Redesigning the Trial Lesson to Improve Learner Conversion

Real-world EdTech project · Existing product · English & Spanish courses

A note on confidentiality: This was a real-world project, so I can’t share the original product screens or proprietary materials. The diagrams in this case study are simplified reconstructions, while the research, design decisions and outcomes reflect the work I carried out on the project.

Context & business problem

Our language-learning product used a flipped-classroom model: learners studied new material independently on the platform before class, then used live lessons mainly for speaking and practice.

The sales team noticed a problem. Learners who took the flipped-classroom trial lesson were converting to paid courses at a substantially lower rate than learners trying a more traditional lesson format.

The business goal was clear: improve trial-to-paid conversion.

From the learner's perspective, however, the trial also needed to do something more fundamental: help them experience the value of the course, learn something useful, and understand whether this way of learning suited them.

My role: instructional design, learner research, experience audit and redesign, and collaboration with teachers, sales and other stakeholders.

The existing experience

The existing trial lesson tried to achieve several things at once.

Before the lesson, learners completed a level test. During the lesson, the teacher:

briefly checked the learner's level;
explained the structure of the trial and introduced the platform;
guided the learner through part of a self-study lesson;
demonstrated different platform activities and explained how they worked;
switched to a separate presentation for conversational practice;
finished by explaining the flipped-classroom model and answering questions.

In theory, this gave learners a complete overview of the product.

In practice, I suspected we were showing too much of how the product worked and not giving learners enough time to actually experience its value.

Research & diagnosis

I deliberately started by investigating the experience rather than immediately redesigning the lesson materials.

1. Checking whether the problem was teacher-specific

One early hypothesis was that the lower conversion might be connected to how individual teachers delivered the trial.

I requested conversion data for teachers running flipped-classroom trials and compared their results.

The data did not support this hypothesis: the problem appeared to be broader than individual teacher performance.

That shifted my focus from "Who is delivering the lesson?" to "What is happening across the trial experience?"

2. Reviewing real trial lessons

I watched several trial lesson recordings to understand how the designed experience translated into an actual lesson.

This helped me identify recurring patterns such as long explanations, frequent context switching and rigid adherence to the lesson structure.

3. Working with teachers

I organized a collaborative lesson-study session with teachers.

We observed a trial lesson using pre-agreed criteria, discussed points of friction and brainstormed possible improvements.

Teacher feedback was particularly valuable because teachers could explain not only what they were doing, but why. It also helped me identify questions to explore later with learners.

4. Looking at competitor experiences

I reviewed trial experiences offered by direct and indirect competitors and gathered additional observations from colleagues involved in regular competitor research.

I used this mainly as context: I wanted to understand what learners might already expect from a language-school trial experience rather than simply copy competitor solutions.

5. Talking to learners

I created a focused interview guide and spoke with nine existing learners about their trial experience.

I asked what had felt confusing, unnecessary, difficult, disappointing or different from what they expected.

Because these participants had ultimately chosen to continue learning with us, this research had an important limitation: it could tell me where the experience created friction, but not fully explain why some trial learners decided not to convert. Interviewing non-converting learners would therefore have been a valuable next research step.

Key insights

Several themes appeared repeatedly across interviews, teacher feedback and lesson observations.

1. Too much explanation, not enough learning

Many learners felt that too much of the lesson was spent explaining the school, the flipped-classroom model or basic platform functionality.

Some of this information had already been provided by the sales manager.

Other explanations, such as which button to click or how automatically checked exercises worked, felt obvious to learners or were things they preferred to explore independently.

The result was that learners felt they had too little time to actually learn and speak during the lesson.

2. The experience felt fragmented

Teachers moved between several tools during a single lesson: for example, Miro, the learning platform and a PowerPoint presentation.

Some learners found this distracting and confusing.

Instead of experiencing one coherent product, they were experiencing several disconnected pieces.

3. The experience did not always adapt to the learner

One learner described receiving a lesson below their actual level because the teacher was unsure how to respond when the preliminary level information was inaccurate.

Others felt teachers sometimes followed the prescribed lesson sequence even when an activity was clearly too easy or unnecessary.

Several learners wanted more time for conversation and more flexibility in how lesson time was used.

4. The content did not always demonstrate enough value

Some learners simply found the trial topic uninteresting.

Combined with the limited time available for meaningful practice, this made it harder for the trial lesson to demonstrate what the full learning experience could offer.

5. There were accessibility and usability issues

Some learners struggled to read small text while the teacher shared their screen in Zoom.

I also observed opportunities to improve contrast and make the materials easier to follow visually.

6. Gaps between touchpoints created unnecessary repetition

One especially useful discovery came from comparing teacher and learner perspectives.

Teachers were explaining the flipped-classroom model because they believed learners needed that information.

However, learners had often already received the same explanation from the sales team.

The problem wasn't simply "teachers talk too much." It was a service-design problem: different parts of the learner journey were not sufficiently coordinated.

The research showed that the problem went beyond the lesson materials. The trial experience was shaped by the handoff from sales, platform content, teacher guidance and the live lesson itself. Improving conversion therefore meant looking at the whole journey rather than redesigning individual activities in isolation.

Design decisions

The research changed the goal of the trial lesson.

Instead of trying to explain every part of the product, the redesigned experience focused on helping learners experience the value of the product first, while making additional information available only when it was useful.

Reduce repeated explanations

We removed information about the flipped-classroom model that duplicated what learners already received from sales.

Teachers could still answer questions at the end of the lesson, but no longer needed to repeat the full explanation by default.

Move platform onboarding out of the trial

Instead of teaching learners how every type of activity worked during the trial, we created a separate introductory lesson at the beginning of each course. Learners could explore it when needed.

During the trial, teachers were encouraged to explain platform interactions only when the learner actually needed help.

Give learners more time to experience the product

By shortening the introduction and removing unnecessary explanations, we could spend a larger proportion of the trial on conversational practice and meaningful learning.

This directly addressed one of the strongest themes in the interviews: learners wanted to spend more time learning, not hearing about learning.

Make the experience more adaptive

I created guidance for teachers on:

checking whether the learner's assigned level was appropriate;
smoothly moving to a different level when necessary;
choosing activities based on the learner's ability and interests;
skipping or reordering activities when appropriate;
handling common technical problems.

The overall lesson structure remained standardized, but teachers had more freedom within it.

This was an important balance: we needed enough consistency for a scalable product while allowing teachers to respond to the person in front of them.

Reduce context switching

I brought the warm-up, platform activities and conversational practice into one platform experience instead of asking teachers to move between several tabs and tools.

We also added a short explanation showing which parts learners would normally complete independently before class.

This allowed us to demonstrate the flipped-classroom model without interrupting the lesson to explain it at length.

Improve the materials themselves

We reviewed trial topics and activities to make them more engaging and introduced more activity options so teachers could adapt lessons to different learners.

I also improved readability by increasing text size and contrast.

Support teachers, not just redesign the interface

Because teacher behavior was an important part of the learner experience, changing the lesson materials alone would not have been enough.

I created and delivered training on the redesigned format, provided opportunities to practice it, recorded the session for teachers who could not attend and updated onboarding materials for future hires.

The training emphasized an important principle:

The trial lesson has a structure, but the learner experience should not feel rigid.

Constraints and trade-offs

Ideally, I would have redesigned every trial lesson from the ground up. That wasn't realistic.

The team had limited time and several competing priorities, so we prioritized the areas most likely to have an impact.

We fully redesigned the trial lessons with the lowest conversion rates.

For the remaining lessons, we made smaller, scalable improvements (better readability, more conversational options, removal of unnecessary activities and clearer teacher guidance) while investing more heavily in teacher training.

This allowed us to improve the wider experience without waiting for a complete content redesign.

Outcome

Before the redesign, the school coordinator identified an unusual decline in trial-to-paid conversion across the flipped-classroom teaching team.

Conversion was monitored monthly by teacher, with rates below 50% flagged for further investigation. At the time, several teachers had fallen below this benchmark, while others who had previously converted more than 70% of trial learners had dropped to around 60%.

Because conversion could vary seasonally, we also considered the time of year, but the decline did not match the school's usual seasonal pattern.

After the redesigned trial experience was introduced, the overall conversion trend began to recover and remained positive as the new format became established. Individual teachers could still occasionally fall below the 50% benchmark and require additional coaching, but the team-wide decline that had triggered the project did not persist.

Because this was an operational redesign rather than a controlled experiment, I would not attribute every change in conversion to the redesign alone. However, the timing and sustained recovery provided a strong signal that the changes were moving the learner experience in the right direction.

More importantly, the project changed how we thought about the trial lesson.

It was no longer treated simply as a collection of teaching materials. It was an end-to-end experience involving the sales team, platform, content, teacher behavior and learner expectations.

What I'd measure next

If I continued the project, I would look beyond overall conversion and track where the improvements were coming from.

I would compare:

conversion by lesson, level and teacher to detect remaining patterns;
the proportion of trial time spent on speaking and meaningful practice;
learner ratings of perceived lesson value and relevance;
recurring questions or points where teachers still need to explain the product;
accessibility or usability issues reported during trial lessons.

I would also interview learners who did not convert after the trial. Their perspective would help distinguish usability and experience problems from other reasons for not purchasing, such as price, timing or product-market fit.

What I learned

The project reinforced three ideas that have strongly influenced how I approach UX and product problems.

Look beyond the most visible part of the problem.
At first, lower conversion could easily have been treated as a content problem or a teacher-performance problem. Looking at data, recordings, teacher feedback and learner interviews revealed that the friction was distributed across the whole journey.

The people delivering an experience are part of the experience.
Teacher feedback helped me understand behaviors that looked unnecessary from the learner's perspective. In several cases, teachers were compensating for missing context or unclear processes rather than simply making poor choices.

Optimizing for the business goal and the learner goal did not require competing solutions.
The business needed better conversion. Learners wanted less explanation, more useful learning and an experience adapted to them. Improving the learner's ability to experience the value of the product was also the most direct way to support the business goal.