Learning path · Level 8 · Tutorial 4

Use Email Metrics to Improve Campaigns

Interpret text-based email and automation reports, calculate meaningful rates, and recommend a focused improvement. Account for unreliable opens, automated clicks, delivery failures, and incomplete workflow observation before judging results.

  • Level: Intermediate
  • Duration: 18 min
  • 8 steps
Start tutorial Back to the path
Use Email Metrics to Improve Campaigns

What you will go through

  1. Read the Report Before Calculating Identify what each Starter Kit campaign report count measures and establish its reporting window before interpreting performance. 2 min
  2. Calculate Click, Bounce, and Unsubscribe Rates Calculate the Starter Kit campaign’s click, bounce, and unsubscribe rates while preserving the denominator and reporting window behind each percentage. 3 min
  3. Choose a Meaningful Conversion Rate Calculate recipient-based and session-based conversion rates for Northstar Stationery’s Starter Kit campaign while preserving the correct unit, denominator, attribution context, and reporting window. 2 min
  4. Treat Opens and Clicks as Imperfect Signals Interpret Reported opens and Reported clicks cautiously, then select corroborating evidence without claiming certainty about deliberate human engagement. 2 min
  5. Turn Delivery Failures and Opt-Outs into Actions Use bounce and unsubscribe evidence to select proportionate list-health and eligibility actions without bypassing suppression or overstating causes. 2 min
  6. Compare Workflow Cohorts at the Same Stage Assess Welcome workflow cohorts by entry timing and complete observation windows before judging performance. 3 min
  7. Choose One Evidence-Linked Improvement Use connected evidence to select one proportionate improvement or test while protecting conversion measurement and list health. 2 min
  8. Complete an Evidence-Based Campaign Review Synthesize campaign and workflow evidence into a concise review with one cautious, measurable next action. 3 min

What you will learn

  • Calculate click, conversion, bounce, and unsubscribe rates from example counts while stating each denominator and reporting window.
  • Explain why reported opens and clicks may not represent deliberate human engagement.
  • Use bounce and unsubscribe evidence to recommend appropriate list-health or eligibility actions.
  • Compare workflow cohorts only after accounting for sequence timing and observation completeness.
  • Recommend one evidence-linked improvement or focused test, including a conversion outcome and an unsubscribe or delivery guardrail.

Before you start

  • Draft an Accessible Email Campaign
  • Map Welcome, Nurture, and Post-Purchase Automations
  • Create Consistent UTM Campaign Links
  • Read GA4 Acquisition, Engagement, and Key-Event Reports
  • Design and Interpret a Simple A/B Test

Step 1 of 8

Read the Report Before Calculating

Identify what each Starter Kit campaign report count measures and establish its reporting window before interpreting performance.

Read the campaign snapshot

Start with the report boundary

Before interpreting Northstar Stationery’s Starter Kit campaign, establish what was counted and when.

Reporting window: First 7 complete days after sending

This is a one-time email: each recipient receives at most one message. The report uses both message counts and recipient counts, so equal-looking numbers must not be treated as the same kind of evidence.

Read the labels before making any judgment. A count alone does not tell you whether delivery, engagement, or purchasing performance was good.

Campaign snapshot

Starter Kit campaign report for the First 7 complete days after sending.

Each attempted send has one delivery outcome in this simplified snapshot.
Report fieldCountWhat it counts
Attempted sends2,000Messages the provider tried to send
Delivered messages1,900Messages accepted by the receiving provider
Bounces100Messages not delivered in this snapshot
Unique clicking recipients152Recipients recorded clicking at least once
Attributed purchasing recipients38Delivered recipients with a qualifying purchase
Unsubscribes19Unique recipients who unsubscribed

Separate delivery from attention

Follow the delivery path

The first split is delivery status. In this simplified report, the 2,000 attempted sends resolve into either a delivered message or a bounce: 1,900 + 100 = 2,000.

That reconciliation is useful here because the snapshot explicitly assigns every attempted send one delivery outcome. In a live provider report, first check the provider’s definitions, statuses, and reporting timing before expecting totals to reconcile this way.

A delivered message means the receiving provider accepted it. It does not prove that a person saw or read it.

Delivery is not viewing

This diagram represents the simplified delivery outcome split, not proof of human attention.

A flow diagram showing 2,000 attempted email messages splitting into 1,900 provider-accepted deliveries and 100 delivery failures; an eye with a question mark shows that acceptance does not prove viewing.

Provider acceptance is a delivery signal, not evidence that a person saw the email.

Protect the meaning of each count

Messages and recipients are different units

The remaining fields use different counting units:

  • Unique clicking recipients counts people or recipient records, once each, even if someone clicks more than once.
  • Attributed purchasing recipients also counts recipients once each. In this report, each is a delivered recipient with at least one purchase following a tracked campaign click within the reporting window.
  • Unsubscribes counts unique recipients in this one-time campaign.

A label such as “total clicks” would need another definition: does it include repeated clicks by the same recipient? Always establish whether activity is unique or total before comparing reports.

Website sessions remain session counts

The separate website report contains:

  • 190 campaign-attributed sessions
  • 38 sessions with a purchase within the campaign reporting window

These are session counts, not recipient counts. The number 38 happens to appear in both reports, but that does not make a purchasing recipient interchangeable with a session containing a purchase.

Keep the unit beside any observation you summarize. Also keep the reporting window beside it; otherwise two counts may cover different periods.

Tip

Comparison check

Before comparing two report values, ask: “Do they use the same unit, definition, and reporting window?” A matching number is not enough.

Practice: match units and missing context

Check the report before comparing

Match each report item or comparison claim to its counting unit or missing requirement.

Tap an item and then its matching pair.

Step 2 of 8

Calculate Click, Bounce, and Unsubscribe Rates

Calculate the Starter Kit campaign’s click, bounce, and unsubscribe rates while preserving the denominator and reporting window behind each percentage.

Set up each rate correctly

Choose the population first

Use the report definition before you calculate. Every rate follows the same structure:

Rate = numerator / denominator × 100%

For Northstar Stationery’s Starter Kit campaign, each recipient can receive at most one message. Therefore, Unique clicking recipients and Unsubscribes are compared with Delivered messages. Bounces are compared with Attempted sends, because a bounce is a delivery outcome of an attempted send.

Keep the reporting window with every result: First 7 complete days after sending. A rate describes observed activity under this definition; it does not, by itself, judge whether the campaign succeeded.

Match the outcome to its denominator

Diagram showing bounces branching from attempted sends, while clicks and unsubscribes branch from delivered messages.

Bounces use Attempted sends as their denominator. Unique clicks and Unsubscribes use Delivered messages in this one-message-per-recipient campaign.

Calculate from the underlying definition

Worked rates

The denominator answers “out of which population?” Write its label beside the percentage so that another reader can interpret the rate without guessing.

Campaign rate calculations

All calculations use the Starter Kit campaign window: First 7 complete days after sending.

Each percentage retains its denominator and reporting window.
RateNumeratorDenominatorCalculationResult
Unique click rate152 Unique clicking recipients1,900 Delivered messagesRate = 152 / 1,900 × 100%8.00%
Bounce rate100 Bounces2,000 Attempted sendsRate = 100 / 2,000 × 100%5.00%
Unsubscribe rate19 Unsubscribes1,900 Delivered messagesRate = 19 / 1,900 × 100%1.00%

Do not rely on the metric name alone

Check the definition behind the label

A provider dashboard displays a metric named “click rate.” Which underlying calculation matches the Starter Kit campaign definition?

Practice: produce a usable rate summary

Calculate and label the three rates

Complete this campaign reporting template. Enter each denominator label and calculated percentage using two decimal places.

markdown
Reporting window: First 7 complete days after sending

Input counts
Unique clicking recipients: 152
Bounces: 100
Unsubscribes: 19
Delivered messages: 1,900
Attempted sends: 2,000

Unique click rate — denominator: ; Rate = 152 / 1,900 × 100% = 
Bounce rate — denominator: ; Rate = 100 / 2,000 × 100% = 
Unsubscribe rate — denominator: ; Rate = 19 / 1,900 × 100% = 

Step 3 of 8

Choose a Meaningful Conversion Rate

Calculate recipient-based and session-based conversion rates for Northstar Stationery’s Starter Kit campaign while preserving the correct unit, denominator, attribution context, and reporting window.

Identify the unit before using a rate

Two reports, different units

Northstar Stationery’s Starter Kit campaign uses two reports for the same window: First 7 complete days after sending.

The email report contains 1,900 Delivered messages—a message count—and 38 Attributed purchasing recipients—a recipient count. In this one-message-per-recipient campaign, each recipient received at most one message. That design makes Delivered messages the prescribed denominator for the recipient-based email conversion rate, but it does not change Delivered messages into a recipient count.

The separate website report contains 190 Campaign-attributed sessions and 38 Sessions with a purchase. Both are session counts. A matching number, such as 38, does not make recipients and sessions interchangeable.

Recipients and sessions are counted differently

A split diagram shows an email-to-recipient path on one side and a website-session path on the other, each ending in purchase-related icons.

The campaign window can be the same while the reports count different units.

Calculate without mixing report types

Keep the report definition attached

The numerator and denominator define the question.

For the email report, a purchasing recipient is counted once when that delivered recipient has at least one purchase following a tracked campaign click within the reporting window. The numerator is therefore a recipient count. The denominator remains Delivered messages, a message count, because this one-time campaign sends at most one message to each recipient.

For the website report, both the numerator and denominator are session counts. A purchasing recipient is not necessarily a conversion event: one person could create more than one purchase event in a different report design. Confirm whether each report counts recipients, sessions, or events before comparing results.

Two valid conversion rates

Both calculations use the Starter Kit campaign reporting window: First 7 complete days after sending.

Each percentage needs its denominator, counting unit, attribution definition, and reporting window.
ReportCalculationQuestion answered
Email recipient reportRate = 38 / 1,900 × 100% = 2.00%What share of Delivered messages corresponds to a unique attributed purchasing recipient in this one-message-per-recipient campaign?
Website session reportRate = 38 / 190 × 100% = 20.00%What share of Campaign-attributed sessions included a purchase?

Example

Interpret attribution cautiously

Do not substitute the 20.00% session-based rate for the 2.00% recipient-based email rate. The denominators answer different questions.

Attribution rules determine which purchases appear in each report. Consent limitations, cross-device journeys, attribution differences, and reporting delays can make email and website totals differ. Even when a report attributes a purchase to the campaign, that observed attribution is not proof that the campaign created an incremental purchase.

Practice: preserve each denominator

Calculate each report's conversion rate

Complete the denominator label and percentage for each report. Use two decimal places for the calculated rates.

markdown
Reporting window: First 7 complete days after sending

Email report
Attributed purchasing recipients: 38
Delivered messages: 1,900
Recipient-based email conversion rate — denominator: 
Rate = 38 / 1,900 × 100% = 

Website report
Sessions with a purchase: 38
Campaign-attributed sessions: 190
Session-based conversion rate — denominator: 
Rate = 38 / 190 × 100% = 

Apply both rates in a defensible review

Assess a conversion-rate comparison

Write a concise review note for the Starter Kit campaign.

  • Email report, First 7 complete days after sending: 1,900 Delivered messages and 38 Attributed purchasing recipients.
  • Website report, First 7 complete days after sending: 190 Campaign-attributed sessions and 38 Sessions with a purchase.

Include both conversion-rate calculations with their denominator labels and reporting window. State the question each rate answers. Then reject the claim that the website session rate can replace the email rate, and explain why a website rate measured over 14 days cannot be directly compared with these 7-day results.

Write at least 260 characters (0/260).

Step 4 of 8

Treat Opens and Clicks as Imperfect Signals

Interpret Reported opens and Reported clicks cautiously, then select corroborating evidence without claiming certainty about deliberate human engagement.

Interpret reported opens cautiously

What an open can—and cannot—show

A recorded open is not proof of reading

An email platform usually records an open when a tracking image is requested. That request can happen without deliberate reading: privacy protections may preload or proxy images, and automated systems may request them.

The signal can also miss real reading. If images are blocked or tracking is otherwise limited, a person may read the email without producing a recorded open.

Treat Reported opens as an imperfect observation in both directions—not a count of people who definitely read the message.

Why open tracking can be distorted

Diagram showing a reported open caused by an automated image request and genuine email reading that is not reported because the tracking image is blocked.

Both false positives and missed reads make Reported opens an incomplete engagement signal.

Warning

Decision safeguard

Do not use Reported opens as the only success measure or as the only condition for an automation branch. A contact who did not generate an open may still have read the message; a reported open may not reflect a person’s attention.

Corroborate rather than assume

Investigate click patterns; do not overclaim

A reported click may also be automated

Security scanners and other automated systems can activate tracked links to inspect them. An unusually fast burst of activity or activation across many links is a reason to investigate, not proof that automation occurred.

For the Starter Kit campaign, the provider reports 152 Unique clicking recipients during the First 7 complete days after sending. This is a recipient count: each recipient is counted once, even if they activate more than one link. It is still a Reported clicks signal and does not prove deliberate interest.

Look for corroborating evidence while keeping each report’s unit and reporting window visible.

Corroborating reported click activity

Use multiple signals to investigate Reported clicks from the Starter Kit campaign. Each signal adds context, but none establishes certainty by itself.

The website figures are session counts for the First 7 complete days after sending, not recipient counts.
SignalWhat it can corroborateWhat it cannot prove
Provider filtering informationPossible scanner or filter activityThat every remaining reported click was human
190 campaign-attributed sessionsSome downstream visit activityThat sessions equal Unique clicking recipients
38 sessions with a purchaseA downstream purchase outcomeThat every Reported click reflected interest
Click timing and link spreadA pattern worth investigatingAutomation or intent with certainty

Practice: avoid a certainty claim

Interpret the pattern

True or false: If many tracked links are activated seconds after delivery, you can conclude that recipients deliberately reviewed all of them.

Apply the signal limitation

Make a cautious evidence note

For the Starter Kit campaign in the First 7 complete days after sending, the provider reports 152 Unique clicking recipients. This is a recipient count, counted once per recipient, and is an imperfect Reported clicks signal. A burst of tracked-link activity occurred soon after sending across multiple links.

Write a short review note that:

  1. identifies one conclusion that would be unsafe from this evidence alone;
  2. names one appropriate corroborating check; and
  3. states why that check still does not prove deliberate human engagement.

Write at least 120 characters (0/120).

Step 5 of 8

Turn Delivery Failures and Opt-Outs into Actions

Use bounce and unsubscribe evidence to select proportionate list-health and eligibility actions without bypassing suppression or overstating causes.

Classify the signal before acting

Start with the delivery evidence

For the Northstar Stationery Starter Kit campaign, the reporting window is First 7 complete days after sending. The report shows 100 Bounces: 60 classified as hard bounces and 40 as soft bounces. It also shows 19 Unsubscribes.

Treat these labels as operational signals, not a diagnosis of every address. A hard bounce generally indicates a permanent delivery failure. A soft bounce generally indicates a temporary failure. The provider’s current classification and handling rules remain the source of truth.

Different signals need different handling

A diagram separates permanent delivery failures, temporary retryable failures, and protected opt-outs.

Hard bounces, soft bounces, and unsubscribes are different signals. Their labels and handling should not be treated as interchangeable.

Follow suppression and retry rules

Proportionate delivery actions

Use the provider’s classification, retry behavior, and suppression rules before deciding what to do with an address.

A bounce is a reason to follow up appropriately—not a reason to keep resending without limits.
SignalWhat it generally indicatesAppropriate action
Hard bounceLikely permanent delivery failureHonor suppression; review the provider failure reason and look for address-quality or signup-source patterns.
One soft bounceA temporary delivery problem may be possibleAllow provider-managed retries and monitor the address under the provider’s policy.
Repeated soft failuresA continuing issue needs reviewFollow provider policy; investigate failure reasons, source quality, and address accuracy rather than resending indefinitely.

Warning

Do not bypass suppression

Do not manually reactivate a hard-bounced address or an unsubscribed address to make reach look better. Marketing suppression protects people and list quality. If an address can legitimately receive a message again, that decision must follow the provider process and the person’s valid, current permissions—not a reporting target.

Use opt-outs as a review signal

Investigate opt-outs without inventing a cause

Unsubscribes can point to a mismatch worth investigating: signup expectations, segment relevance, lifecycle timing, or overlapping marketing sends. They do not prove which one caused an opt-out.

Examine a rate change alongside its underlying count, and compare only reasonably comparable segments and send conditions. There is no universal “safe” bounce or unsubscribe threshold that replaces provider guidance and audience context.

Example

Keep evidence separate from explanation

Evidence: 19 Unsubscribes were recorded for the Starter Kit campaign during the First 7 complete days after sending.

Reasonable checks: Were recipients expecting desk-organizer promotions? Was the segment relevant? Did a recent lifecycle message or another marketing send overlap?

Unsupported conclusion: “The desk organizer offer caused every unsubscribe.” The count supports investigation, not certainty about the cause.

Apply the handling rules

Choose the proportionate action

Match each situation to the most appropriate response.

Tap an item and then its matching pair.

Step 6 of 8

Compare Workflow Cohorts at the Same Stage

Assess Welcome workflow cohorts by entry timing and complete observation windows before judging performance.

Check the observation window

Start with entry timing

A workflow entry cohort groups contacts by when they entered the same workflow. This lets you ask a fair timing question: have both cohorts had the same opportunity to receive scheduled messages and produce the defined outcome?

For Northstar Stationery’s Welcome workflow, messages are scheduled on Day 0, Day 2, and Day 5. The complete outcome window ends after Day 12, leaving 7 days after the final scheduled message for purchases to occur.

A cohort observed only through Day 6 is not yet ready for a final workflow-outcome comparison.

The same workflow, different observation time

Two Welcome workflow timelines: one cohort reaches the Day 12 outcome cutoff, while another stops at Day 6 before the full window is complete.

Compare cohorts only after each has reached the same Day 12 outcome cutoff.

Calculate without forcing a comparison

Keep the denominator and unit clear

Use eligible entrants as the fixed denominator for an entrant-level workflow purchase outcome. Retain entrants in their original cohort denominator even if they exit early after a valid purchase or eligibility rule.

Do not treat a contact still waiting for a scheduled message as a failure. Also keep this outcome separate from per-message delivery or reported-click rates: those measure messages or recipients at a message level, while this rate measures workflow entrants.

Workflow cohort snapshot

Both cohorts entered the same Welcome workflow under the same stated outcome definition: a purchasing entrant by Day 12. Only the Earlier cohort has complete observation.

Entrant-level outcomes; eligible entrants remain the denominator.
CohortEligible entrantsPurchase outcomeObservation status
Earlier cohort20020 purchasing entrants by Day 12Complete through Day 12
Recent cohort2008 purchasing entrants so farObservation incomplete: through Day 6 only

Example

Calculate, then label the limitation

For the completed cohort:

Workflow purchase outcome rate = purchasing entrants by Day 12 / eligible entrants × 100%

For the recent cohort, you may calculate a descriptive result so far, but label both its denominator and its incomplete observation status. It is not a Day 12 result and does not establish that cohort as lower-performing.

Decide whether comparison is fair

Check comparison readiness

Which conclusion is supported by the Welcome workflow snapshot?

Apply the timeline rule

Write a comparison-safe conclusion

Use the Welcome workflow snapshot below to write a short cohort-status note.

  • Earlier cohort: 200 eligible entrants; 20 purchasing entrants by Day 12; observed through Day 12.
  • Recent cohort: 200 eligible entrants; 8 purchasing entrants so far; observed only through Day 6.

Include: (1) the status of each cohort, (2) each calculated rate with its numerator, denominator, and time label, and (3) whether you would name a winner now and why.

Write at least 180 characters (0/180).

Step 7 of 8

Choose One Evidence-Linked Improvement

Use connected evidence to select one proportionate improvement or test while protecting conversion measurement and list health.

Connect the evidence

Diagnose a connected system

A useful diagnosis follows the path from sending to outcome instead of optimizing one isolated number:

  1. Delivery: can messages be accepted?
  2. Reported engagement: did tracking record clicks?
  3. Destination outcome: did campaign-attributed sessions and purchases occur?
  4. List health: are bounces or unsubscribes signaling an eligibility or expectation issue?

For the Northstar Stationery Starter Kit campaign, keep the units and window visible: 2,000 Attempted sends, 1,900 Delivered messages, 152 Reported unique clicking recipients, 38 Attributed purchasing recipients, 100 Bounces, and 19 Unsubscribes during the First 7 complete days after sending.

These observations do not, by themselves, prove a cause or tell you whether a result is “good.”

From delivery to outcome

A diagram connects sending, delivery, reported clicks, website activity, and purchases, with separate bounce and unsubscribe branches. A magnifying glass highlights uncertainty around reported clicks.

Read delivery, Reported clicks, destination outcomes, Bounces, and Unsubscribes together. Reported clicks are a signal to corroborate, not proof of deliberate interest.

Choose a proportionate action

Turn patterns into cautious next steps

Use a pattern only after checking that the comparison is fair: the same outcome definition, a comparable audience and eligibility rules, and a completed observation window. A historical difference can support a hypothesis; it cannot prove that a proposed change caused the difference.

For example, a recent Welcome workflow cohort observed only through Day 6 is still marked Observation incomplete. Do not compare its 8 purchasing entrants so far with an earlier cohort’s Day 12 result as if both had the same opportunity to convert.

Pattern-to-action guide

Examples of proportionate diagnosis. “Weak” means lower than a comparable, fully observed baseline—not lower than a universal benchmark.

Check the evidence and its limitations before choosing one change.
Observed patternPossible explanationProportionate next step
Stable delivery; lower Reported clicks in a comparable groupRelevance, call-to-action, audience, or tracking differenceCheck comparability; then test one relevance or CTA change
Reported clicks without downstream corroborationAutomated activity or measurement issueCheck tracking and provider filtering before changing the destination
Corroborated visits; lower purchase outcome in a comparable groupMessage-to-destination mismatch or purchase-path frictionInspect the path; test one aligned message or destination change
Hard bounces or Unsubscribes increaseAddress quality, expectations, timing, or overlap issueFollow suppression rules; review failure reasons or eligibility before more sends
Recent workflow cohort only through Day 6Observation incompleteWait for the Day 12 outcome window before judging it

Practice proportional diagnosis

Match evidence to the next move

Match each evidence pattern to the most proportionate next move.

Tap an item and then its matching pair.

Apply the diagnosis

Draft one evidence-linked recommendation

Write a concise recommendation for the Northstar Stationery Starter Kit campaign.

Use these observations from the First 7 complete days after sending: 2,000 Attempted sends; 1,900 Delivered messages; 100 Bounces, including 60 hard bounces; 19 Unsubscribes; 152 Reported unique clicking recipients; 38 Attributed purchasing recipients; and 190 campaign-attributed sessions with 38 sessions with a purchase.

Include:

  • observations stated as facts, with units where relevant;
  • one limitation or uncertainty;
  • one proportionate change or focused test—not a list of changes;
  • a conversion outcome to monitor;
  • an unsubscribe or delivery guardrail; and
  • a comparable review window.

Do not claim that a metric proves a cause or use a universal “good” benchmark.

Write at least 250 characters (0/250).

Step 8 of 8

Complete an Evidence-Based Campaign Review

Synthesize campaign and workflow evidence into a concise review with one cautious, measurable next action.

1. Structure the review

Build a defensible brief

A useful campaign review is a decision brief, not a verdict. Separate what the reports show from what they cannot prove, then name one next action and the safeguards for reviewing it.

Use these seven labels in order:

  1. Evidence — observed counts or a calculated rate.
  2. Limitation — a measurement or interpretation uncertainty.
  3. Cohort status — whether workflow results are complete enough to compare.
  4. One change — one feasible, evidence-linked action or test.
  5. Conversion outcome — the outcome metric to monitor.
  6. Guardrail — an unsubscribe or delivery measure that must not worsen.
  7. Review window — when the outcome can fairly be assessed.

A proposed explanation is not evidence. A change can be reasonable to test, but the historical report does not prove that the change will cause an improvement.

From report to decision

A visual flow connecting campaign evidence, limitations, workflow timing, one change, conversion measurement, a list-health safeguard, and a review period.

Use the brief to connect observed evidence to one cautious next action instead of optimizing an isolated metric.

2. Read the evidence and timing

Reports to synthesize

Use these fictional reports for your Northstar Stationery review. The Starter Kit campaign promoted a desk organizer. Its reporting window is First 7 complete days after sending.

The website figures are session counts, not recipient counts. Reported clicks can include automated activity, and attributed purchases do not prove incremental impact.

For the Welcome workflow, messages are scheduled on Day 0, Day 2, and Day 5. A complete outcome window ends after Day 12, allowing 7 days after the last scheduled message. Eligible entrants remain in their original cohort denominator even if they exit early.

Campaign report

Starter Kit campaign — First 7 complete days after sending

Each Attempted send has one delivery outcome in this simplified snapshot.
MetricCounting unitCountDetail
Attempted sendsMessages2,000One delivery outcome each
Delivered messagesMessages1,900Provider acceptance, not proof of reading
Unique clicking recipientsRecipients152Reported clicks
Attributed purchasing recipientsRecipients38Counted once per qualifying recipient
BouncesMessages10060 hard bounces; 40 soft bounces
UnsubscribesUnique recipients19One-time campaign
Campaign-attributed sessionsSessions190Website report
Sessions with a purchaseSessions38Website report

Workflow report

Welcome workflow outcome snapshot

Do not treat an incomplete observation period as a completed workflow result.
Entry cohortEligible entrantsObserved throughPurchasing entrantsCohort status
Earlier cohort200Day 1220Complete
Recent cohort200Day 68Observation incomplete

3. Apply the evidence

Final application: write the review

Write a concise Northstar Stationery review brief using the reports on the previous screen.

Use every required label exactly once: Evidence, Limitation, Cohort status, One change, Conversion outcome, Guardrail, and Review window.

In Evidence, calculate at least one campaign rate using Rate = numerator / denominator × 100%, name its denominator, and state the campaign reporting window. In Limitation, identify an engagement or attribution uncertainty. In Cohort status, address whether the Recent Welcome workflow cohort can be judged against the Earlier cohort. Propose only one evidence-linked change. Include a conversion outcome, an unsubscribe or delivery guardrail, and a fair review window.

Write at least 500 characters (0/500).

4. Finish

Summary

Review checklist

Your review is decision-ready when it keeps these distinctions visible: report evidence versus hypothesis, recipients versus sessions, reported engagement versus deliberate engagement, and completed workflow observation versus pending time in sequence.

  • State the numerator, denominator, and reporting window for every rate you use.
  • Treat Reported clicks and attributed purchases as useful but imperfect signals.
  • Do not judge the Recent Welcome cohort as a completed Day 12 outcome.
  • Choose one change and protect delivery or unsubscribe performance while reviewing it.

Step complete

Congratulations! You completed: Use Email Metrics to Improve Campaigns

You completed an evidence-based campaign review that links one next action to measurement and list-health safeguards.

Download the App now to have access to + 5000 free courses, exercises, certificates and lots of content without paying anything!

  • 100% free online courses from start to finish

    Thousands of online courses in video, ebooks and audiobooks.

  • More than 60 thousand free exercises

    To test your knowledge during online courses

  • Valid free Digital Certificate with QR Code

    Generated directly from your cell phone's photo gallery and sent to your email

Cursa app on the ebook screen, the video course screen and the course exercises screen, plus the course completion certificate