deep-marketing-analyst

What This Skill Does

The deep-marketing-analyst skill is a high-level strategic engine designed for modern performance marketing. It functions by transforming raw advertising performance data from platforms like Meta, Google, TikTok, Amazon, and programmatic DSPs into structured strategic insights. Instead of simply reporting metrics, the skill focuses on hypothesis testing, evidence mapping, and the synthesis of actionable growth experiments. It acts as an objective analyst that evaluates conflicting campaign data, assesses the strength of evidence, and provides concrete recommendations that respect the unique nuances of each ad platform—such as creative-led testing on Meta/TikTok versus intent-based optimization on Google/Amazon.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/danyangliu-sandwichlab/deep-marketing-analyst Ensure your OpenClaw agent has the necessary API access configured for your advertising platforms to enable real-time data ingestion.

Use Cases

• Budget Reallocation: Determining if budget should shift from high-intent Google Search campaigns to top-of-funnel TikTok creative experiments based on ROAS parity.
• Creative Performance Audits: Analyzing why video assets are performing differently across Instagram Reels and YouTube Shorts to inform future asset production.
• Funnel Bottleneck Identification: Investigating drops in conversion rates for Amazon Ads during peak shopping events vs. standard operating windows.
• Strategy Validation: Running hypothesis testing to see if influencer-led content truly lowers CPA compared to static product imagery in your target demographics.

Example Prompts

1. "We have a 48-hour window to optimize our Q4 budget. Analyze our Meta and Google Ads performance from the last 14 days and tell me where to shift spend to maximize revenue while keeping CPA under $40."
2. "My TikTok ROAS is dropping despite stable creative. Test the hypothesis that our current creative fatigue is due to audience saturation and propose a three-stage testing plan for the next 7 days."
3. "Evaluate the current performance of our Amazon Sponsored Products vs. DSP campaigns. Is our brand conquesting strategy cannibalizing our own organic traffic?"

Tips & Limitations

• Data Integrity: The quality of the analysis is directly proportional to the accuracy of your tracking pixel and attribution modeling. Ensure your data streams are clean.
• Platform Specificity: The skill is aware of platform-specific constraints; always include your primary KPI (e.g., ROAS, CPA, or CAC) in your prompt to get the most relevant output.
• Guardrails: Always verify high-spend recommendations. While the skill includes stop-loss conditions in its analysis, final authorization for significant budget changes remains with the human stakeholder.
• Complexity: For the best results, provide as many optional fields as possible (e.g., confidence_target or excluded_assumptions) to narrow the scope of the analysis.