Email Measurement
Choosing an email programme goal that can be measured
Choose an email goal tied to an observable action, define its counting rule and report results within the data's limits.
Choose an email programme goal by naming an action the email can reasonably help produce, then decide how to count it. “More engagement” is too vague to guide a send. “Help eligible workshop subscribers book a session after an invitation” names an audience and action, but leaves a question: can bookings be reliably connected to that email?
Start with the decision
Ask what the team will change if the result falls short. A useful goal might help decide whether to continue a newsletter format, improve a booking page or change the audience for a guide. If a metric cannot inform a decision, it is probably a dashboard number, not the goal.
Choose one primary outcome for the period. Other signals can explain it, but do not replace the chosen measure after seeing the results.
| Intended outcome | Possible measure | Boundary to state |
|---|---|---|
| Find a practical guide | Recipients with a recorded click on its tracked link | A click does not prove the page loaded or was useful. |
| Enquire about a service | Completed enquiry forms | Check whether submissions can be connected to the email and duplicates handled. |
| Book a session | Completed bookings | Define the observation window and treatment of cancellations. |
These are example measurement designs, not observed results. If the final action cannot be reliably connected to an email, measure an earlier action and name it accurately.
Set the counting rule before sending
Define the event, group and observation window. For a guide-click goal, decide whether repeated clicks by one recipient count once, which tracked link counts and whether the denominator is emails sent or successfully delivered. Record sent and bounce counts beside a rate so changes in delivery remain visible. Check that the link leads to the intended page.
Custom campaign URLs can help identify campaign data in Google Analytics when it is set up to collect it. Tags alone do not prove that an enquiry or sale occurred.
Use inbox measures as diagnostics
Email platforms can report opens, clicks, deliveries, bounces and unsubscribes, but availability and definitions vary. Open tracking records a technical event, not proof that someone read the message. Privacy and automated security activity can also affect open and click figures. Check the platform's tracking and filtering settings before interpreting small changes.
If a guide email records clicks but few completed enquiries, investigate the page, offer, tracking and audience fit. The email metric narrows the question; it does not answer it by itself.
Key Email Metrics and Their Limitations
- Open Rate
- Technical event, not proof of reading
- Click Rate
- May not indicate page load or usefulness
- Delivery Rate
- Shows successful delivery, not engagement
- Bounce Rate
- Indicates delivery failure; check filtering settings
Write a short measurement brief
Use this structure: “For [eligible audience], this programme aims to help people [action]. We will count [event] for [group] during [window]. We will also watch [guardrails].”
Guardrails might include unsubscribes, complaints or an operational problem that makes the result misleading.
Report raw counts with any rate, note tracking gaps and state only what the data supports. “The invitation recorded more booking-page clicks” is narrower than “email generated more bookings”.
