ESCAPE THE DECK
Issue #1 · Thursday 2 Jul · Consulting to building · Every two weeks
THE OPEN
Welcome to the first edition of Escape the Deck, and thank you for even reading this!
So here is the start of it. This is issue one, and Escape the Deck exists to work through one shift: what AI is actually doing to consulting, and what to do about it. No hype, no theatre, no slide that ends at "recommendations" and walks away. Every other Thursday, one honest read.
Let me start by sharing the idea the whole thing hangs off.
THE MAIN THING
AI has not killed consulting. It has killed the part of it that was only ever a nicely formatted opinion
For most of my career, what I saw happening, too many times, was that the hard part, the delivery, got left to someone else. You were paid for the thinking, and whether it ever worked in the real world was somebody else's problem to carry. What has changed is not the quality of the advice, it is the distance between a good idea and a working version of it. That distance has collapsed. When a client can get a credible first answer out of a model in an afternoon, the recommendation on its own looks like a thin offer.
So the premium moves to the people who can go past the diagnosis: sit inside the business, redesign the workflow, build the thing, own the value case and stay long enough to make the change stick. Not the cleverest deck, the owned outcome. I have started calling that capability deployment muscle, and I think it is the whole game now.
The full piece, including why agentic AI is the reason this is happening, what the UK's Guardian newspaper calls the "slow death of the prestige career", and what it all means depending on where you sit, is the first article on the new site.
THE OPERATOR'S MOVE
Pick one thing you currently sell, propose, or ask for. One.
If you are an employee, think about a proposal you are about to send
If you run the business, pick one of the service lines or propositions on your corporate site
If you’re a client, think of a scope of work you keep agreeing to.
Now, rewrite it around two things:
First, one owned outcome with a number attached: not "improve the process" but "cut the cycle time from X days to Y, and we are accountable for getting there".
Second, one place where AI visibly does the work, not as a feature you name but as the reason you can promise that number and that timeline.
If you cannot attach a number, you do not yet understand the outcome well enough to own it. That gap is the useful part. This fortnight, do it once, on something real and live. See how differently it reads.
THE UNCOMFORTABLE DEBATE
Here is the question to put on the table at your next pipeline review, partner meeting, or supplier review, depending on which side you sit:
If a client (or you, as the client) only paid when the outcome was actually delivered, would this piece of work still exist?
It is uncomfortable because a chunk of what gets sold and bought would quietly fail that test: the bought-and-forgotten deck, the "strategy refresh" nobody implements, the named but not used AI pilot. I like this because hopefully you’ll agree, it’s better you ask it in the room now than have agentic tools and a tighter budget ask it for you later.
WHAT I'VE BEEN READING
The billable hour is cracking. The Wall Street Journal on professional-services firms scrambling to charge for outcomes instead of time, with more than 30% of McKinsey's fees now tied to results. https://www.wsj.com/cfo-journal/inside-consultants-messy-shift-from-hourly-billing-7bd9b802
The slow death of the prestige career. The Guardian on why consulting is the first place the AI shift really bites. https://www.theguardian.com/commentisfree/2026/jun/22/consulting-ai-prestige-careers
Who, or what, does the work. McKinsey on rewiring talent to value once agents join the team, the idea behind my latest article. https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/rewiring-talent-to-value-in-the-age-of-ai
80% of AI projects fail to deliver value. A sobering figure that I read as an operating-model problem, not a technology one. https://www.consultancy.uk/news/44679/why-80-of-ai-projects-are-failing-to-deliver-value
Corporate AI spend is climbing fast. Business Insider on the jump from 0.8% to 1.7% of revenue, and why "where is the value actually going?" becomes the only question that matters. https://www.businessinsider.com/consulting-ai-spend-token-use-mckinsey-bcg-kpmg-deloitte-ey-2026-6
HIT REPLY
I want to know which of the three you are:
trying to escape into an operating or building role
rebuilding a firm or practice from the inside
a buyer who is done being handed decks
‘Hit reply and tell me which, and the single hardest thing about it right now. I read every reply.
Keep building,
Alex
PS. The free resource is the Consulting to Tech Transition Map: the path I wish someone had handed me when I made the jump from advising to building. It is the whole point of this newsletter in one phrase, from AI ambition to operating reality. It is on the site, no cost. Grab it. The next issue lands in a fortnight.
