The Academy
Creating MyTutor's first content-driven learning platform
We delivered a brand new learning platform form for MyTutor’s tutor-base to empower them to increase the quality of their lessons. My work on this project was two-fold: First, I helped devise and execute the content strategy using human-centred design principles. Secondly, I led the design process for implementing the actual learning platform and ensuring tutors took engaged with the content.
MyTutor
Product Designer
October 1, 2019
This is a project with many moving parts that revolved around implement an online training platform ("the Academy"), initially aimed at new tutors joining MyTutor. I led the Product Design process at the Academy. However, my responsibilities extended into constructing the content strategy (i.e. what lessons MyTutor create?), I set the framework for vetting external provides for the Academy's infrastructure and finally, I facilitated the branding efforts of the video production.
For this piece, I'll primarily focus on how I helped Mytutor construct their content strategy and spend less time explaining the product design process on the academy (which was a reasonably straightforward process anyways)
MyTutor had a big problem on their Schools side. Stats showed that ~25% of schools churned after a term with MyTutor. Of those Schools who churned ~90% cite lesson quality and as a primary or secondary reason.
MyTutor learned through teacher feedback that the quality of school programmes varies significantly and has decreased as the business has scaled.
Part of the problem is that MyTutor matched their most inexperienced tutors with the most difficult pupils. Schools tutors are a subset of retail tutors. Because retail work is higher-paid, as tutors come more experienced, they tend to work with parents exclusively.
The problem means the students are progressing slower, both tutors and students don't find the experience enjoyable and it's hurts retention (70%-75%)
Also, MyTutor wanted to their business further and thus needed a broader tutor supply. To achieve that, they planned on start accepting tutors from less prestigious UK universities.
MyTutor wanted to target the low lesson quality and aimed to achieve retention of +85%. The Schools-team thus decided they wanted to provide specialised training to new and existing tutors on the platform.
My role in this project was to:
When starting out thinking about what content tutors might find valuable, we began by asking ourselves: What are the pain points tutors currently experience? What stands in the tutors’ way of delivering a good, quality lesson?
The content was primarily targeted at brand new tutors so wanted to understand new tutors mindset and struggles. We thus set up interviews with tutors who had yet to have their first lesson, and we set up chats with experienced tutors and had them reflect upon their first lesson on MyTutor and which tools they felt they needed.In addition to interviewing tutors, we review past lesson recordings to look for common problem patterns we might be able to target with the training. On the back of our research, we analysed the data and found amongst other things that our (newer) tutors
Tutors feel unsupported by the MyTutor
“I didn’t get any feedback at all as a tutor on MyTutor. [That] slowed down my improvement. Looking back, I don’t think I improved that much as a tutor."
They're terrified before their first lesson
"I've finally [got accepted to MyTutor] so I'm ready to teach... but I'm absolutely terrified of teaching my student the wrong thing/not knowing the answer to their questions! How does a first-time tutor get over this fear?"
Tutors struggle to strike the right balance between rapport and relationship
“I’m not bothered if we don’t have a close relationship - I’m not there to be a friend.”
Tutors do far too much of the work
“The biggest mistake I see tutors make is that they spend too much time just talking at the student. The students are too passive.”
To create the actual training curriculum, we invited three teachers in-house. We presented back our research insights and used snippets from lesson recordings that support our finding.
Based on the data from our presentation, I had the teacher go through an empathy mapping exercise. Given the teachers’ background (they're specialised in teaching in a classroom setting), I deliberately had them do this activity to help them get into the mindset of our young, often inexperienced online tutors. The exercise helped the teacher empathises with the challenges our tutor faced.
On the back of that empathy, each teacher brainstormed their version of the ideal training curriculum for online teaching. We then had the teachers present back their work and finally had them force rank each segment of their respective curricula, thus creating a master curriculum that could guide of content production.
Before we went full-scale developing the whole training course, we decided to do an MVP-version and gauge its receptions amongst a select group of beta testers. Do run the experiment, we shot 5, lo-fi videos on an iPad and shared it with our testers. We did a split between watching with a tutor gathering they're comments and feedback as we went along. We also sent the MVP-material to a group of tutors with a mixed experience to gather quantitative feedback.
Overall, we got positive feedback from the content of the videos. Though experienced tutors only found minor gains from the content, they unanimously agreed that new tutors would find it extremely helpful:
“Absolutely fantastic for first-time tutors - it’s exactly what is needed. Having tutored over 120 hours of lessons, some of the content was not new. However, I did learn several things and will be adjusting the way I deliver tutorials, particularly first lessons!”
Now that we felt a good convocation about the desirability of the content, the next challenge was: How do we offer it and how do we get tutors to engage?
Whilst I've been heavily involved in advising, planning, researching and facilitating workshops in the content and 3rd party vendor efforts, I lead the UX efforts. From the get-go, the rough idea for the solution was: MyTutor will offer training courses for its tutors in our efforts to increase the lesson quality. The main objective was to get 100% of our tutors to complete the training.
The tricky challenges we were trying to balance were:
When we started the design work, I set up a two-day workshop as I wanted us to move fast with some initial concepts. Since we already had a sufficient level of understanding and empathy of our users, I set up the workshops like so:
Through our pre-validation efforts, we were already confident that the training material was valuable to our tutors. However, since we hadn't formalised a value statement, each of us had a different interpretation of it which I feared muddle our efforts.
So before we even started to think about the user journeys or generating ideas, I wanted the team to align on: What's the offering and how might tutors find that offering exciting? If we had a formal value proposition down, I reasoned, that work would trickle down into all aspects of the project: From the concepts, we'd imagine, to UI copy to general messaging.
I thus took the team through a Value Proposition Canvas exercise to get everybody's perspectives on the table. Doing so really helped the team to get together on what we were offering and why tutors might find that exciting.
Next up was to start brainstorming about ways we could get MyTutor's tutors to take the training. To do that, we went through a few rounds of crazy 8's sketching + presenting back. I then broke the team into smaller groups that each drew their solutions which they presented back.
When we had our ideas down, I split the group into smaller teams and had them each generate their version of the ideal user journey from sign-up to completed training. Through several iterations of presenting back versions of the ideal journey, we settled on our master journey. That journey formed the backbone of our storyboard which I used for our first prototype.
After we'd aligned on a master-story, we prototyped the flow and booked sessions with a handful of tutors. The prototyping and validation processes were straightforward: We built an interactive prototype, put together a test script and invited a bunch of tutors to test with us throughout a week. In between sessions, we were able to iterate on the confusing bits and to trim the steps in the flow. We kept at this process throughout the week til we felt a good conviction about the product.
We’ve already done the hard groundwork leading up to the prototyping stage so from here-on-out everything was pretty straight-forward: Build a flow, write a script and iterate with users till we were confident about the desirability and usability of the feature. After several rounds of validations, we specced up the designs and handed them off to our off-shore developers.
Upon release, the adoption rate was better than expected.
“Have been tutoring with my tutor for a few years now and found it really nice to go back through the basics and get some new inspiration!”
On the impact of the training, we found that top-performing tutors didn't change their lesson quality by any meaningful measure. However, new tutors, inexperienced tutors significantly improved their quality which was a huge net-benefit for everyone involved.
From its initial launch, the academy grew as the tutor team released ever more training courses targeted at different aspects of tutoring. I chimed in on some of the course projects to help facilitate content shaping workshop and to help the team think more strategically about value propositions. Furthermore, the academy became instrumental when MyTutor branched out and started offering partnership courses. And given the success of the Schools-training, the Private side of MyTutor started releasing their own training material
After a few months of adding more content with a different audience in mind, the initial structure and framework of the Academy started to prove unscalable. I thus set out to revamp the Academy, giving it a scalable structure and a visual update.
