Case Studies — AI-Leverage

01 — Hybrid

Partner, boutique strategy firm · 4 people · New York

The founder: Maren thought she needed "better AI knowledge." She'd read constantly, tried three different GPT wrappers, and still felt like she was falling behind peers.
The problem: Proposals were the bottleneck. Each one took two days of her time — rewriting drafts her team produced, burning the creative hours she needed for client strategy.
What we found: Discovery surfaced that the team's drafts were not the issue. The issue was the founder-held context about each client that only lived in Maren's head. No one could draft well without it.
What we built: A client-context system — structured per-client dossiers, a drafting flow that pulls from them, an AI pattern that produces first drafts at 80% instead of 40%. Four-week Hybrid engagement.
Her week now: Proposals take four hours instead of two days. Maren is back inside client strategy work. The team ships more, the bottleneck is gone.

02 — System Build

Founder, 14-person agency · Remote-first

The founder: David was convinced he needed an "AI strategy." He kept hiring consultants and walking away with decks that didn't change anything in his week.
The problem: His lead pipeline was leaking. Leads came in, got one email, then drifted into the void. Sales was dependent on the founder chasing down every good one personally.
What we found: Discovery revealed that AI wasn't the missing piece — structured follow-up was. The AI move was narrower and more powerful than expected: a multi-touch, per-lead sequence driven by real research the team was already doing and not using.
What we built: A follow-up system that uses existing lead research to produce genuinely personal emails, sequenced over three weeks, with founder sign-off at one point in the sequence. Three-week System Build.
His week now: Pipeline has held for four months. Response rate tripled. David's time is back on client delivery, not inbox triage.

03 — Workflow Upgrade

Independent consultant · Brand strategy · London

The founder: Priya assumed AI would erode what made her work distinctive. She was resisting it on principle — and quietly falling behind younger peers.
The problem: Research synthesis — the heart of her work — was eating 60% of every engagement. Her craft was downstream of it, but the time spent wasn't proportional.
What we found: The opposite of what Priya feared: her taste and judgment were exactly the part AI couldn't replicate. What it could do was compress the synthesis phase from days to hours — without flattening the insight.
What we built: A research workflow that front-loads AI on ingestion and clustering, hands Priya the structured material, and leaves all interpretation to her. No build; just a two-week Workflow Upgrade with a written playbook.
Her week now: Synthesis is now 15% of an engagement, not 60%. She's taking on two more clients than before, and her work is more distinctive, not less.

Three founders. Three different leverage points. Three real changes.