When not to code: why simple tools sometimes work best

Context — the prompt and the constraints

Academics asked for a quick way to sense how a cohort was feeling during workshops. Needs:

• Low friction for students;
• Fast, interpretable visual for facilitators;
• Ideally embeddable in Canvas and easy to reuse.

Early approach — Mentimeter AI mocks

Mentimeter's AI assistant produced a workable flow in minutes. It gave us slider-style prompts and optional reflection text.

Pros: speed, existing SSO, easy sharing.
Trade-offs: limited layout control and visual customisation.

Bespoke exploration — Firebase Studio "VibeCheck"

In parallel I built a lightweight prototype to see what extra value a custom tool could offer: circumplex mood input, individual rows, class distribution, AI summary, and QR sharing.

Second iteration improvements

Clearer UI and typography, aggregated cohort dot-plot, facilitator "AI-powered vibe summary", and a share modal with QR code.

Iteration to match the Miro heat map

After seeing Elysia's heat-bar design in Miro, I forked the Firebase Studio project to test whether the bespoke tool could mirror that look and feel (gradients, simple sliders, relatable labels).

Outcome: I could replicate the visuals, but the effortless drag-and-drop, everyone-on-one-board interaction still required hours of additional engineering (state sync, collisions, role controls, etc.). With a working solution already in place, more build time wasn't justified.

The parallel solution — Elysia's Miro board

As a first-time Miro user, Elysia created a vibe-check board with draggable circles over gradient heat bars ("big bleh" → "11/10 awake & keen", "strong yikes" → "totally fine"). Students understood it instantly; it embedded into Canvas; QR sharing was trivial.

Why it won for this context

• Zero new logins or onboarding
• One screen for the class; live movement is self-explanatory
• Rapid authoring and reuse as a template
• Clean embed in Canvas, plus QR option

Problem and constraints (across options)

• Keep student effort near-zero; anonymity toggle desirable
• Show a cohort-level visual at a glance
• Avoid maintenance burden; use existing systems where possible

Approach and evolution (quick cycles)

• Need shaping: define facilitator goal and outputs.
• Platform scan: Mentimeter AI prototypes; feasibility check.
• Custom spike: Firebase Studio app to test value beyond slides.
• Evidence check: trial Elysia's Miro board with a real class.
• Fork & style test: match Miro gradients in Firebase; assess effort.
• Decision: pause bespoke; adopt Miro template; document rationale.

Decision and next steps

• Adopt Elysia's Miro board as a reusable template for workshops.
• Pause Firebase Studio build; keep the repo as a reference for future needs (e.g., analytics, longitudinal trends).
• Share template with LEI as part of a curated Miro template set.

What I learned

• Sometimes not coding is the user-centred choice when an existing platform meets the need with less friction.
• Visual parity ≠ usability parity; replicating "simple" behaviours often hides significant engineering effort.
• Embedding into the LMS trumps novelty when facilitation time is tight.

Acknowledgements

Thanks to Dr Elysia Sokolenko, Dr Bonnie Williams, and Anna Leonard for the collaboration and rapid feedback.

Disclosure on AI use

Mentimeter's AI assisted the initial survey flow; Firebase Studio hosted the bespoke prototype; I used AI to accelerate copy and UI ideas. Final design and decisions were human-led.