AI as a workflow, not a chat.
The AI competencies a digital marketer needs today and how I use them: generative assistants, prompt design, agent-style work with scheduled automations, the ad platforms' own AI, creative production, data analysis, integrations, SEO in the AI era, chatbots and ethics.
The automated part of a week.
A scheduled task pulls last week's channel-level performance, summarises it and flags anomalies; I add the commentary and send.
A script lists wasteful terms, AI proposes negatives; applied with my approval.
10 variants of the winning creative are generated, screened against brand guidelines, and enter testing.
A form/WhatsApp lead lands in the CRM, gets qualified, a notification arrives; a reply draft is waiting.
Beyond the chat.
The AI competencies a digital marketer needs today: generative assistants, prompt design, agent-style work with scheduled automations, the ad platforms' own AI, creative production, data analysis, integrations, SEO in the AI era, chatbots and ethics. The principle is the same in all of them: AI brings speed, decisions and responsibility stay with me.
Generative AI assistantsChatGPT · Claude · Gemini · Copilot
Large language models for research, copy, analysis and code. I use each for its strengths: long context and file analysis, fast ideation, integration with the Google ecosystem.
- Generating ad copy and landing variants, then human editing
- Summarising reports and datasets, querying anomalies
- Writing code: Apps Script, Python, GTM variables
Prompt design & workflowSystem prompt · examples · chaining
Good output isn't luck: role, context, format and success criteria are written up front; guided with examples; output verified. I keep a prompt library for recurring work.
- Reusable templates (brief, report, analysis)
- Output verification and source checks
- Chained tasks: analysis → recommendation → action list
Agent-style work with ClaudeProjects · Artifacts · Scheduled tasks · Cowork
Beyond typing into a chat: Projects for persistent context and files, Artifacts for report/dashboard/code outputs, Scheduled tasks for routines that run on their own at set times, and Cowork/Claude Code for tasks connected to files, email and tools. I turn these into workflows that run automatically on their own.
- Scheduled generation of the weekly performance summary and report draft
- Search-term / negative keyword cleanup suggestions, competitor tracking
- Lead and report flows via Gmail · Drive · Sheets connectors
AI inside the ad platformsPMax · Advantage+ · AI Max · automated bidding
Google's and Meta's own AI: asset generation, automated targeting, bidding strategies. My job is to feed these systems the right signal and keep hold of the control points (brand protection, exclusions, creative quality).
- AI Max for Search and PMax asset-group strategy
- Advantage+ creative and audience automation; using the limits deliberately
- Conversion signal quality: CAPI, enhanced conversions, value rules
Creative production with AIMidjourney · Firefly · Canva · Runway · ElevenLabs
AI tools for image, video and audio: fast prototypes for multi-variant creative testing, backgrounds/scenes for product images, short video and voice-over. Human approval is a must for brand consistency and rights.
- 10+ variants for the weekly creative testing loop
- Product/scene images, vertical video, voice-over
- Brand guideline compliance and usage-rights checks
Data analysis & reportingSheets/Excel + AI · GA4 Insights · Looker · Python
Cleaning large exports, writing formulas, cohort and anomaly summaries, natural-language querying. Cutting report-writing time to spend it on interpretation.
- Clustering search-term and product performance exports with AI
- GA4 explorations and automated insight checks
- Coding recurring analyses with Python/Apps Script
Automation & integrationZapier · Make · n8n · Google Ads Scripts · Apps Script
Leads landing in the CRM, WhatsApp/email notifications, budget alerts, automatic negative keyword lists. By putting AI inside these flows I automate the decision steps too.
- Lead → CRM → notification flows
- Budget/anomaly alerts with Google Ads Scripts
- Moving data between platforms via webhooks
SEO in the AI eraAI Overviews · GEO/AEO · structured data
As AI summaries and answer engines grow in search results, the rules of visibility change: clear answers, expertise signals, structured data, quotable content. Content production with AI, edited by humans.
- Question–answer-driven content architecture
- Structured data and quotability
- AI draft, expert edit, then publish
Chatbots & communicationWhatsApp · Meta · site assistants
AI-powered bots for lead qualification, appointment booking and instant answers to common questions; with human handover. Directly affects conversion rates in services, education and healthcare.
- WhatsApp lead flow and pre-qualification
- FAQ and routing with a site assistant
- Human handover and logging
Ethics, privacy & verificationGDPR/KVKK · hallucination · brand safety
Which data goes to AI and which doesn't; how outputs are verified; how client data is protected. AI brings speed, it doesn't remove responsibility.
- Data-sharing boundaries and anonymisation
- Output verification checklist
- Measurement compliant with consent management
All the tools are on the Digital Marketing page.
Google, Meta, LinkedIn, TikTok, Pinterest, X, Microsoft, programmatic and SEO tools; what each is and how I use it.