Set an appetite, not an estimate

Somebody asks how long the reconciliation view will take. You say two weeks, carefully, having thought about it. Five weeks later you are still going, and every week you have been slightly wrong in a way that felt reasonable at the time.

Nothing in that story is fixed by estimating better. It is fixed by noticing that the wrong question got asked.

Two different objects

Basecamp’s Shape Up draws the distinction more cleanly than anyone:

“An appetite is completely different from an estimate. Estimates start with a design and end with a number. Appetites start with a number and end with a design. We use the appetite as a creative constraint on the design process.”

Read the direction of travel. An estimate takes a fixed scope and produces a duration, which means the duration absorbs every surprise. An appetite fixes the duration and makes the scope absorb them instead.

Their glossary is one line: “Appetite: The amount of time we want to spend on a project, as opposed to an estimate.” Not how long it will take. How much it is worth.

That is why an appetite cannot be wrong in the way an estimate can. It is not a prediction about the world. It is a decision about your own priorities, and you are the authority on those.

The move

Before anyone designs anything, name the number you are willing to spend, and let the design be whatever fits inside it.

Shape Up’s own way of arriving at that number is a set of questions rather than a calculation:

“Is this something worth a quick fix if we can manage? Is it a big idea worth an entire cycle? Would we redesign what we already have to accommodate it? Will we only consider it if we can implement it as a minor tweak?”

Notice none of those asks how hard it is. They ask what it deserves.

What makes the number real

An appetite you will quietly extend is an estimate with better manners. The thing that makes it bind is uncomfortable, and Basecamp state it plainly:

“Teams have to ship the work within the amount of time that we bet. If they don’t finish, by default the project doesn’t get an extension. We intentionally create a risk that the project, as pitched, won’t happen.”

They call the mechanism a circuit breaker: cancel by default rather than extend by default. And the reason to accept something that severe is the one benefit an estimate can never give you:

“a smart bet has a cap on the downside. If we bet six weeks on something, the most we can lose is six weeks.”

ESTIMATE                        APPETITE

"How long will the              "What is fixing reconciliation
 reconciliation view take?"      worth to us right now?"

  scope fixed                     time fixed
  time absorbs surprises          scope absorbs surprises
  "two weeks"                     "two weeks, and we ship whatever
                                   fits, or we stop"
  week 3: "nearly there"
  week 4: "one more thing"        week 2: something shipped,
  week 5: sunk cost                 or the bet is off and we
                                    have lost exactly two weeks
  downside: unbounded
                                  downside: two weeks

Being honest about the evidence

Two things I am not going to claim.

Nobody has compared appetite-setting against estimating. There is no study showing this produces better outcomes. The argument here is mechanical, about which variable absorbs uncertainty, not empirical.

And this is not a scale-free technique. Basecamp described their own context when they wrote it: about fifty people, roughly a dozen in the product team, structured so teams work uninterrupted for a cycle. Their own appendix is careful about that, and about the fix: “Six weeks might not be the exact time frame for your team”, but be deliberate about what you bet on and cap the downside regardless. For a small team they suggest bets of different sizes each time, maybe two weeks here, three there.

So do not import six-week cycles. Import the direction of the question.

What it costs

Work you have already paid for can be thrown away. That is not a side effect, it is the mechanism. If nothing is ever cancelled, you do not have an appetite. Expect this to feel wasteful the first time, and expect somebody to argue that finishing is cheaper than stopping. Sometimes they will be right.

And it requires somebody with the standing to set the number. If you cannot set it yourself, you can still ask for it, and asking is most of the value: what’s this worth to us? is a question a product lead can answer in a sentence and rarely gets asked.

Try this week

Take the next piece of work anyone asks you to size, and reply with a question before a number:

Before I estimate it, what’s your appetite? If I could do a rough version in three days and a proper one in three weeks, which are we in the market for?

Two things happen. You find out what it is worth, which nobody had said out loud. And you have changed a conversation about your accuracy into a conversation about their priorities, which is where the decision actually lives.

Then design to the number you get back, rather than negotiating the number to fit the design you had already imagined.