The Deliverable Fallacy

Tim Ferriss has argued that on some level his wildly successful self-help book The 4-Hour Body was a lookup table: in 2019 it was the best available interface to questions like: "How do I lose fat?" or "How do I fix my sleep?" An AI chatbot whose underlying model has read that book (and thousands of others), and can return a set of personalised recommendations in fifteen seconds. Recent book sales confirm what happens next:

• The Association of American Publishers has adult nonfiction down 7.8% year-on-year in the first quarter of 2026, with self-help the steepest faller at 26.3%.
• Ferriss reports his own catalogue selling at roughly 80% below 2022.

Being the best route to an answer can be very lucrative, until someone comes up with something faster.

Consultancy contains plenty of lookup tables too: regulations, benchmarks, standards, specifications. AI can produce all of those frighteningly cheaply. But were they ever the product? Did a deck containing a list of recommendations actually turn an organisation around, or was it something else? Here is where Ferriss turns, and it is worth following him closely. He sent friends a bullet-point summary of his own book. They thanked him, agreed it all made sense ... and changed nothing. Yet the readers who went through the full book, with its sequencing and all its detours, routinely changed the habits they had failed to shift for years.

The information in his bullet-point summary was identical to the information in the book. The outcome was not. His diagnosis: the book worked because it changed what the reader believed about their own situation before it handed over the solution. The summary delivered the solution without the reframing, so it changed nothing. Skip the build up and the solutions don't work.

In consultancy, the slide deck is the bullet-point summary and the engagement is the long-form book. A consultant's report details their solutions. What makes them stick is the process that produced them: the workshops where staff members argued through tradeoffs, the interviews that surfaced what everyone knew and nobody said. The client organisation changes because it experiences something, not because it receives something. Confuse the two and you have the deliverable fallacy: mistaking the final product for the mechanism that produced it. Edgar Schein described the difference between expert consultation and process consultation: the expert model delivers conclusions, process consultation changes the client's capacity to reach them.

Ferriss makes a seller's point: produce transformation, not information. The buyer, a consultant's client, now has a faster route to the bullet points than ever before. A polished strategy regurgitated by a chatbot will read exactly like the six-week engagement's deck, because people asked to distinguish AI-generated text from human work perform at roughly coin-flip accuracy. It will not produce the change, for precisely Ferriss's reason, and the client will be satisfied right up until the recommendations hit the organisation and nothing changes.

AI is attacking the visible part of consulting, not the valuable part. Commoditising the deck does not touch what mattered, because the deck was never where the real value sat. But it does strip away the disguise. The firms in trouble are the ones who had quietly become deck-factories, selling the summary and hoping nobody noticed the transformation was not included. Most don't do this deliberately. Over time the deck slowly becomes the product because it's the easiest thing to price. The firms that will be fine are the ones whose engagements genuinely change how the client thinks, where the deliverable is the least valuable part of the project. This is the verification gap: the client cannot inspect the thing that matters, so they inspect the proxy - and that bit has just become free to produce.

None of which resolves into advice. If the engagement is the value and the deck is residue, a consultancy firm still has to sell something it can demonstrate. "You are paying for a transformation you cannot inspect in advance" is a brutal pitch against a competitor brandishing a polished AI deck at a tenth of the price. The firms that understand the mechanism are left selling the thing the client never realised they were paying for, against a substitute that looks identical and costs almost nothing.

So, how do you sell a process to a client trained to buy an output? The firms that work this out are not the ones producing better decks - those are freely available without needing a 26 year old with an MBA and a lanyard. They are the ones who make the value of the engagement visible before the sale.