Case study: conversion optimisation

tl;dr version

40.7% increase in conversion rate

136.8% increase in revenue

Client happy.

About the client

An international school offering learning in a specialist subject, with face-to-face teaching and online courses.

What they wanted

Primary goal: Get more people signing up to online classes

Secondary goal: The owner had grown the business, so the copy needed to reflect it was now a bigger operation

This was to be done in 2 stages:

1. Copy

  • rewrite course copy to sell the courses more
  • review overall website copy from a conversion optimisation angle, using words to remove barriers to signing up

2. Conversion rate optimisation

  • identify and fix where the website was losing visitors
  • analyse the % of abandoned carts and optimise the checkout process
  • install Hotjar to analyse how visitors use the website, what they click on, where they go – and review this against the client’s objectives

The process

1. Diagnosing with data

A doctor can’t tell what’s wrong with a patient by looking at them (unless we’re talking broken bones or things like that). It’s the same with websites.

The website had 5,000–7,000 sessions per month, so this was perfect for gathering enough data to uncover any problems. Where people are leaving the site, any pages not performing well on mobile or particular browsers, any places where people are getting lost, confused, or clicking on the wrong stuff..

2. Copywriting

While the data is coming in, it was time to look at the copy and fix any obvious problems. This involved making:

  • headlines, subheadings and paragraphs show the benefits (why people should attend the school)
  • the copy user-friendly (no jargon or confusing words, particularly important because this website has an international audience)
  • changes to copy layout (sometimes the price was shown before the course description. For this website it was better to first show people why a course is good, and then show them the price)
  • button copy show a benefit of clicking (view course info, see syllabus are always better than ‘submit’ or ‘download’
  • sure visitors were reassured (adding elements that built trust, including testimonials, secure padlock symbols, major credit card logos)

3. Final report

This covered problems found, and how to fix:

  • Technical analysis
    (analysing site speed, checking it works on different devices)
  • Page performance
    (finding and fixing areas which are performing well/badly in terms of people leaving the site before converting)
  • Design
    (assessing how easy it is to use the website, checking the layout follows website best practice)
  • SEO
    (optimising html code, finding and fixing broken links, crawling site to find errors)
  • Removing barriers to converting
    (analysing data and doing user testing to find out why a high % abandoned at checkout, and fixing the problems)
  • Recovering abandoned carts
    (implementing remarketing strategy)

4. Results

In the 3 months after working with the client, we got the following:

40.7% increase in conversion rate

136.8% increase in revenue

Client verdict

“I contacted Steve Alphabet because I needed to freshen up the copy and wanted some fresh (experienced) eyes to look over the site and suggest any changes that would make the conversion process easier. I found the service great.  I liked the way that Steve took the time to understand my business and offer smart suggestions. I recommend Steve Alphabet to people who need either fresh eyes on an existing site or suggestions when building a new website.”
Business owner

By the way… it’s still possible to do conversion optimisation for websites with much fewer visitors, it just means that I go for a different approach. Either the process takes longer (until we have enough sessions and page views in Google Analytics to be able to make some hypotheses and draw conclusions). Or we don’t use Google Analytics, and do user testing (where I get 5–10 people to visit the website, carry out a few tasks, and then tell me how easy/difficult it was).