Metrics·Glossary term

Click Rate

Click Rate A/B testing Reference guide

Click Rate is a concept used in metrics, kpis & business outcomes.

Quick definition: Click rate is the share of a defined delivered, displayed, or eligible audience that clicks at least one tracked link or interface element.

What is click rate?

Click rate describes audience response at an interaction point. It is most often used in email and lifecycle messaging, where it usually means unique recipients who click at least one link divided by delivered emails. Product teams also use it for the share of users who click an in-app element. The label is ambiguous, so a metric definition should say exactly what counts as a click and what population supplies the denominator.

Click rate is often confused with click-through rate (CTR). In some reporting systems they are synonyms. In others, CTR means clicks divided by ad or page impressions, while click rate means unique clickers divided by delivered emails or recipients. Click-to-open rate (CTOR) is different again: unique clickers divided by reported opens. Since email opens are affected by image blocking and privacy features, CTOR is a diagnostic rather than a reliable substitute for click rate.

The outcome is usually a micro-conversion, not proof of value. A recipient may click from curiosity, a scanner may follow a link, or a user may discover the destination does not satisfy the message. Measure downstream registrations, purchases, completed tasks, unsubscribes, spam complaints, and retention when they relate to the campaign objective.

Click-rate formula and denominators

A common email formula is:

unique click rate = unique delivered recipients who click at least once / delivered recipients × 100

“Delivered” generally excludes hard bounces and rejects, but should include recipients who do not open or engage. This makes it useful for assessing a campaign’s overall ability to prompt action. A total-click rate uses all clicks as its numerator and can exceed 100% if recipients click multiple links. Use unique clickers when the question is how many people responded; use total clicks when repeated interactions or link usage are the subject.

MeasureNumeratorDenominatorUse
Unique click rateRecipients clicking at least onceDelivered recipientsCampaign reach and response
Total click rateAll tracked clicksDelivered recipientsLink interaction volume
Click-to-open rateUnique clickersReported opensCreative diagnostic, with caveats
In-app click rateUnique users clickingUsers with eligible rendered elementInterface response

Use the same identity level for both parts. If a recipient has several devices, link clicks should resolve to the recipient where feasible. Do not include clicks made after an unsubscribe or conversion if the campaign definition says the opportunity ended earlier. State the observation period: an email click rate measured after 24 hours is not comparable with one measured after seven days.

Click rate in A/B testing

Click rate is appropriate for testing subject lines, email content, link placement, notification copy, and navigation where the goal is to help people reach relevant content. Randomize before delivery, preserve the assignment, and analyze all eligible delivered recipients rather than only people who opened. Conditioning on opens can bias a comparison because the treatment may affect opening behavior and because open tracking is incomplete.

For an email test, hold audience criteria, send time, suppression rules, and tracking links constant except for the planned change. A subject line that creates urgency can increase opens and clicks while increasing unsubscribes or disappointing readers. A better decision includes completed destination actions, complaint rate, unsubscribe rate, and any relevant revenue or support metrics. See how to select primary and guardrail metrics.

Predefine the click event and whether a click on any link or on a focal CTA counts. Multiple links can create a misleading aggregate: a treatment may raise clicks to “learn more” while lowering clicks to the intended action. Inspect link-level results as secondary diagnostics, but do not search dozens of links and announce a winner without accounting for multiple comparisons. Our multiple-comparisons guide explains why.

Worked click-rate calculation

A subscription service tests a renewal reminder. The population is eligible customers whose email was accepted by the provider; the primary click is a unique recipient clicking the account-renewal link within 72 hours. Control is delivered to 18,000 recipients, with 1,080 unique clickers. Treatment is delivered to 17,800 recipients, with 1,246 unique clickers.

control click rate = 1,080 / 18,000 = 6.00%
treatment click rate = 1,246 / 17,800 = 7.00%
absolute difference = +1.00 percentage point
relative lift = 16.7%

The treatment is more effective at generating visits, but renewal completion is 2.10% in both arms. Investigation shows the message made the link easier to find, yet the renewal page has a payment-error problem. The experiment identifies a helpful messaging improvement and a separate funnel constraint; it would be wrong to describe the click lift as a renewal lift.

Data-quality caveats

Security scanners, corporate gateways, and privacy tools can automatically follow links, inflating clicks. Filter known bot signatures where reliable, compare unusually fast clicks, and document exclusions. Do not over-filter selectively after one variant underperforms. Shortened URLs and redirect services need campaign and variant parameters that survive redirects without exposing personal data.

Delivery logs and click logs can disagree because of provider delays, user identity changes, opt-outs, and link-tracking failures. Reconcile counts by send batch and variant. Keep a stable recipient identifier, message identifier, link identifier, and timestamp. In product interfaces, distinguish a rendered, visible element from an element merely present in DOM code; lazy loading and feature flags can alter real opportunity.

Segment comparisons require care. Device, country, email client, and acquisition source can all change click behavior. Predeclare important segments, report their sample sizes, and treat unplanned patterns as hypotheses for future validation rather than conclusions.

Common click-rate mistakes

  • Using “click rate” without a denominator: recipients, opens, and impressions answer different questions.
  • Counting automated scans as human intent: validate abnormal timing and user agents.
  • Comparing different maturation windows: later clicks are missing from newer sends.
  • Optimizing any-link clicks: the intended action may decline while curiosity clicks rise.
  • Analyzing openers only: this removes a population affected by the treatment.
  • Ignoring unsubscribe and complaint effects: attention can be purchased with trust.

Frequently asked questions

Is click rate the same as CTR?

Sometimes vendors use the terms interchangeably. Define the numerator and denominator explicitly, especially for email where click rate commonly uses delivered recipients.

What is a good email click rate?

Benchmarks vary by audience, message purpose, deliverability, and link count. Compare comparable campaigns and downstream outcomes rather than relying on a universal target.

Why might total click rate exceed 100%?

It can when total clicks, rather than unique clickers, are divided by delivered recipients.

Should clicks from security scanners be removed?

Remove them only using a documented, consistently applied rule. Keep the raw data so the effect of filtering can be audited.

Can click rate be a primary experiment metric?

Yes when a relevant click is the immediate objective, but pair it with quality or completion measures that confirm the click created value.

Summary

Click rate measures the share of a defined audience that clicks, commonly unique clickers per delivered recipient. It must be distinguished from CTR and click-to-open rate through an explicit denominator. In A/B tests, measure all assigned eligible recipients, defend against automated traffic, and judge a click lift alongside the outcome reached after the click.

Sources