I use AI as a thinking and research layer across my entire product marketing workflow. Claude handles competitive research synthesis, first-draft content creation, and pattern analysis across market data. Fireflies captures and structures meeting insights so nothing falls through the cracks. n8n automates the repetitive distribution work that used to eat 5-10 hours a week. The result: roughly 60% less time on research and first drafts, and significantly more time on strategy, stakeholder conversations, and the judgment calls that actually move the business.

Workflow-by-Workflow Breakdown

1. Competitive Analysis

Before: Manually scanning competitor websites, G2/Capterra reviews, press releases, and LinkedIn posts. Synthesizing patterns across 5-8 competitors into usable battlecards. This took 3-4 full working days per competitor set, and the output went stale within weeks.

Now: Claude ingests competitor positioning from websites, analyst reports, product changelogs, and review site data. It identifies messaging patterns, feature gaps, and positioning shifts. I guide the analysis with targeted questions and validate against what I hear from sales and customers.

Full competitive landscape: 3-4 days → 6-8 hours. Individual battlecard: 2 days → 4 hours. Quarterly refresh: 1.5 days → 3 hours.

2. Content Creation

Before: Staring at a blank page. A single case study could take 3-5 days from interview to polished draft.

Now: Fireflies transcribes customer interviews and sales calls. I feed the transcript and key data points into Claude with a specific brief. Claude produces a structured first draft in minutes. I then rewrite for voice, verify all claims, and add the narrative arc that makes the story compelling rather than just accurate.

Case study: 3-5 days → 1-1.5 days. One-pager: 1 day → 2-3 hours. Sales deck narrative: 1.5 days → half a day.

3. Sales Enablement

Before: Sitting in on sales calls or reading CRM notes to understand objections. Building guides manually. Distributing via email and hoping reps read them.

Now: Fireflies records and transcribes calls automatically. I use Claude to analyse patterns across multiple transcripts: top objections, how best-performing reps handle them. This produces data-backed enablement material rather than anecdote-based content. n8n handles distribution to the right Slack channels and CRM fields.

Objection analysis (20+ calls): 2 days → 3-4 hours. Talk track creation: 1 day → 3 hours. Distribution: 2-3 hours → automated.

4. Campaign Planning and GTM Strategy

Before: Building campaign briefs from scratch. A full brief might take a week of focused work.

Now: I use Claude as a sparring partner. I feed in positioning, persona details, and business objectives, then work through the campaign structure iteratively. Claude helps stress-test messaging angles, generate A/B test hypotheses, and draft channel-specific copy variants. GoodNotes captures daily thinking, which I export to Claude for synthesis.

Campaign brief: 5 days → 2-3 days. Messaging variants: 1 day → 2-3 hours. Email sequence (5-7 emails): 2 days → half a day.

5. Market Research and Positioning

Before: Reading analyst reports, conducting interviews, surveying the landscape — often 2-4 weeks for a proper positioning exercise.

Now: Claude helps process and synthesise large volumes of research. I use it to run positioning exercises based on the April Dunford methodology. Claude does not make the positioning decisions. It helps me organise inputs, identify patterns, and draft the positioning document so I can focus on judgment calls.

Research synthesis: 1-2 weeks → 3-5 days. Positioning document: 3 days → 1 day. Persona development: 2 days → half a day.

Tool Stack

Claude

Thinking & research layer

Research synthesis, first drafts, competitive analysis, positioning exercises, document creation

Fireflies

Capture layer (meetings)

Call transcription, meeting summaries, action items, pattern analysis across calls

n8n

Automation layer

Content distribution, CRM updates, Slack notifications, recurring report generation

GoodNotes AI

Capture layer (daily thinking)

Daily priorities, brain dumps, brainstorming, idea capture with iPad + Apple Pencil

GoodNotes captures raw daily thinking. Fireflies captures meeting data. Both feed into Claude for synthesis and document creation. n8n handles the "last mile" of distributing outputs. The human judgment sits between capture and output: deciding what matters, what the strategy should be, and what the final message needs to say.

What I Don't Use AI For

Strategic positioning decisions. AI organises competitive data and drafts options, but the positioning decision requires understanding company strategy, sales capabilities, and competitive dynamics that AI doesn't have access to.
Relationship building and stakeholder management. Getting a VP of Sales to champion a new battlecard format requires reading people and navigating organisational dynamics.
Final messaging approval. AI generates 10 messaging variants in minutes. Choosing the right one requires understanding brand voice, customer sentiment, and what the sales team can deliver on.
Customer and prospect conversations. Interviews, discovery calls, and relationship-building require empathy and the ability to follow unexpected threads.
Ethical and legal judgment. Claims validation, compliance review, and decisions about what we can and cannot say require human accountability.

Overall Impact

50-70% reduction in time spent on research and drafting work, freeing 10-15 hours per week for strategic and relationship work that actually differentiates a PMM.

The biggest shift isn't speed — it's depth. Before AI, I had to choose between being thorough and being fast. A competitive analysis could be comprehensive or it could be done this week, but not both. Now the quality ceiling went up because I'm no longer cutting corners on research to meet deadlines.