Improve Youtube Studios

First, I’d clarify what a better learning experience means. Someone can finish several videos and still struggle to apply what they watched. I’d focus on helping learners understand the content, practise it, and confirm that they’ve achieved their goal.

I’ll start with the product and scope, choose a learner segment, identify their needs, and then prioritise solutions based on impact and effort.

1. Understand the product and clarify the objective

YouTube supports different learning needs through video: understanding a concept, following a tutorial, developing a professional skill, or exploring a hobby. The experience also extends beyond the video through features such as chapters and discussions.

Before choosing an improvement, I’d ask:

• Are we focusing on the main video experience or another offering, such as Studio?

• What learning outcome or business objective should we improve?

• Is there a particular geography or learner group?

• What are the timeline and budget?

For this case, I’ll focus on the main video platform, with 12 months and a $200,000 budget. The goal is to make YouTube a preferred destination for achieving learning outcomes.

I’d design something applicable across markets, but test it with a limited set of subjects and creators. A global rollout would be too broad for this initial budget.

YouTube’s advertising and subscription model provides business context, but I’d judge this improvement primarily by learning outcomes and repeat learning behaviour.

2. Choose a learner segment

I’d consider three groups, using illustrative personas:

These groups reflect different life stages, occupations, and interests. I’d also consider language, connectivity, available study time, and accessibility needs.

I’d focus on Peter’s segment: people who already use YouTube regularly to learn a specific skill.

They’re familiar with the product and have repeated learning needs. Their goals can also be broken into assessable milestones, such as building a simple financial model or playing a chord progression.

I’d validate their available time and resources rather than assume every student has them. Other segments also have measurable goals, but this group offers a focused starting point.

3. Identify the barriers to learning

I’d map the journey as:

Set a goal → Find relevant content → Choose a sequence → Watch and understand → Practise → Get feedback → Apply the skill

I see four needs:

• Content relevance: “Will this video help me achieve my goal, and what should I watch next?”

• Progress tracking: “How much have I learned, and am I ready to test myself?”

• Validation and accreditation: “Can someone assess my work, and will that assessment have value outside this platform?”

• Understanding and engagement: “Can I follow the explanation, clarify unfamiliar terms, and stay involved?”

I’d investigate these through learner interviews, observation, and data on where people stop or repeatedly revisit content. I’d also check how existing features and creator practices address them before proposing additions.

4. Prioritise the needs

I’d compare customer impact with business impact. Here, business impact means increasing the likelihood that learners achieve a goal and return for another one.

Using a 1–5 scale, my initial hypotheses would be:

These scores are assumptions to validate, not research findings.

I’d prioritise understanding the content and validating the learning. They address two connected questions: “Do I understand this?” and “Can I actually do it?”

I’d distinguish validation from accreditation. A tutor can provide useful feedback without offering a formally recognised qualification.

5. Explore possible solutions

I’d consider three approaches.

Assess the learner’s starting level.

After choosing a goal, the learner could complete a short assessment, perhaps ten questions, to identify suitable beginner, intermediate, or advanced content. The results could inform a personalised playlist. For practical skills, I’d include a task because multiple-choice questions alone may not reveal ability.

Connect learners with tutor assessment and support.

Participating tutors could offer scheduled group Q&A sessions and assessments of learner work. A paid assessment could compensate the tutor through a revenue share. Any certificate should clearly state who issued it and what was assessed. University recognition would require an actual partnership.

Make each chapter more useful for learning.

Creators could add short summaries, timestamped notes, takeaways, and homework. A conversational learning assistant could help explain concepts or answer questions using the lesson material. A Gemini-powered assistant is one possible implementation, but it would need accuracy checks and ongoing cost estimates.

6. Prioritise the solutions

I’d compare expected learning impact with development effort and ongoing delivery costs.

My recommendation would be chapter summaries and notes, scheduled tutor Q&A, and a limited tutor-led assessment.

For example, someone learning financial modelling could review a chapter summary, save a note against a confusing section, complete a small modelling exercise, and discuss difficulties in a group session. They could then submit a final task for assessment.

I’d use group Q&A and assess a defined milestone rather than arrange a live assessment for every video. That gives us a more manageable pilot.

7. Evaluate trade-offs and feasibility

Tutor involvement could improve feedback, but it also creates recurring costs and scheduling constraints. I’d offer recordings and asynchronous question submission alongside live sessions.

Summaries can help revision, but learners may mistake reading a summary for mastering a skill. Practice tasks would help reveal that gap.

Paid assessments could exclude some learners. I’d keep core learning support and self-check activities accessible, with paid tutor review as an optional layer.

For the $200,000 budget, I’d estimate engineering and design, tutor compensation, content preparation, accessibility, evaluation, and contingency. I’d narrow the pilot until those costs fit. I wouldn’t claim the full solution is affordable without that exercise.

8. Measure success and make the recommendation

My north star would be the number of learners who demonstrate their chosen learning outcome through a meaningful assessment. I’d track the corresponding success rate among learners who start, so growth in participation doesn’t hide poor outcomes.

Supporting metrics would include:

• Assessment participation and completion.

• Pass rates and improvement from an initial assessment.

• Use of chapter summaries and notes.

• Q&A participation and resolution of learner questions.

• Return rate for another learning goal.

I’d also monitor abandonment, repeated assessment failures, tutor workload, cost per assessed learner, and complaints about feedback quality. Failure rates would need interpretation: an easy test could inflate success without improving learning.

I’d start with a small “learn, practise, and get feedback” experience. If it helps learners demonstrate a skill and return for another goal, we’d have evidence to expand into personalisation or more advanced learning assistance.