SourceWhale email tracking: what to measure and how to test a change

Decide whether this is a measurement or delivery question

“The campaign has many opens and few replies” is a measurement question. “A recipient sees a warning on the link” is a message or filtering question. They can occur together, but they need different evidence.

Start a confirmed spam incident with the SourceWhale troubleshooting guide. This page focuses on interpreting engagement and comparing a specific message change. It does not claim that disabling tracking fixes spam.

Know what each event can establish

Open tracking commonly uses remote image loading. Apple documents privacy protection that can download remote content independently of a person viewing the email. Image blocking can also leave a genuine read unrecorded.

Mailchimp documents automated opens and clicks from security services and link previews. Its filtering features are Mailchimp features; do not assume SourceWhale uses the same filtering or reports the same definition. Ask SourceWhale how its current report handles automated activity, unique events and repeated events.

Avoid follow-ups that tell a candidate you know they read an email. The tracking event is not reliable evidence for that statement.

Inspect the generated links before changing settings

In an authorised test message, compare the editor’s intended destination with the link in the received message. A tracking redirect can send a request through another domain before reaching that destination. Mailchimp’s documentation illustrates this mechanism, rather than documenting SourceWhale’s implementation.

Record these separately:

• Open measurement: whether the received message contains a measurement image, where identifiable.
• Click measurement: whether a link passes through a tracking redirect.
• Custom tracking domain: whether your account uses a branded domain for a tracking function, if supported and configured.
• Destination: the final page the recipient is meant to reach.
• Signature content: logos, social links, banners and any additional redirects.

A branded tracking hostname does not itself prove correct setup, recipient trust or inbox placement. Use only the domain records issued for your account. We have not verified a universal SourceWhale tracking-domain setup or a current menu path for every account, so this guide does not supply one.

If you cannot establish which control affects which part of the message, ask support before changing it. A setting label is less useful evidence than the received message showing the intended change.

Run a comparison you can interpret

Our suggested method is a small, authorised diagnostic comparison, not a campaign to real candidates.

1. Write the question. For example: “Does the generated link-tracking variant produce a different placement observation through this connection?”
2. Save a baseline. Record sender, route, sanitised message version, selected test environments and current settings. Repeat the baseline if practical to see whether it is stable.
3. Change one thing. If your account permits it, compare the click-tracking setting while keeping open tracking, copy, signature, destinations and attachments unchanged.
4. Verify the message changed as intended. Inspect an authorised received copy. If links still redirect, record that instead of assuming the setting removed all redirects.
5. Repeat comparable observations. Keep the route and test coverage consistent. Record sending time, missing messages and any other changes you could not control.
6. Choose a bounded conclusion. Say what happened in these tests, what remains uncertain and what the account owner will do next.

The manual GlockApps workflow explains how a seed test can support step 2. A direct mailbox test is a different route if the affected message was generated through a SourceWhale campaign. There is no claim here of a native integration or a completed experiment.

Keep signature and attachment comparisons separate

A reduced signature can be a sensible preview to investigate broken images or unexpected destinations. It is not automatically an inbox improvement. Keep accurate sender information and the required opt-out mechanism intact.

For an attachment question, use a harmless synthetic file of the relevant type that you are authorised to share. Do not upload a real CV, client brief or confidential offer into external test mailboxes. Record if sanitisation changes a property relevant to the investigation, such as the destination domain or attachment size.

If you remove a logo, replace a tracked link and drop an attachment in one variant, label it a combined message revision. You cannot attribute its result to tracking alone.

Write down the result without inventing a cause

An illustrative conclusion might read:

The two message versions produced different outcomes in the selected test accounts. The sender and connection were held constant. Testing time differed, and recipient-side filtering was not controlled. The result warrants a further check; it does not establish that tracking caused the original incident.

No result is being reported in that example. It is a template for recording the limits of your own observations. Use the blank test record to keep a baseline and follow-up together.

Judge the campaign by its purpose

For candidate outreach, a useful human response matters more than an impressive open count. For an expected interview update, successful arrival and a clear next action matter. Keep those message types separate in reporting.

Use delivery failures, complaints, opt-outs and genuine responses alongside cautious engagement metrics. Do not move unresponsive recipients into a more persistent sequence solely because a tracking event appeared. The delivery metrics guide explains how to keep the denominators and evidence separate.