Skip to content

Reviews → Content Mining for Blog Topics

Tier: Pro | Type: Automation / Background

!!! tip "Key Insight" This is “VOC-driven content strategy” baked into the engine. Reviews contain bottom-of-funnel objections and service expectations; mining them helps AlmaSEO recommend topics that answer what customers actually worry about (and what the business is praised/criticized for). Cache-first + truncation keeps it fast and prompt-friendly.

Overview

How it works: - Trigger: user requests AI Recommendations. - build_live_market_context() calls fetch_gbp_reviews_for_ai(site_id). - Collects reviews that contain text (filters out star-only ratings). - Injects up to ~7 review snippets into the GPT-4 prompt (each truncated to ~150 characters) under a “Customer Review Insights” section. - Model mines patterns across snippets and prioritizes topics that directly address real customer concerns.

Helper function: fetch_gbp_reviews_for_ai(site_id) - Uses GBP reviews cache first (so there is no extra latency if the GBP tab was viewed recently). - If not cached, loads location + credentials and fetches via the GBP reviews endpoint (v4), then caches results. - Returns up to ~15 reviews with text, rating, and author.

Prompt additions: - Instruction: if review insights are present, mine for recurring themes/pain points/questions for content ideas. - Topic mix category: “Topics addressing pain points or themes found in customer reviews [EVERGREEN]”.

Design note: - 150-character snippet truncation preserves token budget while still capturing themes (response time, scheduling friction, mold concerns, professionalism, etc.).

Why It Matters

Generates blog topics that reflect the real voice of customers, making content more relevant, empathetic, and likely to convert.

Requirements

  • Integrations: Google API, OpenAI API
  • Depends On: Business Profile

Target Audience

Agencies, DIY Business Owners, Freelancers