CTA comparison

CTA comparison lives in the Analyze area of the dashboard, under Behavior.

At a glance#

What it does#

CTA comparison ranks your calls to action by click-through rate — clicks divided by impressions. Both sides come from autocapture: click events and impression events, with no bespoke instrumentation to add.

Dividing by impressions rather than by page views is the whole point. A CTA with few clicks might be ineffective, or it might simply never be seen — and those are opposite problems with opposite fixes. Only a view-rate distinguishes them.

When to use it#

When conversion is lower than you expect and you need to know which element is responsible. Also whenever you have several CTAs competing on one page: this is the view that tells you they are competing, and which one is losing.

Workflow#

1. 1

   Sort by CTR, then check impressions

   A 40% CTR on 20 impressions is not a finding. Read both columns together.

2. 2

   Separate "not seen" from "not clicked"

   Low impressions means a placement problem — see scroll maps. High impressions with low CTR is a copy or design problem.

3. 3

   Look for cannibalisation

   Two CTAs on one page each getting mediocre CTR usually beat one clear CTA when combined into it.

4. 4

   Slice by segment before rewriting

   A CTA that works for returning users and fails for new ones does not need new copy; it needs to not be shown to the wrong audience.

Permissions and prerequisites#

Web apps only — CTAs and impressions are HTML concepts. Requires autocapture, including impression reporting, in the SDK.

Limits and edge cases#

• An impression means it entered the viewport, not that it was noticed. CTR is therefore a floor on how ineffective a CTA is, not a measure of attention. Pair with attention maps.

• Autocapture identifies elements structurally. A refactor that changes markup can make an element look new, splitting its history in two.

• A click is not a conversion. People click and abandon. Follow the CTA into a funnel before celebrating.

• Very small samples produce dramatic rates. Ignore rows with low impressions.

Troubleshooting#

A CTA I know exists is missing. Autocapture did not identify it, or it was never in a viewport during the window. Check that impression reporting is enabled.

CTR above 100%. Clicks recorded without a matching impression — usually an element clicked before its impression fired, or a programmatic click. Treat the row as unreliable.

Where the data comes from#

Click-through rate is autocaptured clicks over autocaptured impressions, so it needs impression reporting as well as click reporting.

From the SDK

• Events

Served by

• Ingestion

Related#

• Attention maps

• Forms

• Events

• Event Debugger

• Scroll maps

• Exit intent

• Analyze overview