agent-analytics-autoresearchv1.0.12

A B2B SaaS company wants to improve its homepage hero copy, CTA, and supporting proof points to increase free-trial signups. The growth team has Google Analytics and Mixpanel funnel data but lacks a structured process for generating and prioritizing new landing-page variants. This skill runs a 5-round autoresearch loop from the existing analytics snapshot, produces two review-ready variants with a recommended split-test shape, and logs all rounds for auditability.

An online retailer notices a high cart-abandonment rate at the shipping and payment steps. Without dedicated CRO resources, the team needs a repeatable way to brainstorm, critique, and rank checkout copy and microcopy experiments. The skill ingests CSV exports from the checkout funnel, generates candidate variants for the payment step, and blind-ranks them with Borda scoring to surface the two most promising tests.

A consumer fintech startup sees that new users install the app but fail to complete KYC and first transaction. The growth lead can paste a mix of SQL query results and crash-log summaries into the run folder. The skill treats the brief and pasted data as product truth, runs an autoresearch loop focused on onboarding copy, and outputs variants for a push-notification or in-app tooltip experiment.

An API platform wants to test whether repositioning its free tier and changing plan descriptions increases paid conversion. The team has PostHog event data but no in-house experimentation analyst. The skill reads the pricing page brief and PostHog export, generates and critiques pricing copy variants, and produces a final_variants.md with two distinct pricing-page tests ready for engineering review.

A small conversion-rate-optimization agency manages multiple client landing pages and needs to systematize ideation while keeping human approval in the loop. Analysts run the skill per client surface using pasted client analytics dashboards and notes. The inner loop generates variants, but the outer experiment loop only proceeds after the client signs off, preserving the review-only default and client trust.

The product is sold on recurring monthly or annual plans, and the primary growth metric is often trial-to-paid conversion or activation. This skill helps product and growth teams turn funnel analytics into structured, ranked A/B test candidates for pricing, onboarding, and feature-adoption surfaces. It is especially useful when the team lacks a dedicated experimentation platform or statistical support.

Revenue comes from direct purchases, so small changes in checkout conversion, cart abandonment, and product-page engagement directly affect the bottom line. The skill ingests order-funnel snapshots from CSV, SQL, or analytics tools and produces reviewable copy and layout test variants. It supports a high-velocity testing culture without needing a full experimentation team.

Agencies bill clients for growth, CRO, or analytics engagements, often managing many landing pages and experiment backlogs. This skill standardizes the generation and ranking of experiment ideas from client data while enforcing human approval before implementation. It allows agencies to deliver structured, data-informed experiment recommendations faster and more consistently.