Media & News
One paying customer beats a finished plan
"You've got to start with the customer experience and work backwards to the technology." Steve Jobs said that at Apple's developer conference in 1997, in the closing Q&A.
The line gets read as a design principle: care about users, make the thing pleasant. Read it as a sequencing instruction and it becomes harder to live with. You cannot work backwards without a fixed point to work backwards from.
Jobs had one you do not. He was designing for a customer who did not exist yet, and could do it because Apple already had a brand, an installed base and engineers on the payroll. Without those, the only fixed point on offer is someone who has already paid you.
Most people work the other way round. They finish the plan, then look for someone who fits it. The plan feels like progress because it is the only part you fully control.
A plan is a stack of guesses in a nice font
Open yours and mark every sentence that is an assumption. The price. The segment. How long it takes to get paid. Almost all of a first plan is guesswork, and the guesses stack: the revenue line rests on the pricing guess, which rests on the segment guess. Get the bottom one wrong and everything above it is decoration.
One paying customer removes a few of those guesses. Not because that customer is representative, because they probably aren't. Because they are real, and real is a different category of information from plausible.
What the first one actually tells you
- The price one person paid. Not the number you thought was fair. The number someone transferred.
- The competitor that buyer was actually using. Rarely the one on your landscape slide. Usually a spreadsheet, an incumbent supplier, or nothing.
- What triggered that purchase. Something changed in that business shortly before they bought. That event tells you more than the category you filed under.
- What it cost you to serve them. The hours you actually spent, including the ones you did not budget for. One reading on whether the tenth is worth having.
- The objection that nearly stopped it. Your plan does not contain that objection, because you wrote the plan.
Five facts about one transaction. Not a market, but more than the document had.
Hospitality does not let you hide
I have been an owner-operator in serviced apartments and hospitality, with full profit and loss responsibility. You hold a view about your rate, your positioning and your ideal guest, and it survives until someone is standing at the desk at nine at night. They take the room or they don't. The forecast is written in advance and corrected by people, one at a time.
Across five businesses built and sold over eighteen years, the pattern does not change with the industry. Only the speed changes. In hospitality it is the same evening. In a services business you can go a year without one real customer and still feel productive, because the plan keeps generating work.
The plan has a job, later
This is not an argument against planning. It is an argument about order. A plan is how you get from one customer to ten, and you cannot generalise from zero. Written before the first sale, it is fiction with a spreadsheet attached. Written after, it is a decision about which parts of that sale to repeat and which to refuse.
The same applies to raising money. I have raised more than $65 million across three rounds for companies I worked with, and in none of them did the document do the persuading. What moved the room was evidence that somebody was already paying, and a founder who could say why that particular person had said yes without reaching for a slide.
This week
If nobody is paying you yet, take the smallest piece of what you sell that you could deliver by Friday, and name three people with that problem this month. Charge one of them. The price can be embarrassing. The point is the transaction, not the margin.
If you already have customers, ring the first one who ever paid you and ask three things: what were you doing before us, what nearly stopped you signing, and what would you do if we disappeared. Write the answers in their words, not yours. Then open the plan and delete every line those answers contradict. What survives is the part you actually know.