The $500K deck trap why most mid-market AI projects die before shipping.

The anatomy of the deck trap

Every quarter, mid-market companies across Europe and the US pay a consultancy a serious budget to produce a document. The document is called an “AI strategy” or an “agentic transformation roadmap.” It is, almost always, a PDF. It circulates for three to six months. At the end of those months, no agent is running in production.

The deck itself is not the problem. What it replaces is: decisions. A sixty-page strategy document usually means the organization still has not decided who owns the first workflow, which system the agent will read from, and who gets woken up at 2 AM when the agent does something unexpected.

I have watched this cycle as the operator hired to ship after the deck fails. The pattern holds. The failure is organizational, not technical.

The agent that replaces the deck usually takes six weeks and costs a fraction of what the deck did. That is the median case I see, not a one-off. Companies that ship agents fast do not skip strategy — they compress it. A two-page scope. A named owner. A deadline that means something. And a team allowed to fail fast on a narrow wedge before anyone asks them to transform the whole department.

Four decisions that separate shippers from stalls

Across the production engagements I have run, the pattern is clear. Teams who ship production agents in weeks (not months) make four decisions early that stalled teams never make at all.

1. Name the workflow, not the department

A good agent scope is “automate the invoice-to-PO matching workflow in Procurement” — not “bring AI to Procurement.” Teams that ship name a specific workflow in the first week. Teams that stall produce a department-wide assessment and call it progress.

2. Assign a production owner before writing a line of code

An agent without a production owner is a demo. Someone inside the client organization needs to be on the hook for what happens when the agent runs at 2 AM on a Saturday and does something wrong.

3. Set a six-week shipping window

Six weeks is not a magic number. It is a forcing function. A six-week timeline makes a four-month discovery phase impossible. It forces a scope narrow enough to deliver, test, and iterate inside one cycle. Every engagement where we have shipped successfully used this constraint.

4. Kill the staging environment mindset

Agents do not work in staging. They work in production, with real data, real users, and real consequences. Companies that ship fast design for production from day one — monitoring, guardrails, human-in-the-loop fallbacks. Companies that stall spend months in a sandbox and then discover their agent cannot handle the one data format their ERP actually produces.

What this means for your next pilot

If you are about to start an agentic AI pilot: do not start with a deck. Start with a decision. Name the workflow. Name the owner. Set the clock. Build for production from the first commit.

The $500K deck trap is real, and it is not malicious — the consultancies producing those decks believe in them. The distance between a document and a deployed agent is measured in decisions, not pages. Teams that ship understand this. Teams that do not produce beautiful PDFs.

If this resonates and you want to talk through your situation, I take a limited number of 30-minute calls each week. No pitch, no deck — just the conversation this essay opened.