diff --git a/surfsense_web/components/homepage/compare-table.tsx b/surfsense_web/components/homepage/compare-table.tsx
index 85ad84277..856170a00 100644
--- a/surfsense_web/components/homepage/compare-table.tsx
+++ b/surfsense_web/components/homepage/compare-table.tsx
@@ -1,36 +1,61 @@
 import { Reveal } from "@/components/connectors-marketing/reveal";
 import { MarketingSection } from "@/components/marketing/section";
 
+const COLUMNS = [
+	{ label: "Browser agents", examples: "Browserbase, Browser Use" },
+	{ label: "Scraping APIs", examples: "Firecrawl" },
+	{ label: "Search APIs", examples: "Exa, Tavily, Parallel" },
+	{ label: "Scraper marketplaces", examples: "Apify" },
+];
+
 const ROWS = [
 	{
-		feature: "Who does the work",
-		suites: "Analysts read human dashboards",
-		scrapers: "You build everything on raw data",
-		surfsense: "Your AI agents gather, analyze, and alert",
+		feature: "Built for",
+		browser: "Web tasks that need clicking, logins, and forms",
+		scraping: "Turning individual pages into LLM-ready content",
+		search: "Finding and reading pages about a topic",
+		marketplace: "Thousands of community-built scrapers, one per site",
+		surfsense: "Live platform data as research primitives for agents",
 	},
 	{
-		feature: "Intelligence layer",
-		suites: "Curated but slow, human-in-the-loop",
-		scrapers: "None; data only",
-		surfsense: "Agent harness turns live data into briefs",
+		feature: "How retrieval works",
+		browser: "An LLM drives a real browser, page by page",
+		scraping: "Fetch a URL, get markdown or schema-extracted JSON",
+		search: "Query an index, get ranked results and page content",
+		marketplace: "Pick an actor per site, learn its input, run it",
+		surfsense: "One typed REST call per platform, no LLM in the retrieval loop",
+	},
+	{
+		feature: "Platform data (comment trees, transcripts, reviews)",
+		browser: "Whatever the LLM extracts from rendered pages",
+		scraping: "Page-level extraction; social platforms aren't the focus",
+		search: "Page text and snippets, not structured platform items",
+		marketplace: "Yes, but schema and quality vary per actor",
+		surfsense: "Native items: posts, comment trees, transcripts, reviews, SERPs",
+	},
+	{
+		feature: "Consistency",
+		browser: "Depends on the model and the page",
+		scraping: "One API, you define schemas per page type",
+		search: "One API, snippet and page-content shapes",
+		marketplace: "A different schema and quality bar per actor",
+		surfsense: "One API and one schema style across every connector",
+	},
+	{
+		feature: "Research workspace & knowledge base",
+		browser: "No",
+		scraping: "No",
+		search: "No",
+		marketplace: "No",
+		surfsense: "Cited briefs, knowledge base, scheduled automations, deliverables",
 	},
 	{
 		feature: "Pricing",
-		suites: "Enterprise contracts, annual quotes",
-		scrapers: "Usage-priced infrastructure",
-		surfsense: "Self-serve, pay per item, free tier",
-	},
-	{
-		feature: "Developer surface",
-		suites: "No agent surface",
-		scrapers: "APIs, but no MCP or harness",
-		surfsense: "REST API with your API key, plus an MCP server for agents",
-	},
-	{
-		feature: "Open source",
-		suites: "No",
-		scrapers: "No",
-		surfsense: "Yes, self-hostable",
+		browser: "Per browser minute (1-minute session minimum) plus the LLM tokens driving it",
+		scraping: "Credits per page; schema extraction costs extra credits",
+		search: "Per search request",
+		marketplace: "Per event or result, set by each actor",
+		surfsense: "Per item returned; failed calls never billed",
 	},
 ];
 
@@ -40,24 +65,23 @@ export function CompareTable() {
 			<Reveal>
 				<h2 className="text-2xl font-bold tracking-tight sm:text-3xl">How SurfSense compares</h2>
 				<p className="mt-3 max-w-2xl text-muted-foreground leading-relaxed">
-					CI suites sell dashboards to analysts. Scraper APIs sell raw data to engineers. SurfSense
-					gives AI agents the data and the harness in one open-source platform.
+					SurfSense is the only open-source product that combines a NotebookLM-style research
+					workspace for people with live-data primitives for agents. Here is how that stacks up
+					against each class of tool.
 				</p>
 			</Reveal>
 			<Reveal>
 				<div className="mt-8 overflow-x-auto rounded-xl border bg-card">
-					<table className="w-full min-w-2xl text-sm">
+					<table className="w-full min-w-4xl text-sm">
 						<thead>
 							<tr className="border-b bg-muted/40 text-left">
 								<th className="p-4 font-medium">Feature</th>
-								<th className="p-4 font-medium text-muted-foreground">
-									CI suites
-									<span className="block text-xs font-normal">Crayon, Klue</span>
-								</th>
-								<th className="p-4 font-medium text-muted-foreground">
-									Scraper APIs
-									<span className="block text-xs font-normal">Bright Data, Apify</span>
-								</th>
+								{COLUMNS.map((col) => (
+									<th key={col.label} className="p-4 font-medium text-muted-foreground">
+										{col.label}
+										<span className="block text-xs font-normal">{col.examples}</span>
+									</th>
+								))}
 								<th className="p-4 font-medium text-brand">SurfSense</th>
 							</tr>
 						</thead>
@@ -67,8 +91,10 @@ export function CompareTable() {
 									<th scope="row" className="p-4 text-left font-medium">
 										{row.feature}
 									</th>
-									<td className="p-4 text-muted-foreground">{row.suites}</td>
-									<td className="p-4 text-muted-foreground">{row.scrapers}</td>
+									<td className="p-4 text-muted-foreground">{row.browser}</td>
+									<td className="p-4 text-muted-foreground">{row.scraping}</td>
+									<td className="p-4 text-muted-foreground">{row.search}</td>
+									<td className="p-4 text-muted-foreground">{row.marketplace}</td>
 									<td className="bg-brand/5 p-4 text-foreground">{row.surfsense}</td>
 								</tr>
 							))}