How AI Optimization Works

Updated February 14, 2026

An overview of the AI pipeline that analyses and improves your listings.

How AI Optimization Works

Selletto Pro uses Google's Gemini 2.5 models to analyse your listings and generate improvement suggestions.

The Pipeline

Data collection – We gather your listing data (title, tags, description, stats) and anonymised competitor signals.

Model routing – A cost-optimised router selects the appropriate Gemini model:

- Simple tasks (keyword suggestions) → Gemini Flash-Lite

- General optimisation → Gemini Flash

- Complex strategy (pricing, A/B design) → Gemini Pro

Analysis – The model evaluates your listing against Etsy SEO best practices and category-specific benchmarks.

Insight generation – Structured suggestions are returned in seven categories: title, tags, description, pricing, keywords, category, and image guidance.

Caching – Results are cached in Redis for 24 hours to minimise cost and latency.

Applying Suggestions

Suggestions are shown in the AI Insights Panel on each listing page. You can:

Preview the suggested change without applying it.

Apply a single suggestion with one click.

Apply All to accept all suggestions at once.

Dismiss suggestions you do not want.

Applied suggestions are tracked in the Performance Insights dashboard so you can measure their impact.

Monthly Budget

AI usage is capped at $500/month across all users. Individual usage is tracked and shown in Settings > AI Usage.

Was this article helpful?