01
Marketing
AI drafts a campaign in an afternoon. Brand review: 11 business days. Legal: another 8. Design slot: three sprints out. The tool is not the bottleneck. The approval chain is.
Slide 01
Every executive conversation starts with "which tool is fastest?" That is the wrong question. You are standing in Marseille — a city with streets laid out over centuries by people who never imagined your vehicle — and asking which car has the best zero-to-sixty.
Slide 02
Speed and agility. Get the pad thai there hot. You do not need a CDL or a dock. You need to move fast on narrow streets.
Volume, a dock, patience for I-465. The vehicle matches the cargo. Every executive gets this part of the logic right.
Across town in thirty minutes. Maybe a rocket car, if they made one street-legal. Pure speed, minimal cargo.
"Which tool is the fastest? Which one scored highest on SWE-bench? Which one has the most impressive demo?" You are asking which car has the best zero-to-sixty on streets that were designed for donkeys.
The fastest form of ground transportation depends on the cargo. That logic is incomplete. It also depends on where you are operating.
Slide 03
The streets curve without warning, narrow until two walls brush your mirrors, dead-end into staircases, and angle at intersections that make no geometric sense. They were never planned. They emerged over centuries from foot traffic and stone walls.
Marseille's streets were designed a thousand years before the internal combustion engine existed. Ground transportation — no matter how advanced or powerful or expensive — cannot overcome streets that were never built for it.
Slide 04
The AI tool did its job. Fast. Impressive. Exactly what the vendor promised in the demo.
Approvals, reviews, environment provisioning, security scan queue, two rounds of manual QA. The streets.
They optimized the only part that was already fast and left 97% of the cycle time untouched. This is not unusual.
Your approval processes were designed in 2014. Your team topology assumes one developer writes 200 lines per day. Your deployment pipeline has seventeen gates because someone got burned in 2019 and you never removed the checkpoint.
Those are your streets. And they were designed before the vehicle you just bought existed.
Slide 05
AI drafts a campaign in an afternoon. Brand review: 11 business days. Legal: another 8. Design slot: three sprints out. The tool is not the bottleneck. The approval chain is.
Knowledge base requires four sign-offs to update. Your agents are reading answers written before the product shipped its last three releases. The streets are stale.
Demand forecasting model gets rebuilt once a quarter because the data pipeline requires a ticket, a sprint slot, and a deployment window. Different department, same medieval streets.
Slide 06
These gates catch real problems. They exist for good reasons. You redesign them for speed, but you do not remove them.
The load-bearing gates need to be redesigned for speed. The scar tissue gates need to be removed. Knowing the difference is the entire job.
Slide 07
I asked one question: how long does it take a change to get from your best engineer's laptop to production? He did not know.
The tool was generating code faster than the organization could absorb it. Nobody noticed because they were measuring the wrong thing.
Slide 08
Every executive still running evaluations and satisfaction surveys 18 months later chose a tool. The entire tool evaluation process is itself a form of organizational procrastination. It has a spreadsheet and a timeline. It feels productive.
Every executive who got real results chose a partner over a tool. Choosing a partner requires admitting you need help with something harder than a purchasing decision.
Slide 09
Fully loaded cost of $200,000 per engineer per year. Salary, benefits, tooling, overhead.
Your total engineering investment. This is before you add marketing, operations, and customer service teams.
60% of cycle time is wait time in approval queues and handoffs. That 60% is conservative based on measured engagements.
Your competitor who fixed their streets six months ago is spending the same payroll on two to three times the output. That gap compounds every quarter. In 18 months, the difference is not 20% or 30%. It is multiples.
The compounding is the part that boards underestimate. Every quarter you spend evaluating tools instead of fixing streets is a quarter your competitors are pulling ahead.
Slide 10
Pick one workflow that matters. Map every step from idea to live. Mark which steps are work and which are waiting. The ratio is usually 15-20% work, 80-85% waiting. If your number is better, congratulations. If it is not, you now know where the problem is.
For every approval, review, or handoff: what does this gate catch? If nobody can name a specific incident in the last two years, it is scar tissue. Could it be automated, parallelized, or reduced from a committee to a single owner?
Some gates exist because the people who operate them derive authority, headcount, or budget from their existence. Removing the gate threatens someone's role. That is not cynical. That is organizational physics. You cannot fix the streets without understanding who built their house on the current road.
Slide 11
Those people need a role in the new design, not a pink slip. They are subject matter experts who know why the gate was built. They will tell you which gates are load-bearing faster than any audit will.
Streets are not just process. They are people. The pattern is the same every time: the engineers who cannot use AI agents do not have a tools problem. They have a streets problem, and the streets are made of organizational decisions that someone benefits from preserving.
Slide 12
If you do not know the answer, it does not matter what you buy. You are going to end up like me in Marseille, gently crunching into a 400-year-old wall in the most advanced vehicle the rental company had to offer, wondering why the car did not save you.
Have you mapped your wait time? Do you know which gates are load-bearing and which are scar tissue? Do you know who benefits from the current design?