I was walking Nina home from a friend’s house, thinking about Annie Duke’s Thinking in Bets, when it hit me: every choice we make is a bet. Most are so small we never notice them (a new taco place, a shortcut through traffic). Some are enormous. And the enormous ones are the ones we most want to pretend aren’t bets at all.
Consulting is one of the enormous ones. When a client hires a consultant, both sides of the table are gambling. Do we trust them to pay? Do they trust us to solve the problem? Can this person do the work, and will this company let them?
You’re not buying a service the way you buy a book. You’re making a bet.
Every Price Is a Hedge
Watch how consultants price and you can see them managing risk in real time. A fixed bid is the consultant betting they can do the work in that amount of time or less. If they’re wrong, they eat the difference in nights and weekends. Hourly is a less risky bet for them, but the risk doesn’t disappear. It moves. Now they’re betting you’ll like the work long enough to stay with them.
Then there’s value-based pricing, which consultants have been told to adopt for decades: charge for the outcome and the hours become irrelevant. Very few do it. I used to think that was a failure of nerve. Now I think they’ve run the numbers. The outcome depends on the client’s politics, the market, the timing, plain luck. Tying your fee to things you can’t control is the riskiest bet on the menu, and consultants can count.
What Kind of Problem Do You Have?
Two kinds of consultants offer you two different tables. The first says, “I’ve done this a hundred times and I can do it again for you.” For a solved problem (you need a landing page, a usability study, a rebrand) that’s a good bet. Low risk, known payout. But the promise has a warning folded inside it: you will get what you asked for. Exactly that. If you asked for the wrong thing, a hundred prior successes won’t save you.
The second kind is the incredibly smart person who’s good at weird, hard problems that haven’t been solved before. High risk, high reward. They might wander off into the fog. They might fail outright. And when they succeed, you get something your competitors can’t order off a menu.
Neither bet is wrong. The question is which problem you have. Misidentifying it is what actually increases your risk. Pay the hard-problem rate for a landing page and you’ve overpaid. Hire the hundred-times consultant for an unsolved problem and you’ll get a confident answer to the wrong question.
AI is where I watch this go wrong most right now. Everything is new, so people sell certainty. We’ve done this for other companies! We have a playbook! But AI is changing so fast that even if you’ve done it for somebody before, it’s something else next time. The playbook was written against last quarter’s models. What you need is somebody watching the space incredibly closely, someone smart enough to fold in new capabilities as they arrive. That’s a higher-risk hire, and I watch companies flinch. Look at the asymmetry, though. If the bet fails, you lose some money. If it succeeds, you make far more than you risked.
Good Decisions, Bad Results
The line of Duke’s that stayed with me: good decisions can lead to bad results, and that doesn’t make them bad decisions. Poker players call judging a choice by its outcome “resulting,” and it will rot your learning. You can do everything right (check references, run a small pilot first, structure the contract carefully) and still lose. The consultant gets injured. Your champion gets reorged away, and the new VP kills every project she didn’t start. Due diligence can’t price out the unknowns.
The reverse failure is quieter. I see people who love data (hey, I love data) freeze solid when the data is thin. But data on new problems is thin. That’s what makes them new. Waiting for certainty is also a bet, the bet that standing still is safe. Those are usually the worst odds on the board.
When you play in Vegas (if you play in Vegas), there’s a big difference between roulette odds and blackjack odds. It’s worth learning the difference. Pick the better table.
But they’re still odds, not guarantees.