diff --git a/e2e/demo.test.ts b/e2e/demo.test.ts
index bc15a4df..4fa120a9 100644
--- a/e2e/demo.test.ts
+++ b/e2e/demo.test.ts
@@ -16,6 +16,10 @@ test('home page explains model-agnostic intelligence and paid Starter pricing',
 	await expect(page.locator('.evaluation-method')).toContainText(
 		'80 paired comparisons across 40 code-navigation tasks'
 	);
+	await expect(page.locator('.evaluation-results')).toContainText('23% higher');
+	await expect(
+		page.getByRole('link', { name: 'TokensToGold methodology and scoring code ↗' })
+	).toHaveAttribute('href', 'https://github.com/noodlbox/tokens-to-gold');
 	await expect(
 		page.getByRole('heading', { name: 'Organization Knowledge', exact: true })
 	).toBeVisible();
diff --git a/src/routes/+page.svelte b/src/routes/+page.svelte
index 2b7c6c4f..a262527c 100644
--- a/src/routes/+page.svelte
+++ b/src/routes/+page.svelte
@@ -350,7 +350,7 @@
 		Higher accuracy. <span class="gradient-text">Fewer calls. Fewer tokens.</span>
 	</h2>
 	<div class="evaluation-results">
-		<div><strong>+12 points</strong><span>file-and-symbol accuracy</span></div>
+		<div><strong>23% higher</strong><span>file-and-symbol accuracy</span></div>
 		<div><strong>14% fewer</strong><span>tokens per task, on average</span></div>
 		<div><strong>14% fewer</strong><span>tool calls per task, on average</span></div>
 	</div>
@@ -362,6 +362,12 @@
 			from 52.2% to 64.3%. Average tokens decreased from 1.01M to 0.88M and tool calls from 25.5 to
 			21.8.
 		</p>
+		<p>
+			Public retrieval-benchmark reference:
+			<a href="https://github.com/noodlbox/tokens-to-gold" target="_blank" rel="noopener noreferrer"
+				>TokensToGold methodology and scoring code ↗</a
+			>. The Context Engine figures above come from our separate agent A/B evaluation.
+		</p>
 	</details>
 </section>