The Problem with Most AI Engagements
Most consulting engagements addressing AI transformation follow one of three patterns, and all three produce disappointing results.
The Strategy-Only Engagement. A consulting firm conducts extensive discovery, interviews dozens of stakeholders, and produces a comprehensive transformation roadmap. The document is excellent. Eighteen months later, it sits on a shelf while the organization has moved on to its next priority. Nothing was implemented.
The Tool Deployment Engagement. A technology vendor deploys an AI tool. Usage metrics are tracked. Adoption is modest. The tool works as designed, but the processes it touches haven't been redesigned to take advantage of it. ROI is marginal.
The Training Engagement. An AI training program improves individual digital literacy. Staff learn about AI tools and capabilities. But without changes to workflows and without deployed systems to use, the knowledge isn't applied. Skills decay. Nothing changes.
The Remolda Cycle was designed specifically to avoid all three failure modes.
Why Five Phases
The five phases of the Remolda Cycle correspond to five distinct challenges that every successful AI transformation must address. Skip any one of them and the transformation is incomplete.
Phase 1: Audit. You cannot design a transformation without an honest picture of where you are starting. The Audit phase creates this picture — not the idealized version that appears on org charts, but an accurate assessment of actual workflows, data quality, talent capabilities, and organizational culture. Many transformations fail because they were planned against an inaccurate baseline.
Phase 2: Strategy. A list of AI opportunities is not a strategy. A strategy designs the target operating model — how the organization should function when AI is embedded — and then maps the path from current to future state. The Strategy phase produces this map, with realistic timelines, investment modeling, and change management integration.
Phase 3: Implement. Strategy without execution is a document. The Implement phase is where AI systems are deployed — not in a single big-bang launch, but in structured waves that build confidence, validate assumptions, and generate the early wins that sustain organizational commitment.
Phase 4: Empower. Technology that people don't use is wasted investment. The Empower phase rebuilds competency across the organization — not through generic training, but through systematic capability development tied to the specific tools and workflows being deployed.
Phase 5: Evolve. AI transformation is not a project with an end date. AI systems degrade, organizational needs change, and new AI capabilities emerge continuously. The Evolve phase creates a sustainable operating model for continuous improvement.
The Timing Design
A feature of the Remolda Cycle that is non-obvious but important: Phase 4 (Empower) runs in parallel with Phase 3 (Implement). This is deliberate.
The most common failure in AI training programs is that training is delivered before the tools are available or after the tools have already been deployed without support. Neither timing works. Training before tools gives staff knowledge without application. Training after deployment without support produces frustration and abandonment.
Empower running in parallel with Implement means staff receive capability development exactly when they need it — as each new AI workflow goes live in their area.
What This Looks Like in Practice
A typical Remolda engagement for a mid-sized organization runs 12–18 months from Audit kickoff through the first Evolve quarterly review.
The first 6–8 weeks are the Audit and Strategy phases — establishing the honest baseline and designing the transformation roadmap. The following 9–12 months are parallel Implement and Empower phases — deploying AI capabilities wave by wave while building organizational competency. The final phase transitions to ongoing Evolve engagement on a quarterly cadence.
This is not the fastest possible timeline. It is the timeline that produces durable transformation rather than impressive pilots that don't scale.