Scale at Speed AI Advisor
An AI advisor built on Felix’s 2Y3X methodology — ten years of agency scaling in an interactive knowledge base. Ask a specific question about your agency’s AI readiness, governance, or strategy. Free. No credit card.
Boston Agency Meetup · Rochambeau · April 29 2026 AI and the structural decisions that matter.
You were at Rochambeau. Here’s the full picture — the frameworks from the postcard, expanded.
The 5 Foundation Stones of AI Adoption
Most founders think AI maturity is about which tools you’re using. It isn’t. It’s about the structural decisions that allow AI to compound — or fail catastrophically — inside your organisation.
- Level 01
Aware
Experimenting individually
Individuals are using AI tools on their own initiative. Productivity is uneven. No visibility, no policy, no shared benefit. The agency is at this level and doesn’t know it yet.
- Level 02
Active
Defined use cases live
Specific use cases agreed: brief writing, research, copy drafts. Some teams using them consistently. But governance hasn’t caught up with usage — and that gap is growing.
- Level 03
Governed
Policies, security & compliance
You have a written AI policy. You know which tools process which data. Client contracts cover AI use. Staff know the rules. Governed doesn’t mean slow — it means safe to accelerate.
- Level 04
Optimised
AI-driven decisions
AI now informs the decisions that matter: resource allocation, client strategy, pricing, hiring signals. Feedback loops exist. This is where competitive advantage becomes real.
- Level 05
Native
AI-first business model
The agency’s model has changed because of AI, not just been improved by it. Delivery, pricing, team structure, and client propositions are rebuilt around AI capability.
- NB: Level 3 is GOVERNED, not ‘Integrated.’
You can’t scale AI without governance. Most founders skip this level and pay for it later.
Most agencies at an event like Boston are actively moving from Level 1 to Level 2 — and figuring out what Level 3 actually requires. That’s the right conversation to be having.
Six ways AI goes wrong in agencies
These aren’t theoretical risks. They’re patterns that show up repeatedly in agencies that moved fast without the governance to support it.
- 01
Shadow AI
Staff using personal ChatGPT, Gemini, or Claude accounts to process client briefs, strategy documents, and competitive data. No visibility, no control. Every brief that crosses a free-tier AI boundary is a potential breach of your client contract — and you won’t know it happened.
- 02
Data Leakage
Client-confidential information entering AI training sets. Most free-tier tools use your inputs to improve their models. Your NDAs were written before this problem existed. They don’t cover it. You’re accountable for data your team is handing to a third party.
- 03
IP and Copyright
AI-generated content with no ownership chain. Client deliverables that may infringe third-party intellectual property. No audit trail. “The AI wrote it” is not a defence — and it’s not an answer your client will accept.
- 04
Compliance gaps
GDPR, CCPA, and SOC 2 implications your legal team hasn’t mapped to your AI stack. If you work with clients in healthcare, financial services, or regulated sectors, the exposure multiplies. Most agencies don’t know what frameworks apply to them.
- 05
No team buy-in
This is the failure point most founders overlook — and the most expensive. Imposing AI on a team that hasn’t shaped the change produces passive resistance, surface-level adoption, and eventual rollback. The fix is the same principle behind the GLT system: get the people who will be affected by the change to set the goals and design the change themselves. AI transformation imposed from the top fails. AI transformation co-designed by the team compounds.
- 06
Seeing AI through the agency’s eyes, not the client’s
The final failure point is a commercial one. Most agencies ask: what can AI help us produce faster, or deliver more cheaply? That’s the wrong question. Your clients aren’t worried about your margins — they’re worried about what AI means for their business. The agencies that win will be the ones that answer that fear directly: not with efficiency gains, but with genuinely new things clients can’t build themselves.
What this means for you
One question to ask yourself.
Every failure point in the list above exists because governance hasn’t been built yet. Shadow AI, data leakage, IP exposure, compliance gaps don’t get fixed incrementally. They get fixed when you reach Level 3 and build the foundation properly. That’s not the finish line. It’s the baseline.
The most useful next step is a clear picture of where you actually are — so the right decisions become obvious.
What to do next.
About
Felix Velarde
Felix Velarde’s first business was one of the first web design agencies in the world. His second became the world’s most awarded digital creative agency. Over 30 years he has founded and led a series of pioneering businesses, co-founded a roll-up, and completed 17 M&A transactions. He has spent the last ten years helping founders scale fast and exit at industry-leading multiples through the 2Y3X programme and as an industry-leading Board Advisor.
Felix speaks on AI transformation, agency growth strategy, and the structural decisions that separate agencies that scale from those that don’t.