Most marketing teams are paying for tools they underuse and skipping tools that would move the needle. Here's a transparent look at the exact stack we use — and why each piece earns its seat.
Every marketing agency in 2025 claims to be "AI-powered." Almost none of them explain what that actually means in practice. We think that opacity is a problem — for clients trying to evaluate what they're paying for, and for the industry trying to understand what's genuinely new versus what's just rebranding.
So here's our stack. Every tool we use, what we use it for, and where we've found it to be genuinely transformative versus where it disappoints.
Claude is our primary reasoning model for strategy work. We use it for brief synthesis, strategic scenario planning, audience psychology profiling, and first-draft positioning frameworks.
Why Claude over GPT-4o for strategy? The reasoning depth. Claude produces outputs that are more willing to say "I'm uncertain about this" and more capable of holding complex multi-dimensional problems without collapsing them into oversimplified answers. For marketing strategy, where nuance matters, that disposition is more valuable than confident-but-wrong answers.
Where it delivers: Synthesizing complex research into a clear strategic narrative. Pressure-testing assumptions against a structured rubric. Writing that sounds like a senior strategist rather than a content mill.
Where it disappoints: Real-time information. For anything requiring current market data, we route to Perplexity first, then bring the synthesized intelligence into Claude for strategic interpretation.
Before any strategy session, we run structured Perplexity queries to map the competitive landscape, identify recent market movements, and surface signals that haven't made it into training data.
Perplexity's value isn't search — it's synthesis with sourcing. We can ask "What are the 5 dominant positioning angles in the enterprise security software category as of Q2 2025?" and get a structured, sourced answer in 90 seconds instead of 90 minutes of desk research.
Where it delivers: Competitive intelligence, recent press coverage synthesis, category trend mapping, pricing landscape research.
Where it disappoints: Nuanced interpretation. Perplexity gives you the facts. Claude gives you what the facts mean for your strategy. They're complementary, not interchangeable.
For high-volume creative work — headline variation, ad copy testing matrices, email subject line generation, social post calendars — GPT-4o is our workhorse. It's fast, produces consistent output quality, and handles structured prompts with reliable formatting.
We use GPT-4o with a custom prompt library we've developed over the last 18 months. Not generic prompts — structured frameworks that encode our messaging methodology directly into the prompt architecture, so the output is already aligned with our strategic approach before any human edits it.
Before a designer opens Figma, we run creative direction through Midjourney. This isn't about replacing design — it's about eliminating the blank canvas problem and dramatically reducing the cost of exploring creative directions that don't work.
A mood board that used to take a designer 4 hours to build (finding references, creating layouts, iterating) takes 45 minutes with Midjourney. The designer's time goes to executional refinement, not exploratory ideation. The result: more creative directions explored, more informed creative decisions made, fewer expensive late-stage pivots.
Semrush for data, Claude for interpretation. We pull keyword opportunity data, content gap analysis, and backlink landscapes from Semrush, then synthesize the strategic implications using Claude. The output: an SEO strategy that's grounded in data but expressed as a clear editorial narrative rather than a spreadsheet dump.
The tools above collectively save us 60–70% of the research and production time in a standard campaign cycle. That's meaningful. But it doesn't change what actually requires human judgment:
We're an AI-native agency because AI genuinely changes what's possible for our clients. Not because AI does everything. Because AI does the right things — and that frees us to do what machines can't.
The 30-day brief process was built for a different competitive environment. Here's how modern marketing teams are compressing insight-to-execution cycles — and what it means for pipeline velocity.
When acquisition, content, and conversion operate in isolation, growth plateaus — regardless of budget. Here's the integrated demand generation architecture that changes the equation.
Ready to apply this thinking to your business?
Apply for a founding spot
