's too late to do anything.

## Demote open rate

Apple Mail accounted for **roughly half of all email opens** in early 2025 and pre-fires the tracking pixel via proxy servers — opens get logged whether the subscriber read the email or not. Real-world cases show campaigns jumping from ~28% to ~52% open rate with no change in clicks or conversions.

Treat open rate as a directional signal at best. Put your target on something further down the funnel — clicks, conversions, replies, time-on-site after a click.

## Be specific enough to act on

"Total revenue this year" is too broad and too slow — by the time it's slipping, three quarters of decisions are already made. "Revenue from promotional campaigns to active subscribers, this month" narrows the *who* and the *when*, so a missed week has an obvious next move (test subject lines, swap creative, re-segment).

A good target points at something you'd know how to change if it started slipping.

## Pick a metric you have clean data for

A target is only as accurate as the data feeding it. If a metric depends on a UTM that's often missing, or an attribution window that doesn't match how your audience buys, the pacing won't reflect reality. Start with metrics you trust today.

## Limit the total list

Two to three metrics per plan, no more than a dozen total across all plans. Past that, no one looks at any of them.

## Examples for email teams

* Track **revenue per email sent** (Tier 1) for promotional campaigns, paired with **unsubscribe rate < 0.5%** (Tier 3) as a guardrail. *Requires an ESP that exposes attributed revenue.*
* Target **click-through rate from the welcome series** (Tier 2) paired with **complaint rate < 0.1%** (Tier 3) as a guardrail — both universally available.
* Target **email-attributed revenue from active subscribers** (Tier 1) with **complaint rate** (Tier 3) — keep the focus on the segment that buys without burning the rest of the list. *Revenue metric depends on your ESP.*
* Build an **engagement rate** target as a custom metric — formula: `(opens + clicks_unique) / sent`. Universally available. For more recipes, see [Custom metric recipes and combined targets](custom-metric-recipes.md).
* Use **unsubscribe rate** (Tier 3) as a guardrail metric — catch list fatigue before it eats deliverability. GDMA's global average is ~0.14%; below 0.2% is healthy; above 0.5% is a warning signal.

## Start with one or two

You don't need a target on every metric. One revenue target plus one guardrail is usually enough for a first plan. You can grow from there once the rhythm is established.

**Tip:** Not sure which metric to start with? Ask yourself "If this number moved 10% in the wrong direction, would I notice — and would I know what to do?" If the answer is yes, it's a good candidate.