Does changing your CTA button or re-writing your headline even matter?
In this piece I look at risk vs. reward to understand where we can make quick, safe optimizations
Having spent a lot of time reviewing A/B testing data over the last 12 months, here is what I am seeing.
Imagine all changes you make exist on a graph, y-axis is risk, and x-axis is impact.
A low-risk, low-reward test might be something like changing your color scheme, or changing a product image vs. a stock image.
A high-risk, high-reward test might be changing out your whole hero section for an LLM input box. Very likely to substantially and statistically significantly change your funnel. It could go up or down, but it will move.
Then you look at something like fixing your CTA button text.
A common example I use because there is so much data around it, is replacing “Learn More” with more specific next-steps verbiage (“View use case”, “See pricing for XYZ”, “Read how it works”).
Talking with teams that tested this, I saw that 78% of brands saw negligible impact in changing the button text, 22% saw a lift… but zero brands that changed out “Learn More” for more specific copy saw a decline in CTR.
So this is an extremely low-risk move, with low-to-medium upside (some brands like Glean saw a whopping 100% CTR increase by swapping out their “Learn More” CTAs).
Not to mention that the speed of a change like this is minutes, not hours, to implement.
So when it comes to small tests or experiments, I think the mental re-frame here is… What is the downside risk?
Some tests have very real downside risk, and it’s best to approach those carefully and with a data-backed experiment plan.
But don’t sleep on the low-risk, medium reward vector, there is some gold there.
Here are three other low-risk, solid upside ideas…
Have a comparison of your product vs. Claude Code (Framer does this and is ranking in LLMs)
Clarify your company stage segmentation (i.e. instead of saying Startups, Mid-sized, Enterprise, do what Wayflyer does and specify 10-100k, 100k-1m, 1m+)
Link your case studies to your customer logos (Clay led the charge here and now hundreds of brands do this)
At DoWhatWorks, our world-class research team works with many of the largest brands in SaaS, fintech, streaming, and eCommerce, to power their website optimization programs.
While our primary competitive intelligence offering is aimed at brands at $100m+ ARR that have established experimentation programs, we have an exciting opportunity for smaller brands.
We are currently building an agentic version of our database that will make searching for tests and insights easier and more streamlined. While this is in early beta, we are opening up slots for folks to hop on a call with our team, provide feedback, and in exchange, get some complementary analysis on your site.
Find a time with our team that works for you.