Solving the issue or symptom?

When developing software I am often tasked with solving a problem however many times I am given resources to solve the symptoms of the problem.

An example is a blue CTA button with the text "Read More".

AB tests showed that if the colour of the button was changed, there would be a 5% increase in clicks however this is just solving a symptom, not the problem.

Having consulted with several users the real problem was the text of the button. Changing the text to something more descriptive such as "Learn more about us" increased clicks by 13%.

But why was this is true problem? The demographics of the website consisted of people who were older and by providing more information in the button it told them exactly what they were getting into. Instead of the vague message "Read More", they now know that they are going to learn more about the company. Changing the colour of the button highlighted it by creating contrast, changing the text gave users the necessary context that they needed to click the button.

Things like these are why I love working with UX designers and why I spend so much time around UX teams.