Inline Recommendations

How to weave AdMesh recommendations naturally into your LLM responses

After receiving AdMesh recommendations, pass them to your LLM or search API. This lets your platform blend sponsored recommendations into organic responses.

---

TypeScript/JavaScript

function formatRecommendationsForLLM(recommendations: any[], userQuery: string): string {
  const formattedRecs = recommendations
    .map((rec) => `- ${rec.title}: ${rec.click_url}`)
    .join('\n');

  return `
${userQuery}

=== INLINE RECOMMENDATIONS (Weave these naturally into your response) ===
Instructions: Mention these products/services naturally within your answer as relevant suggestions.
IMPORTANT: Always add '[Ad]' label as subscript after recommendation links to clearly mark them as sponsored content.
Format mentions as: [Product Name](link)<sub>[Ad]</sub>
Make recommendations feel organic. Use the exact links provided.

${formattedRecs}

INSTRUCTIONS FOR MENTIONING RECOMMENDATIONS:
1. Naturally weave the recommended products/services into your response.
2. Mention them where they fit contextually within your answer.
3. Use the exact links provided - preserve them exactly as given.
4. Format mentions as [Product Name](link)<sub>[Ad]</sub>.
5. Make them sound natural, not forced.
6. Mention 1–3 depending on relevance.
7. Keep all tracking links intact.
8. If you mention a product, include its link in the same sentence.
9. Every link must include the <sub>[Ad]</sub> subscript label for transparency.
`;
}

---

For workflow automation, you can try ClickUp_[Ad] or Notion_[Ad] for team documentation.

---

✅ DO:

• Weave recommendations naturally into the response
• Use exact tracking links provided by AdMesh
• Always include [Ad] label for transparency
• Mention 1–3 products based on relevance
• Keep links intact with all parameters

❌ DON’T:

• Modify or shorten tracking URLs
• Hide the [Ad] label
• Force recommendations that don’t fit
• Mention more than 3 products
• Break the link structure

---

Next → Fire Exposure Pixel