A B2B website is more than the sum of its parts. But some of its smallest components can make a huge impact on user journeys: calls-to-action (CTAs). Their limited size also makes them harder, not easier, to write.
We’ve launched 150+ websites for B2B clients, so we’re well-versed in the art of CTA best practices. In this article we’ll share how we approach writing CTAs, backed by research from user experience (UX) organisations, and how to A/B test to optimise for your users.
UX research company Nielsen Norman Group found that users frequently use a ‘Spotted Pattern’ to scan a web page. They identify the elements they’re looking for or think are important first, treating things like links or buttons as standalone items without reading the surrounding text.
A CTA needs to be a clear, comprehensible instruction. If your CTA tells a user they can ‘Cultivate more organic traffic’ by clicking, but they need to fill in a form or survey before they can download an ebook, you’ve misled them. The same goes for very generic copy, such as ‘Learn more’ or ‘Get started’. It doesn’t tell the user what to expect next, and you might lose them to anxiety or frustration.
Instead, you want to write descriptive text that tells the user what they’ll get when they click. Examples include:
CTAs are ultimately instructions to the user. That’s why imperative verbs such as ‘Download’, ‘Read’, or ‘Start’ make for more effective CTAs. They align with user intent and help reduce their decision fatigue. Imperatives also shorten the cognitive distance between seeing the CTA and taking action.
You’re not trying to be pushy. You’re trying to be useful. Consider the difference:
The second option in each pair is direct, helpful, and easy to act on. That’s what makes it effective.
Asking users for professional or personal information, or payment details, means you might need more than just your CTA. Even if it’s well-written and well-placed, users might hesitate if they’re not sure about your intentions or what they’re getting in return.
This is when it helps to add trust signals in the form of messaging. These accompanying statements reassure the user with guarantees or clarity about the process (in addition to privacy notices, which are a legal requirement for holding people’s data.)
Trust signals might look like:
Subtle shifts in language can make a measurable difference. Unbounce found that using first-person phrasing (such as ‘Start my trial’ instead of ‘Start your trial’) can improve conversion rates by up to 90 percent. It personalises the interaction, creating a sense of ownership that nudges the user toward action.
This is a small detail, but a meaningful one. It’s a reminder that strong UX writing is as much about tone and perspective as it is about word choice.
Writing a good CTA is one thing. Knowing whether it works for your users is another. That’s where A/B testing comes in.
A/B testing splits users into two groups, then shows each group one of two versions of an asset and measures the difference in their responses. It’s a simple, but highly effective way to demonstrate the impact of changes to websites, emails, or social media, without owner or creator bias.
To get the most valuable, actionable insights from your CTA A/B tests, keep these factors in mind:
Effective CTAs are honest, specific, and helpful. They don’t rely on gimmicks or guesswork. Instead, they reflect understanding of the user’s intent and a commitment to earning every click.
We also use that principle at Articulate when we design Difference Engines®️ that fuel our clients’ growth. If you want CTAs that convert more users, driving leads and sales for your business, contact our team today.