Suneet Singal on the premortem

Suneet Singal's page on the premortem teaches an exercise Gary Klein described for teams that are about to commit. It sits with the other decisions essays beside the podcast The Opportunity Notebook with Suneet Singal. Nothing here is offered as a transcript of that podcast, and nothing is offered as his personal method. A premortem asks people to imagine that the plan has already failed and then to say why. The imagination is a permission slip. In an ordinary meeting, loyalty and momentum make reservations sound like disloyalty. In a premortem, a reservation is the assignment. Klein's Harvard Business Review article is the source for the exercise. This page retells the habit in plain language, shows where it fits before a hard-to-reverse step, and marks a labeled hypothetical so the story stays fictional. His published background in real estate and corporate finance is context for why planning meetings need a way to hear the quiet risk, not a claim about any particular project.

Klein's exercise

Gary Klein's Harvard Business Review article Performing a Project Premortem starts from a plain trouble. Projects fail often, and people who see the weaknesses stay quiet during planning because the room has already leaned toward go. A premortem changes the assignment. The team imagines that the project has failed, then generates plausible reasons. Klein's point is that the failure frame makes it safer for a knowledgeable person to say the worrying thing. Prospective hindsight, the stance of looking back from an imagined future, is the mechanism. You are not asked whether the plan might fail, a question that sounds like a vote against your colleagues. You are asked why the failure, now treated as a fact inside the exercise, occurred.

The steps, as this page uses them from that article's idea, are short. The person leading the exercise states the plan in a sentence. Everyone is told to imagine that it is a year later, or whatever horizon fits the plan, and that the plan has failed badly. Each person writes reasons silently, on their own page, before anyone speaks. Then people read reasons aloud, one at a time, until the list is on the wall. Only after the list exists does the group ask which reasons deserve a change in the plan. The silence at the start matters. A shared brainstorm rewards the first confident voice. Silent writing gives the quieter reason a chance to exist before it is edited by the room.

This is a meeting habit, not a prophecy. The imagined failure is a prompt. It does not mean the plan will fail, and it does not produce a probability. If someone leaves the room saying the premortem proved the project is doomed, they have misunderstood the exercise. It produced a list of ways the project could fail, some of which the group can now block, some of which it can watch, and some of which it may accept. The value is the list and the changes, not a mood of doom or a mood of reassurance.

Why a failure story is different from a risk list

A risk list made while everyone is still selling the plan tends to be polite. People offer risks that sound manageable, because an unmanageable risk feels like an attack. A premortem asks for the story of failure first, and for the management second. The order changes what can be said. A person can write the vendor misses the date and we have no second source without immediately proposing a fix that softens the sentence. The sentence gets to be true as a fear before it is made comfortable. Klein's article is aimed at that social fact. The knowledge is often already in the room. The room's loyalty is what keeps it off the table.

There is a second difference. A risk list is usually a list of categories: time, money, people. A failure story is particular. The treasurer is also moving house in the month we assumed she would reconcile the books. Particulars are what you can act on. Categories invite a slide titled Risk Management and no change in the plan. When you run the exercise, push the reasons toward the particular. Who does what, by when, that turns out to be impossible? Which assumption, if false, makes the rest of the schedule decorative? Kahneman, Lovallo, and Sibony describe overconfidence and the planning fallacy among the biases that make a base case too rosy. A premortem is one way to put a less rosy story in the notes without asking anyone to be the official pessimist.

Keep the exercise from becoming a contest in gloomy cleverness. The instruction is plausible reasons, not theatrical ones. A meteor is not a contribution. A key person with no backup is a contribution. The leader should say that distinction at the start. You are looking for failures this group could have seen coming, because those are the ones the group can still change. Distant absurdities waste the hour and give skeptics a reason to mock the method. The method is sturdy enough to be used in ordinary language. It does not need a costume.

How to run the hour

Give the exercise a clock. Ten minutes of silent writing is enough for a plan the group already understands. If the plan is not understood, you do not have a premortem problem. You have a naming problem, and you should return to the page on what decision you are actually trying to make. Each person should aim for several reasons, not one polished essay. A reason can be a sentence. Then go around the room. Each person offers one reason that has not already been said, and you continue until the pages are empty. This round-robin is how a senior person's first idea does not become the whole list. Write the words the person used. Paraphrase is how a sharp reason becomes a mild one before it reaches the wall.

When the list is complete, mark each reason in one of three ways. Change the plan now. Watch this on a stated date. Accept this and proceed with open eyes. The third mark is legitimate. Not every risk can be removed, and pretending otherwise turns the premortem into theater. An accepted risk should be accepted by the person with authority to close the decision, not by a shrug. A watched risk needs a date and a person. A change needs an edit to the plan before the meeting ends, or a scheduled decision about the edit. Reasons that produce no mark were entertainment. The page on checklists can hold the watches and the changes so they are not left in a photograph of a whiteboard that nobody types up.

Invite the right people. Klein's point is that knowledgeable dissenters are the ones with the reasons. A premortem of a building plan that includes only the enthusiasts, and not the person who knows the permit calendar, will produce eloquent risks and miss the calendar. Include the person who will do the work, not only the person who will announce it. If someone cannot attend, ask for their page in advance and read it in the round. Absence is one of the ways a known problem stays out of the notes.

What to do with the stories

The stories are inputs to a revision, not a verdict. After the marks are made, read the decision sentence again. Does it still describe the commitment? A premortem sometimes reveals that the group is about to decide something larger than the sentence admitted. The series cannot be announced because the speaker has not agreed. That discovery renames the decision. Good. Rename it. A premortem that is not allowed to change the plan, or to delay it, is a ritual of anxiety. People will notice, and the next premortem will be attended with cynicism. If you ask for reasons, budget time to respond to the ones you marked change.

Some reasons should travel to other pages in this set. A reason about what the plan uses up belongs with opportunity cost. A reason about effects that arrive after the launch belongs with second-order effects. A reason about how often this kind of plan fails belongs with base rates. You do not need to turn the premortem into a seminar on those ideas. You need to notice when a reason is actually a request for one of them. The vendor has slipped the last three times we hired them is a base-rate remark wearing a story. Write it as a question about the class of jobs, not only as a hunch about this vendor.

Keep the raw list. The polished minutes will remember the changes and forget the reasons you accepted. Six months later, when a watched risk arrives, the raw sentence is what tells you whether you saw it. Store the list with the decision memo. A one-page memo cannot hold every reason. It can hold a line that says the premortem list is attached, with the date and the names of the people who wrote. Names matter. An unsigned risk is easy to dismiss as atmosphere. A signed sentence is a colleague's judgment, which the future meeting owes the courtesy of a reply.

A premortem is not a postmortem

A postmortem looks back at a failure that happened. It is valuable, and it is late for the project it studies. A premortem borrows the honesty people sometimes find only in a postmortem and spends it while the plan can still change. The emotional difference is the point Klein is using. After a real failure, blame and grief crowd the room. Before the launch, the failure is fictional, so the same sentence is cheaper to say. Do not wait for the cheaper sentence to become the expensive one. If your group only tells the truth in postmortems, the premortem is the meeting you are missing.

The two meetings also ask different questions. A postmortem asks what happened and what we will do differently next time. A premortem asks what could happen and what we will do differently before we start. Mixing them produces a vague lesson about communication, which is the lesson groups write when they do not want to name a particular change. Insist on the particular. The next time we will communicate better is not a mark on a premortem list. We will name a second person who can reconcile the books if the treasurer moves in May is a mark. Specifics are how the exercise respects the people who will have to carry the plan.

There is a limit to the transfer. A premortem does not replace a review after the fact. Imagined reasons can be wrong, and real failures can arrive from a direction nobody wrote down. Run the postmortem later without embarrassment that you also ran the premortem. Finding a new reason at the end does not mean the earlier hour was wasted. It means the list was incomplete, which you already knew, because no list of the future is complete. The claim for the method is modest. More of what the group already feared will be said in time to matter.

A labeled hypothetical about a conference

This is a labeled hypothetical. The conference, the prices, and the failures are invented. A library wants to host a one-day conference for neighborhood historians. The plan says two hundred people, four speakers, a printed booklet, and lunch. The director is proud of the plan and has told the board it will happen in June. Before the invitations, a colleague asks for a premortem. The room imagines it is July and the day failed. Silent pages produce these reasons, among others. The hall's air conditioning fails on hot days and June is a guess about the weather. Two of the four speakers have not actually said yes. The booklet printer needs three weeks, and the biographies are not written. Lunch for two hundred was copied from a conference in another city that has a larger audience than this library has ever drawn.

None of those sentences is a research finding. They are the fictional team's plausible stories. The useful work is the marking. The speaker reason is marked change: no public date until three speakers have confirmed, and the fourth is a backup from the library's own staff. The printer reason is marked change: the booklet becomes a two-page sheet that can be printed in the building, unless the biographies are done by a written deadline. The attendance reason is marked change to the success test: the day is a success at eighty people, a figure chosen for this story, not taken from a study. The weather reason is marked watch, with a person assigned to check the hall's backup plan two weeks prior. The director's pride is not on the list, because pride is not a reason the day fails. The unconfirmed speakers are.

The hypothetical shows a premortem earning its hour by editing the plan, not by scolding the director. It also shows a success definition moving, on purpose, from a borrowed number to a local one. That move belongs on the page about whose definition of success you are using. The story is not about Suneet Singal, and it is not a template for running a conference. If your plan is a conference, your reasons will be your own. The pattern to copy is silence, then sentences, then marks, then edits before anyone is invited.

Limits of the method

A premortem cannot supply facts the room does not have. If nobody knows the permit calendar, the exercise will not invent a true one. It may reveal the ignorance, which is worth revealing. Go get the calendar. The method also fails when the leader signals the right answer. If the opening line is we all know this will work, but let's pretend, the pretend will be thin. State the failure as the premise of the exercise and then stop talking. Leaders who add reasons first will hear their own reasons restated. Speak last if you speak at all.

The method can be overused. A two-way door that could be settled by a one-week trial does not need a solemn premortem. Use the tool when the commitment is expensive to undo, when the room has a reason to stay polite, and when the plan is specific enough to fail in a particular way. Klein offered it for project plans. A life, a personality, and a vague hope are poor subjects. You will get fiction of the unhelpful kind. Point the exercise at a plan with names and dates.

Finally, do not cite the premortem as evidence that you were careful if you ignored the list. The existence of a meeting is not a change in the plan. The page on decision checklists makes the same warning about lists that nobody reads aloud. A premortem is a way to hear the quiet risk. Hearing it is the first half. The marks, the edits, and the stored list are the half that makes the hour honest. Suneet Singal's podcast may raise neighboring questions about judgment. This page does not claim the premortem as his technique. It claims only that Klein's exercise is a clear, public habit, and that a group about to step through a heavy door can use it before the door swings shut.

What this entry does not claim

Suneet Singal’s published professional background is in real estate and corporate finance. Company filings linked from this site record a chief executive officer departure from FC Global Realty Incorporated effective 2 January 2018 and an acting chief executive officer appointment at First Capital Investment Corporation effective 12 September 2017. The Suneet Singal Scholarship, Grant and Giving websites describe him as the founder of those programs. This entry explains a general subject. It is not a private track record, a current job title, an investment result, or a quote.

Pair the premortem with second-order effects, reversible and irreversible decisions, and checklists for high-stakes choices.