Last Saturday, I had the opportunity to lead a session at MeasureCamp London on a question that is on everyone’s mind in digital analytics right now:
👉 “Superpowers for Digital Analysts: Can Generative AI Really Deliver?”
This wasn’t just an isolated conversation. It built on the lively debate we started at MeasureCamp Paris (June 2025) with the provocative title: “Is Generative AI the End of the Digital Analyst?” 📖 If you missed it, you can read the recap here: https://lnkd.in/eTWmC6cf
At London, we pushed the conversation further — from fear of replacement to opportunity for empowerment.
⚡️ The Energy in the Room
The session kicked off with a simple question to the audience: “Do you believe AI will provide superpowers to Digital Analysts?”
The response was striking: 90% said YES.
This optimism set the tone. Clearly, analysts are not just bracing for change; they’re ready to embrace it. But optimism must also be tempered with realism. Technology promises much — but how do we make sure it enhances, rather than diminishes, our role?
That’s where the discussion got interesting.
🧰 Where AI Already Delivers (KEEP)
The group shared dozens of ways they’re already using Generative AI to enhance their work. These use cases show that AI is moving well beyond hype and into daily practice:
- SEO & Web Performance: mapping websites for SEO issues, surfacing problems faster than manual audits.
- Data Enrichment & Classification: using GPT-powered plugins to categorize, clean, and enrich data.
- Code Writing & Debugging: leveraging ChatGPT, Cursor, and other tools to write, check, or optimize code in JavaScript, SQL, or BigQuery — sometimes in minutes instead of hours.
- Pattern Recognition & Prediction: finding outliers, spotting trends, and even surfacing the “unknown unknowns” hidden in large datasets.
- Customer Insight: digging deeper into user behavior, generating new questions, and framing hypotheses analysts might not have thought of.
- Research: using AI to break down complex documentation and technical references into digestible summaries.
In short: AI is already a copilot — accelerating repetitive or technical tasks, freeing up analysts to focus on higher-order thinking.
🛠️ Where We Need to Get Better (IMPROVE)
Not every use case is mature. Several areas still need refinement before they truly empower analysts:
- SEO Auditing: using GA and Search Console data still requires careful human oversight to avoid misleading recommendations.
- Real-Time Analytics: promising, but accuracy and latency remain challenges.
- Prompting Skills: better prompts = better results; this is a new literacy we all need to master.
- Machine Learning Exploration: GenAI can help generate models and code, but analysts must understand the math and assumptions behind them.
- Repetitive Task Automation: GenAI can become the new “Excel macro” — but only if properly integrated into workflows.
- Tool Awareness: many analysts don’t yet know the full range of AI tools available (e.g., Gamma.app).
Lesson learned: GenAI superpowers don’t just happen. They require training — both of the machine, and of ourselves.
🚀 Where We Should Start (START)
The frontier is wide open. A few participants highlighted new areas where AI could create fresh value:
- Advanced Infrastructure: experimenting with MCP server installation and configuration.
- Exploring Uncharted Territory: testing copilots in areas where analysts haven’t yet applied them — and being open to surprises.
🛑 Where to Hit the Brakes (STOP)
Perhaps the most important part of the session: recognizing what not to do. The group was clear: superpowers don’t mean switching off our brains.
Digital Analysts should avoid:
- Blindly trusting AI without validation.
- Outsourcing their analytical reasoning to algorithms.
- Copy-pasting text, code, or results without checking.
- Asking AI to generate findings or insights without human interpretation.
During the session, a very interesting discussion emerged around the lack of reliability when performing analysis directly through LLMs connected to a data source. The consensus was that LLMs should primarily be used to generate SQL queries, which must then be validated by experts. The validated queries should be executed directly in SQL to avoid potential hallucinations and ensure trustworthy results.
Because here’s the reality: AI can accelerate errors just as quickly as insights. If analysts abdicate responsibility, the superpowers quickly become weaknesses.
🕹️ The New Rules of the Game
The conversation naturally shifted to the future. If Generative AI is here to stay — how will it reshape our profession?
Three emerging “rules of the game” came up:
- Shared Roles: Analysts and AI agents will divide work differently. Some tasks will be delegated, others retained, and new ones will emerge.
- AI as Junior Professional: Think of the AI agent like a junior analyst — quick, eager, but requiring supervision, mentoring, and guardrails.
- AI as Copilot: The most powerful metaphor. Analysts remain in the cockpit, setting direction, validating decisions, and taking responsibility, while AI accelerates execution.
This shift reframes the analyst’s value: less about execution, more about orchestration, creativity, and strategic judgment.
💡 My Takeaway
Generative AI can give Digital Analysts superpowers — but only if we use it wisely.
The real danger isn’t AI replacing us. The danger is analysts replacing themselves by outsourcing judgment, creativity, and curiosity.
The real opportunity? To become the professionals who know when to trust AI, when to challenge it, and how to combine its speed with our own expertise.
The Digital Analyst of the future is not a number-cruncher, nor an AI operator. They are an orchestrator of hybrid intelligence — balancing human insight and machine power to deliver better outcomes.
🤔 Over to You
So what do you think? Is Generative AI your:
- Copilot (helping you fly higher),
- Apprentice (learning under your guidance),
- Or Competitor (fighting for your role)?
I’d love to hear your perspective!
The journey continues! I’ll be leading a follow-up session at MeasureCamp Brussels on November 15, 2025 — join me for more insights and discussions!
Thank you to everyone who joined — your energy and interaction made this session a real success!
