The Ultimate Guide to Email Calls to Action
A campaign can have excellent copy, a well-segmented audience and correct authentication, and still fail because the reader did not know what to do next. The call to action is the smallest piece of text in the email and the one the whole send depends on.
One ask per email
The most common structural mistake is asking for several things. Book a demo, read the guide, follow us, and upgrade. Presented with four options of equal weight, readers take none, because deciding costs more than deferring.
Choose the single action that matters for this audience at this moment and make everything else supporting rather than competing. Secondary links are fine as plain text; they should not look like additional buttons.
Writing the label
A good call to action completes the sentence "I want to..." from the reader's point of view.
- Specific beats generic. "See the pricing" communicates more than "Learn more".
- First person works. "Start my free trial" reads as the reader's intention rather than an instruction.
- State the outcome. "Get the deliverability checklist" tells them what arrives.
- Avoid novelty. A clever label that requires interpretation loses to a plain one that does not.
Two to five words is usually right. Long labels wrap awkwardly on mobile, and the reader scans rather than reads.
Placement
Placement is more mechanical than most advice implies, because reading position is predictable.
- Above the fold for anyone who already knows what they want. Some readers will act immediately.
- After the main argument for the majority, who need the case made first.
- Near the end for the remainder, who read to the bottom before deciding.
An email with one action but three placements of the same link performs better than three different links, because the reader who is ready at the top and the reader who is ready at the bottom both get an obvious next step.
Design and contrast
A button must look like a button and be the most visually distinct element on the page. Failures are common and predictable:
- Low contrast against the background, particularly light grey on white.
- A button the same colour as the surrounding design, so it reads as decoration.
- Text baked into an image, which fails when images are blocked, which is the default in many clients.
- Tap targets too small for a phone, which is where most of your readers are.
Use real text on a high-contrast background with generous padding. Test on a phone, not a desktop preview.
What makes people hesitate
Even a well-written, well-placed call to action can fail because of what it implies.
- Unknown cost. "Contact sales" with no indication of price is a barrier; a range reduces it.
- Commitment ambiguity. "Start your trial" is clearer than "Get started", which could mean anything.
- Effort. A booking flow that requires a long form loses people who were ready to act.
- Trust. A first-time reader needs more reassurance before clicking than a returning customer does.
What to test, and what not to
Test one variable at a time and test on a segment large enough to be meaningful. In practice, the variables with the largest effect are:
- Where the primary action appears
- The label itself
- Whether there is one ask or several
Button colour is the most frequently tested and the least consequential, and it is where teams spend the most effort because it is the easiest to change. Judge tests on clicks and conversions, not opens, which are inflated by client-side image proxying and prefetching and will not respond to a call to action at all.
Common mistakes
- Multiple competing buttons. Splits attention and reduces total clicks.
- "Click here". Fails accessibility and communicates nothing when read as a list of links.
- One placement only. Misses readers who act at a different point in the email.
- Testing colour first. Low payoff relative to effort.
- Judging on opens. A metric that cannot respond to the change being tested.
Pricing
Campaign features including templates and AI-generated variants are on every paid plan:
| Plan | Price | Contacts | Emails / month |
|---|---|---|---|
| Free | $0 | 50 | 50 |
| Professional | $29/mo | 5,000 | 5,000 |
| Premium | $49/mo | 25,000 | 25,000 |
| Ultra | $99/mo | 55,000 | 55,000 |
Email clients that break calls to action
A call to action that works in your preview can fail in the client your reader uses. The failure patterns are predictable.
- Outlook on Windows renders with a different engine. Padding on a button often collapses, and the clickable area shrinks to the text itself.
- Dark mode inverts backgrounds. A button whose text colour is set but whose background is an image becomes unreadable, and a light button on a dark background can lose its border entirely.
- Images blocked by default. A call to action built as an image disappears, which is why the label must be real text.
- Clipped content. Gmail clips long messages, and a call to action in the last third of a very long email can end up behind a "view entire message" link.
Preview in at least one light and one dark client, and on a phone, before sending to a real audience.
Context and the ask
A call to action does not stand alone. The same button performs differently depending on what the reader was doing immediately before it.
An ask that follows an explanation of a problem the reader recognises will land. The same ask, placed after a paragraph of company history, will not, regardless of how it is worded. If a call to action underperforms, check the paragraph above it before rewriting the label, because the problem is more often the setup than the label.
The short version
One ask per email, labelled as the reader's intention rather than an instruction. Repeat the same action at three points in the email rather than presenting three different ones. Make the button visually unmistakable, real text and high contrast, and check it on a phone and in dark mode.
Write the paragraph before the ask as carefully as the ask itself. Test placement and labels before colour. Judge the result on clicks and conversions, because opens cannot measure a call to action.
Continue learning
Related Cresca resources
External references