Marketing & growth executive
Michael Preuss
I lead growth for AI infrastructure and technical B2B companies, with positioning and brand as core drivers of demand. I get the programs working, hands-on or with the existing team, then hire people to run and scale what we’ve proven.
Explore by lens
Selected writing and work
Writing
Known, Trusted, Loved, Chosen
2026-10-28 · 5 min
AI is making brand more valuable. Product, proof, language and service shape whether people know us, trust us, love us and choose us.
Brand is demand
Signal
- S1
As features converge, brand is demand. The product, the proof, the language and the service all shape whether someone chooses us.
- S2
Trust and love work together, and both contribute to that first choice. We have to earn it before we can become someone's default.
Work
Blaxel / AI Marketing Systems
Joined June 2026 · Head of Growth & Marketing · 2 min
The marketing tools I built at Blaxel: paid media review, market research, answer-engine citation analysis and a shared go-to-market knowledge grap…
Daily marketing reviews
Writing link
Signal
- S1
I connected ad delivery, product activity and payment outcomes in a daily review. Proposed campaign changes still required human approval.
- S2
A shared go-to-market knowledge graph kept the tools working from evidence and review rules.
Proof points
- P1
At Blaxel, paid media review moved from monthly to daily.
Writing
You Don’t Need Better Prompts. You Need a Knowledge Graph.
2026-03-09 · 23 min
A practical guide to why prompting is a tactic, context is infrastructure, and a knowledge graph is how you stop GTM drift.
Context infrastructure
Signal
- S1
When company context is missing, the model guesses. A shared source of truth gives each session the same positioning, evidence and constraints.
- S2
Versioning that context lets the team review a correction once and reuse it across future work.
Work
Blaxel / Paid Growth & Measurement
Joined June 2026 · Head of Growth & Marketing · 2 min
How I ran paid acquisition at Blaxel, connecting campaign decisions to product activity and payment outcomes.
Acquisition efficiency
Signal
- S1
The review connected Google Ads delivery with PostHog product activity and Stripe payment outcomes.
- S2
I reported overall product activity separately from paid-attributed results.
Proof points
- P1
At Blaxel, paid-ad conversions rose 46%, spend fell 36% and cost per paid-ad conversion fell 56%, versus the prior 30 days.
Work
Blaxel / Customer Stories & Cookbooks
Joined June 2026 · Head of Growth & Marketing · 2 min
How I turned Blaxel customer architecture into technical stories and companion GitHub cookbooks that developers could inspect, run and adapt.
Technical customer proof
Writing link
Signal
- S1
Each story explained the customer's architecture and what Blaxel operated. Selected companion cookbooks isolated a mechanism developers could reuse.
- S2
I prepared customer and company distribution copy, approvals and release waves alongside the technical drafts.
Proof points
- P1
The companion examples demonstrate files surviving sandbox deletion and code being built, checked and repaired in a retained environment.
Writing
Babe, wake up - it's 3/6/9 weeks now.
2026-09-09 · 6 min
The 30/60/90-day plan, updated for marketing leaders working with AI.
Leadership transitions
Signal
- S1
I propose a 3/6/9-week transition: assess the business, make choices and prove one, then establish an operating rhythm the team can run.
- S2
Use the time AI gives back to learn faster and validate decisions. Trust still takes repetitions.
Related work and writing
Writing
The Lock-In Is Moving
2026-05-11 · 3 min
AI tools can hold the reasoning behind your team's decisions. What happens when you switch, and what does that mean for customer retention?
Retention through value
Signal
- S1
Switching AI tools doesn't automatically bring the team's accumulated decisions, corrections and reasoning with it.
- S2
If customers could take their context to a competitor, what would they still choose us for?
Writing
Be the Reference
2026-02-16 · 4 min
Answer the questions that decide purchases, check what AI answers say about you, and use a weekly review to decide what to improve.
Being found in AI answers
Signal
- S1
Answer the questions that decide purchases, then check whether AI answers include the company and represent it accurately.
- S2
At Blaxel, I built a tool that turned answer-engine citation data into a weekly queue of Reddit conversations to review.