DocumentationMCP ToolsAPI ReferenceGuidesResources
Guides/Overview

The Discovery Optimization Guide

Discovery Optimization is the unified practice of making your brand findable everywhere people now discover things: algorithmic search (Google), AI answer engines (ChatGPT, Claude, Perplexity, Gemini), and the SERP itself. It then wires every gap directly to the work that closes it. It centers the outcome of being discovered, not the mechanism of any one ranking algorithm.

This guide is a growing set of techniques for doing that work. Each chapter is a repeatable methodology, something you (or an agent) can run end to end. They are organized by the question that actually separates them: which signal does the technique read?

How this maps to SEO and AEO#

You probably know these surfaces by their usual names. SEO is optimizing for algorithmic search engines like Google. AEO (also called GEO or LLMO) is optimizing for AI answer engines like ChatGPT and Claude. They are not competing disciplines. They are two surfaces of the same job, getting discovered, and the SERP increasingly blends them. Discovery Optimization is the practice that treats all of it as one.

You may call itIt optimizes forIn this guide
SEOAlgorithmic search (Google, Bing rankings)Search Performance Analysis (your own search data)
AEO / GEO / LLMOAI answer engines (ChatGPT, Claude, Perplexity)Category Visibility and the answer-engine methodologies
The SERP (incl. AI Overviews)The results page where the two now blendAgentic SERP research

Your own search data#

The signals you already own: the impressions, clicks, and conversions sitting in your own Search Console and analytics. This is the traditional-search end of the spectrum, and the cheapest place to start, because the data is already yours.

Search Performance AnalysisFind SEO opportunities in the Search Console data you already have. Read your impressions and CTR, diagnose intent from the gap between them, restructure to match, and measure conversions instead of just rank.Technique

The answer engines#

What the LLMs actually say when a prospect asks, measured by probing them with realistic, conversational questions. This is the AI-answer end of the spectrum. In Rampify these run as Discovery research sessions, and each is its own methodology.

These run inside Discovery (start_discovery), and each probes the same business from a different angle in the buyer's journey:

Category VisibilityDo the answer engines name you when prospects ask category-level questions, before they know any brands? The default first session.Research methodologyComparison ReadinessDo you make the shortlist when a buyer is already narrowing between named competitors?Research methodologyScenario SurfacingAre you recommended when a real problem shows up in natural, first-person language, before the buyer names a category?Research methodologyDeveloper WorkflowWhen an engineer asks a coding agent how to build, do you get suggested, rather than only when buyers ask which tool to pick?Research methodologyNarrative AuditWhat does the model know about you from training alone, with search off, and is it right?Research methodologyMarket OpportunityPoints the lens at the market itself: density, gaps, and the positioning wedge you could own, read through four buyer archetypes.Research methodology

The SERP#

The results page itself: who ranks, which result shapes win, and the AI Overviews now embedded in it. This is the bridge between the first two. The SERP increasingly contains the answer engine, so reading it agentically connects ranking research to answer-engine research in one pass.

  • Agentic SERP research. Coming soon. The agent retrieves and analyzes real search results instead of paying a SERP API, then turns what it finds into opportunities.

New to Rampify? Start with How Rampify Works, or see the MCP tools that power these methodologies.

Common questions