StepGuide
Provides interactive, step-by-step solutions that scaffold learning from foundational skills to final answers, making math accessible for every student.
MY ROLE
Product Designer
TEAM
1 PD, 1 PM, 2 engineers
TIMELINE
Dec'23-Mar'24
What is StepGuide and VoyageMath?
StepGuide is a step-by-step explainer built into VoyageMath, a classroom-first math tool designed to offer structured practice and boost student confidence - especially for those who struggle with the subject.
StepGuide is where the magic happens. It breaks down tough math problems into clear, visual steps so students can focus on solving - not second-guessing.
It's especially useful for:
Students who feel overwhelmed by multi-step problems
Learners who struggle with confidence in math
Kids who get stuck when there's too much text or not enough structure
StepGuide helps by:
Showing one step at a time
Using simple visuals to explain what to do
Keeping the experience interactive, not lecture-like
The goal:
Make math feel less scary
Help students feel progress after every step
Turn mistakes into small wins
🧩 Why We Built StepGuide
Post-COVID learning loss left middle schoolers in the U.S. years behind in math
Teachers were stretched thin and couldn’t offer personalized support at scale
Most practice tools were either too basic, too rigid, or didn’t give helpful feedback
Our response
VoyageMath was built to restore student confidence through guided practice - especially for those who had fallen behind.
🪜 Step by Step: How We Built StepGuide
Back in school, I loved math - but I watched so many classmates freeze at the sight of numbers. They’d ace science or English, but math? It brought panic. I always wondered why. One big reason: there was never enough personal help to go around. One teacher, twenty students - not everyone got the attention they needed, especially those falling behind.
That memory stuck with me.
With StepGuide, we wanted to build the kind of tool we never had. Something that could step in when a teacher couldn’t, explain things calmly, and walk students through a tough problem without judgment. A tool that worked at their pace. A tool that felt like solving math with a friend.
We didn’t get it right the first time. In fact, we didn’t get it right the first five times. Here’s how we iterated our way to what finally worked:
The Friendly Chatbot (First Prototype)
Our first design was a chat-style interface inspired by ChatGPT. The problem was broken into 3-4 steps, and each step had:
A step name pill
The question being asked
A hint
MCQ options
Feedback like “Brilliant! Step 1 complete.”
Students got two chances per step, and if still wrong, we showed the answer. Then came the next step.
💡 What we learned:
Students don’t read long contexts. Many gave up during the context step before even reaching the actual question.
The quiz screen's question was enough to establish context - repeating it in StepGuide didn’t help. They preferred breaking it down step-by-step.
Encouraging feedback was liked but keeping it permanently visible felt like filler and disrupted the connection between steps.
A significant number of students completely missed the hint section.
Most couldn’t recall the full sequence of steps when asked to explain the solution afterward - defeating the purpose of StepGuide as a learning tool.
Making Steps Sticky
🔄 What we changed:
💡 What we learned:
Shortening the context helped reduce overwhelm and built confidence
Styling validation messages separately improved readability
Highlighting the hints worked against us - students read them first before even looking at what the step was asking them to do
Sticky headers went unnoticed
Students still couldn’t recall the full flow after solving, which meant StepGuide wasn’t sticking
Sidebar Steps & Integrated Hints
🔄 What we changed:
💡 What we learned:
Students ignored the sidepane even though steps were visible
The screen felt noisy and overloaded with elements
Merging hints into the question helped a bit, but it wasn’t enough
Students still needed more visual scaffolding to make sense of each step
Flow between steps wasn’t memorable - they were still seeing steps as isolated
Collapsing the Noise
🔄 What we changed:
💡 What we learned:
Collapsing previous steps made the interface feel less crowded
Students focused better and followed the steps without distraction
Recall improved - they remembered the solving flow across problems
Example box received mixed feedback - helpful for some, confusing for others
Some students misunderstood the example as part of the actual question
Pen and Paper, Reimagined
🔄 What we changed:
💡 What we learned:
Students finally understood the connection between steps
They could explain the full flow after solving 3-4 problems
Clean layout and pen-paper mimicry helped them feel in control
Visual structure clarified what was being asked in each step
Example box continued to confuse - still felt redundant
Cutting the Clutter
Final Version
🔄 What we changed:
💡 What we learned:
Removing the example box removed distractions
Students completed StepGuide without getting stuck or confused
The design felt intuitive and non-threatening
Students described it as a helpful companion rather than extra work
Final version saw universal preference during testing
StepGuide wasn’t just explaining math. It was helping students understand how to solve it, one clear step at a time.
But this wasn’t the end of the story - there was still a motivation gap.
🚧 The Motivation Gap
We had finally built a StepGuide experience that made sense - it was clear, structured, and visually digestible. But there was still one critical problem:
students didn’t want to use it.
We noticed students skipping StepGuide as soon as it appeared. Many didn’t see its value. To understand this better, we broke down the ROI (return on investment) from a student’s perspective:
ROI Perspective: Why Students Chose to Skip StepGuide
Aspect
Current Incorrect Question
(Using StepGuide)
Moving to Next Question (Skipping)
Effort Required
Higher (Several Steps to Correct Mistake)
Lower
Mastery Score Impact
None (Current Setting)
Possible Gain if Guess is Correct
Step Rewards
Half the Steps Compared to Correct Initial Answer
Full Steps if Answer is Correct
Learning Potential
Higher (Detailed Understanding of Mistake)
Lower (Missed Learning Opportunity)
Question Limit Awareness
Not Apparent (Students think questions are unlimited)
Possible Gain if Guess is Correct
Clearly, from the student’s perspective, skipping felt easier, faster, and potentially more rewarding - especially for students who were guessing their way through or felt behind.
We also observed three behavioral cohorts:
40%
Skippers
Never used StepGuide. Saw it as an interruption.
30%
Burnt
Completed at least one StepGuide, but skipped most others after that.
20%
Religious
Used it consistently. These were rare.
Why? Because StepGuide felt like a punishment for getting it wrong.
And when StepGuide launched, we replaced the mountain view (our gamified progress bar) with the StepGuide panel. Students couldn’t see the connection between their efforts and the game anymore.
Even though solving StepGuide did help them climb the mountain - they just couldn’t see it.
What we did to fix it:
Repositioned StepGuide as a reattempt, not a forced help screen:
Students could opt-in to StepGuide after a wrong answer
Added incentives: bonus break room time, loot boxes, progress boosts
Added a mini mountain view inside StepGuide:
Students saw themselves climbing with each correct step.
No movement if the step was incorrect.
Made progress feel tangible again.
Reframed effort as power:
We made StepGuide feel like a smart move, not a fallback.
StepGuide became part of the game loop, not a detour.
📈 What changed after this experiment
Our goal was to convert the 40% of students who skipped StepGuide entirely into students who would at least try one StepGuide. At the same time, we aimed to help the 30% who only did one or two become consistent users.
📉 Before the change:
40% of students skipped StepGuide entirely.
30% did StepGuide at least once, then skipped the rest.
20% were consistent users (the "religious" cohort).
📈 After repositioning StepGuide as a reattempt with rewards:
Skippers dropped to 30%
That 10% moved into the "at least one complete" cohort
StepGuide felt more like a strategic retry and less like a consequence
This repositioning, paired with the mini mountain view, significantly increased voluntary engagement. Students could now see the impact of their effort and felt rewarded for trying.
Suddenly, students weren’t avoiding StepGuide. They were choosing it - not because they had to, but because it helped them win.
Learning became a part of the game. Not something that paused it.
🎯 Final Design
We now have almost 500 skills in VoyageMath, each with three difficulty levels - nearly 1500 unique StepGuide designs. Multiply that by ~35 questions per skill, and you're looking at over 52,000 StepGuide experiences.
While I laid the foundation of StepGuide with my PM, EM, and co-founder - starting with just 5–6 skills - the early momentum brought more hands into the fold. Engineers, designers, and content experts all pitched in.
Today, a 20-person team (10 content designers and 10 subject matter experts) continues to scale and evolve StepGuide. With every skill, we learn something new - and fold it back into both new and old designs.
It’s become a rare and powerful feedback loop. There’s still nothing quite like it in the market.
Each StepGuide interaction is designed to feel effortless, visual, and rewarding.
What it includes:
It’s built to be:
Visually Digestible
Text-light and to the point
Structured without being linear
Deeply connected to the game loop
Most importantly, it doesn’t feel like help - it feels like momentum.
📊 What We Saw
500k+
StepGuides completed
30%
accuracy boost after 2–3 StepGuides
Correct next response
after using StepGuide
STUDENT RESULTS
Group work, review & intervention
is how teachers use StepGuide
“Miss, I don’t get it”
is now rarely heard
Independent Practice
is what teachers trust StepGuide for
TEACHER IMPACT
🪄 Closing Thoughts
StepGuide isn’t just a support system. It’s a confidence-builder.
We didn’t just want students to get the right answer. We wanted them to say:
“That wasn’t so bad. I can do the next one.”
And for the kids who usually give up on math?
That’s everything.
~ THE END ~
Explore More Work
Ready to see how these StepGuide principles came together in a complete student experience?
