Case study: +59% growth in registrations for the online school of English
An online English language school approached us with a request to increase the conversion rate on the site. The students are children from 4 to 12 years old who learn English without using their native language. For better assimilation of materials, graphics, games, and motivational techniques are used. This is what attracts and interests young schoolchildren in the educational process.
The main goals:
- Increase the conversion rate: increase the number of leads for registration for the first free lesson.
- Collect effective conversion elements for the development of the following promotional pages.
- Develop CRO processes and A/B tests from the hypothesis generation stage to the launch of tests and the implementation of the winning variant.
Interesting fact: Novakid has a large volume of traffic from European countries — one million sessions every month.
Stage 1 – Preparation and Data analysis
- We found that the site’s landing pages only contain general information about training. But this is not enough to cover the entire segment of the audience and stimulate it to make decisions faster.
- Our client already had a team working with analytics (Google Analytics). Therefore, most of the information was set up and transmitted. In addition, we used the Hotjar tool, which helped us collect information on heat maps, record and study video sessions, and also launch a survey for users.
- Based on the collected materials, we built several hypotheses and tests for individual audience segments and began testing variations.
What did we do?
- The CRO checklist assisted in determining the company’s weak points.
- Heuristic analysis – based on experience, we collected the hypotheses and observations.
- Hotjar – helped to analyze user behavior, active and inactive areas of pages, to obtain information about advanced desires of platform clients.
- Google Analytics – thanks to this tool, it was possible to validate several ideas and concepts of how interaction with certain elements affects key indicators.
- As a result, we compiled a list with final recommendations, hypotheses and priorities.
Stage 2 – Implementation
For the process to be not only high-quality but also fast, we divided the work into four directions:
a) Pre-configuration of analytics and events for collecting additional information.
b) Creation and launch of designs developed thanks to hypotheses
c) Preparation of surveys for users to collect feedback and insights.
d) Launching and tracking the effectiveness of tests.
Plans and deadlines: 2 tests per month in 3 countries = 6 tests per month. For sure, we could run more tests, but there was a complexity in creating content. After all, it was necessary to develop strategies for different locations and make a translation for them.
Interesting fact: During the tests, we saw the best winning result of +163% improvement and the worst result of -42% in conversion rate.
Eight months of cooperation did not pass in vain:
- Validated 57 hypotheses;
- Conducted 25 more and less successful A/B tests;
- The average improvement that we saw across all tests was +25%. Here we include the best winning results +163% improvement and the worst result -42% in conversion rate.
- Launched polls that helped us better understand the client’s desire;
- Scheduled the tests to distribute the workload to different departments of the marketing team.
- Average test duration – 2 weeks.
An example of the implemented test:
Observation: There is no general information and convincing benefits on the landing page. There is not enough social proof, unique advantages and benefits visualization.
Our hypothesis: Сreate visual confirmations of the school’s reliability, quality of service provision, and guarantees to get users’ results.
Recommendation: Display an infographic (the number of people who trust the institution, how much training has been completed, and which countries are children from). Also, we proposed to show various awards and other elements of quality confirmation.
Testing objective: the first step of the sales funnel – registration for the first free trial lesson.
Results: +59% growth of registrations for the first lesson. See the details on the image below:
As a result, it is important to note that before you run A/B tests, you need to configure the analytics as well as possible and ensure that it works correctly.
Our job is to make hypotheses based on data, not personal experience. That’s why in every new project, we use several tools to collect data. It’s important for us to track every click on the page to perform a correlation analysis and understand what affects conversion rates. Also, it’s important to use heat map analysis and video session recording. This will help us understand exactly what the user is doing on the page and why, and it will also help us build a user behaviour framework.
If the data is not enough to create a hypothesis, we conduct the survey among the visitors of the page and this way, we collect more information and validate the hypothesis. In general, as we wrote earlier in this article, it’s important to first collect data and conduct a UX audit, then make a list of hypotheses based on the data and only then start running tests.
If you want to know your “money-leaking” places on the website and fix them, you can leave a request for a short strategy call, and our manager will contact you.