UX case study: SynthAI's multimodal prompt editor and the cost of flexibility

AI · 6 min read

SynthAI's initial product shipped a single canvas that let users combine text, images, and structured data into prompts. Power users loved the expressiveness, but onboarding tests showed novices were overwhelmed. Error rates increased and help queries spiked, signaling that flexibility had a steep learning curve.

The product team conducted task-based usability testing and persona mapping, which revealed two dominant user journeys: exploratory novices who wanted guided templates and experts who required full composability. The redesign introduced split paths: 'Guided Builder' with templates, inline explainers, and safety defaults, plus 'Advanced Canvas' preserving the original composable experience for experts.

Post-launch analytics indicated the Guided Builder drove a 2.5x increase in successful first prompts and reduced help-desk tickets by 38%, while Advanced Canvas users retained their deep workflows. SynthAI's case emphasizes the trade-off between flexibility and discoverability and shows designers can reconcile both by offering mode-aware flows rather than one-size-fits-all interfaces.