A CTA's click rate measures visibility. Whether that click converts depends on whether the copy promised what the destination delivers.
TL;DR
- A CTA fails for one of two reasons: the user doesn't see it (visibility failure) or sees it and reads it as something else (intent failure).
- Visibility failures are loud — low CTR, low scroll engagement on the CTA region. Most teams catch these.
- Intent failures are quiet — _high_ CTR, _low_ click-to-conversion rate, bounce-back from the destination. Most teams miss these.
- The diagnostic is per-source-page click-to-conversion rate. If clicks from pages where audience intent doesn't match the destination dominate, the CTA copy is pulling the wrong audience.
- Fixing it: change the verb to predict the destination, or suppress the CTA on source pages where the audience can't be the intended one.
The two failure modes, mapped
Most CTA test reports tell you whether you're in the top half (visible) or the bottom half (not visible). They rarely tell you whether the click came with the right intent.
| | Copy clear about destination | Copy unclear or mismatched |
| ----------------------- | ---------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------- |
| Visible CTA | ✅ Real win. Healthy CTR + healthy click-to-conversion. Conversion grows. | ⚠️ The intent trap. Healthy CTR but low click-to-conversion. Users bounce from the destination. |
| Hidden / buried CTA | ⚠️ Visibility problem only. Low CTR but healthy click-to-conversion when clicked. Fix with prominence. | ❌ Compound failure. Low CTR AND low click-to-conversion when clicked. Fix both. |
The standard CRO playbook fixes the bottom half — visibility. Make the CTA bigger, brighter, sticky, above the fold. It does almost nothing for the right column — copy and destination misalignment. Most "we shipped the CTA test" failures live in the top-right cell. The CTA was visible. Users clicked. The clicks didn't convert because the copy promised one thing and the destination delivered another.
How to tell which cell you're in
Pull two metrics for the CTA in question:
| Metric | What it tells you |
| ---------------------------- | -------------------------------------------- |
| Click-through rate (CTR) | Visibility / placement / visual treatment |
| Click-to-conversion rate | Whether the click came with the right intent |
| If CTR is... | And click-to-conv is... | Failure mode | Fix |
| ------------------- | ---------------------------- | ----------------------------------- | ----------------------------------------------- |
| Healthy (>baseline) | Healthy (~baseline or above) | None — real win | Ship |
| Healthy | Low (<half of baseline) | Intent failure (top-right cell) | Change copy or suppress on wrong-intent pages |
| Low | Healthy | Visibility only | Make CTA more prominent, sticky, above-fold |
| Low | Low | Compound failure | Diagnose visibility first, then re-examine copy |
The intent failure is the cell most teams skip. It looks healthy at first glance because the click numbers are up. The diagnostic that exposes it is per-source-page breakdown.
The "Sign Up" trap, in numbers
A sitewide nav button test ran in this brand's portfolio earlier this year. The button said "Sign Up" and routed (via a modal) to plan selection. The aggregate looked positive. The per-source-page breakdown told a different story.
| Source page | Clicks | Click-to-page-entry rate | Audience intent |
| --------------------- | ---------- | ------------------------ | ----------------- |
| Ways to Pay (support) | 2,420 | ~5% | log in / pay bill |
| Customer Homepage | 1,080 | ~5% | manage account |
| Account dashboard | 870 | ~5% | account access |
| Prospect Homepage | 1,220 | ~22% | shop for plan |
| Total | ~7,000 | ~6% aggregate | mixed |
47% of clicks came from support and account pages. On those pages the click-to-page-entry rate was about a quarter of the rate on prospect pages. The CTA was working perfectly on the prospect homepage. It was actively damaging engagement on customer-support pages — because users on those pages were reading "Sign Up" as "log in to my account," clicking, and bouncing when the destination was a plan-selection page.
The CTA was visible. The clicks happened. The conversions did not — because the click was placed by a user looking for something the destination doesn't provide.
Three rules for CTA copy that binds intent to destination
| Rule | Bad | Good |
| ------------------------------------------------------------- | ----------------------------------------------- | ------------------------------------------- |
| The verb predicts the destination | "Continue", "Submit", "Next" | "View Plans", "Get Quote", "Apply Now" |
| The noun sets a specific expectation | "View" | "View Plans" or "View My Quote" |
| Avoid phrases with strong existing associations elsewhere | "Sign Up" (= account creation in most contexts) | "Shop Plans", "Browse Plans", "Get Started" |
If the click promises shopping, the destination must show shopping. If the click promises account access, the destination must show account access. Mismatched verbs are the most common copy-intent failure.
When you cannot rewrite the copy
Sometimes the CTA copy can't change. Brand approval, global navigation owned by another team, six-week deploy timeline. The cheaper fallback is per-page suppression: hide the CTA on source pages where the audience cannot be the intended one.
In the sitewide example above, the cleanest fix was suppressing "Sign Up" on customer-support and account pages — keeping it only on prospect-side surfaces where shopping intent existed. Per-page render conditions are simpler than copy changes and don't require global brand alignment. They also remove the worst sources of click-to-conversion drag without requiring a copy test.
What user research keeps saying
Across qualitative studies in the deregulated retail energy vertical and beyond, the same themes show up:
- Users want to know what happens after the click. Generic copy ("Next", "Continue", "Submit") is unsatisfying because it doesn't preview the destination.
- Action-oriented copy beats generic copy. Across competitor benchmarking — Octopus, Gexa, TXU, Think Energy — the standard CTA copy is "View Plans" or "See Offers". Not "Next". Not "Sign Up". The pattern has been validated industry-wide because it works.
- Benefit-led copy outperforms feature-led copy. "Save up to 20%" beats "View Plans". "See Apartment-Friendly Plans" beats "View All Plans" for users who self-identify as renters. The benefit framing pre-qualifies the click.
- Users notice when the CTA promise doesn't match the destination. Session-replay evidence is consistent: users back-button out within seconds, often returning to the same page they came from.
This aligns with Jakob's Law — users spend most of their time on other sites, so they expect your site's CTAs to behave like CTAs on the sites they already use. The vocabulary is industry-trained. Fight it and you lose. Match it and the click-to-conversion rate climbs without any other change.
The behavioral mechanism
A click is a curiosity event. A conversion is a commitment event. The two are correlated when the path between them is well-designed and the copy sets accurate expectations. They decouple when:
- The verb / noun in the CTA implies a different destination than what the user lands on.
- The user's intent on the source page doesn't match the destination's audience.
- The destination has new friction (modal, form field, redirect) that wasn't implied by the CTA copy.
Each of these creates the same signature in the data: healthy clicks, broken conversion. The diagnostic per-placement breakdown catches all three.
Bottom line
Most CTA tests measure visibility. A CTA's value depends on visibility _and_ intent match — and intent failures show up as healthy click rates with broken conversion rates. Pull per-source-page click-to-conversion rate on every CTA test. If a high-volume source page is converting clicks at less than half the rate of a known-good source page, you have an intent failure. Fix it with copy or with per-page suppression — whichever is cheaper to ship.
Visibility is the easy half. Copy intent is the half that determines whether a CTA test ships a real win or a cosmetic one.