A metric is useful only when its event, denominator, time window, and scope are clear. “Reply rate went up” is incomplete until those four details are visible.
A practical metric dictionary
| Metric | Definition | Important distinction |
|---|---|---|
| Sent | Messages handed to the sending path in the selected scope and period. | This differs from unique contacts and delivered messages. |
| Delivered | Messages with no recorded delivery failure in the reporting pipeline. | Delivery does not confirm placement, reading, or attention. |
| Hard bounce | A permanent delivery failure such as an invalid recipient address. | This differs from a temporary deferral. |
| Soft bounce | A temporary or policy-related failure that may require retry or investigation. | Retry decisions still need limits. |
| Reply | An inbound message associated with campaign activity. | A reply may be positive, negative, or neutral. |
| Reply rate | Replies divided by delivered messages in MailSequence aggregated campaign reporting. | Always confirm the denominator before comparing rates. |
| Positive reply rate | Replies classified into predefined positive dispositions divided by a stated base. | This excludes objections and stop requests. |
| Complaint or unsubscribe | A recipient signal that requires suppression and possible broader intervention. | This is more serious than ordinary negative sentiment. |
| Health | A scoped operating signal based on available inbox or domain evidence. | It is not a placement guarantee or universal reputation score. |
Read the operating funnel in order
- Attempt and delivery: Was the system able to send, and what failed?
- Recipient response: Did people reply, unsubscribe, complain, or remain silent?
- Reply quality: Which replies show interest, meetings, objections, wrong-person outcomes, or do-not-contact requests?
- Sequence contribution: Which version and step produced the outcome?
- Sender health: Are mailbox, domain, DNS, or provider signals changing?
Keep denominators and scope visible
“Replies divided by delivered” and “replies divided by sent” can both be legitimate definitions, but they are different metrics. State the formula, date range, workspace, campaign, version, and whether the count includes unique people or messages. Keep the definition stable when comparing periods.
Separate warmup from campaign activity. Warmup messages have a different purpose and recipient system; mixing them into campaign engagement would inflate volume and obscure the actual outreach result.
Delivery and inbox placement measure different outcomes
A receiving system can accept a message and still classify it away from the primary inbox. MailSequence does not market universal placement measurement. Use provider diagnostics, authentication results, bounce responses, complaint evidence, and campaign outcomes together.
Base decisions on stronger signals than opens
MailSequence does not market campaign open or click tracking as shipped. Tracking pixels can be blocked, cached, or preloaded, and Google says it does not track open rates or verify third-party open-rate accuracy. Optimize for truthful delivery evidence, qualified replies, and the downstream conversations the campaign was meant to create.
What MailSequence reports
- Campaign: sends, delivered messages, delivery success, replies, reply rate, selected sequence version, step breakdown, and disposition breakdown.
- Inbox: campaign and warmup sends, delivery, hard and soft bounces, replies, reply rate, and health.
- Domain: volume and available health evidence alongside DNS status.
- API: workspace-scoped daily analytics for customer-owned reporting.
Sources and product basis
- IETF RFC 3463: Enhanced mail system status codes
Standard categories for persistent and transient delivery outcomes. - Google: Email sender guidelines
Sender requirements, recipient feedback, unsubscribe practices, and open-rate caveats. - MailSequence analytics and reporting
Current reporting levels, formulas, API access, and product boundaries.