Running Pre-Mortems in Project Kickoffs
Most projects fail for predictable reasons that teams overlook during the excitement of launch. Experts in risk management and project planning recommend running pre-mortems at kickoff to identify potential failure points before they derail timelines, budgets, and stakeholder trust. This article presents twenty-five practical strategies from practitioners who have used structured anticipation to prevent common project disasters.
Emphasize Lifetime Income Over Market Hopes
When preparing for a risky retirement initiative, I run a focused, brief pre-mortem to test the plan against the most likely failure points we see in practice. We center the discussion on exposures highlighted in our work: longevity risk, inflation, rising healthcare costs, and tax changes.
The team walks through a credible worst-case scenario in which a client outlives their assets or faces unaffordable care, then identifies which parts of the plan would break first. We then select mitigations from our established toolkit, such as diversifying income streams with Social Security, pensions and annuities, applying a guardrails withdrawal approach, and reviewing tax moves like Roth conversions.
From that quick exercise, we prioritize the few actions that materially reduce the largest single failure risk. For example, uncovering elevated longevity risk early commonly leads us to shift emphasis toward stable lifetime income and stronger withdrawal guardrails rather than relying solely on portfolio returns. That change preserves spending flexibility while materially lowering the chance that the retirement plan runs out of money.
Protect Quiet Contributors With Structured Feedback
Moving away from tracking presence -- and moving towards measuring output and impact -- was a big shift at Lock Search Group, and one that I spearheaded with a little trepidation.
I knew, in the long term, it would pay off, if we could stick the landing -- but that was a big if. In a business like recruiting where coordination, pace, and accountability matter a lot, stepping back can feel terrifying.
So, before we launched, we ran a bit of a pre-mortem exercise, and what it exposed was surprising. There was a real issue with employee visibility that I hadn't anticipated. Not everyone advocates for themselves in the same way, and in a more flexible environment, introverts started to become overlooked.
This dry run allowed us to build in the kind of flexibility that suited everyone, making it frictionless for even the quietest people to speak up. We made feedback systematic, so that no one felt wary of contributing.
We also stepped back and trained managers to be more deliberate about how they assessed contribution.
It slowed the rollout slightly, which at the time felt like a trade-off. But in hindsight, it probably made the difference between something that sounded good and progressive, and something that actually worked for everyone.
Expose Vendor Bottlenecks Before Contracts
Conducting a pre-mortem begins by gathering your team and asking everyone to imagine that the project has already failed spectacularly one year into the future. By shifting the perspective from what might go wrong to what already did go wrong, you bypass the natural optimism and groupthink that often cloud new initiatives. Each participant spends a few minutes working individually to write down every possible reason for this hypothetical collapse, ranging from internal technical debt to external shifts in the market. This psychological trick allows team members to voice concerns without appearing pessimistic or unsupportive of the mission, as the exercise frames the identification of flaws as a creative and necessary part of the planning process.
Once all potential failure points are surfaced, the group categorizes them by their likelihood of occurrence and the severity of their impact. This allows the team to move beyond a long list of fears and toward a focused strategy for risk mitigation. For each high priority failure point, you develop an early warning signal and a specific countermeasure. This turns the pre-mortem from a simple brainstorming session into a living document that guides the project's execution. By formalizing these risks early, the team gains a shared language for discussing problems before they become crises, which significantly increases the agility and resilience of the entire operation.
One risk uncovered early during a large scale systems integration was the assumption that a critical third party vendor could handle a specific data throughput. During the pre-mortem, a developer imagined a total service outage caused by a bottleneck at that external API. This prompted a quick load test that revealed the vendor was indeed incapable of meeting the projected volume. Because this was identified before the contract was signed, the team was able to pivot the architecture to include a localized caching layer and a secondary failover provider. This change added two weeks to the initial development phase but ultimately prevented a catastrophic system failure that would have occurred on launch day.
Let Data Maps Reveal Workflow Snags
Run a quick pre-mortem by pairing a 30 to 60 minute focused meeting with a rapid pull of system logs into a process mining tool to reveal how work actually flows. I start by extracting a sample of CRM and support workflow logs, then let the tool draw the real flowcharts so the team can see variants and hidden loops instead of debating theory. In the meeting we review the top variants, highlight any approval or manual handoff bottlenecks, and ask what could make each variant fail under stress. We then prioritize the few fixes that reduce the most risk and lock in simple guardrails before launch. For one mid-market SaaS client we expected a single onboarding flow but process mining revealed 87 variants. About a third of those included loops that were consuming roughly 15% of staff hours each quarter, which led us to change the plan to tackle approval bottlenecks first and target the highest-impact variants. Fixing the top five variants alone freed over 500 person-hours a year, validating the value of a quick, data-driven pre-mortem.
Choose Sites From Customer Gravity
I almost built a warehouse in the wrong city because everyone on my team was too excited to tell me it was a bad idea.
Here's what actually works: gather your smartest people in a room and force them to assume the project already failed spectacularly. Not "what could go wrong" but "we're standing in the rubble, what happened?" The difference is subtle but massive. When you ask what could go wrong, people hedge and stay diplomatic. When you say it already blew up, they get honest fast.
Before I built my 140,000 square foot facility, I ran this exercise with my warehouse manager, CFO, and our biggest client. I said "it's 18 months from now, we're hemorrhaging money and brands are leaving, why?" My warehouse manager immediately said "we picked the location based on your commute, not customer geography." Brutal. True. We'd been eyeing a site 20 minutes from my house when 60% of our client base was shipping to the opposite corner of the country.
That one comment made us pull real shipping data and map customer distribution. We moved the site search 90 miles away. Added 40 minutes to my drive but cut average transit times by a full day and saved clients about $180,000 annually in zone charges. The facility succeeded partly because we caught that error before pouring the foundation.
The other thing that surfaced was staffing. Someone said "we can't hire fast enough during Q4 and we're shipping orders five days late." That fear made us build relationships with three temp agencies six months before we needed them and create a training program that could onboard someone in four hours instead of two days.
Run the pre-mortem before you're pot-committed financially or emotionally. Once you've signed leases or made announcements, your brain won't let you see the problems clearly. Do it when walking away is still free.
Gate Work With Load And Heat Audits
I'm Clay Hamilton, President of Grounded Solutions in Indianapolis, with 20+ years in electrical/mechanical work and ops--so my "risky initiatives" usually involve power, panels, and people being safe while we move fast.
My quick pre-mortem is a 12-minute "assume it failed" walkthrough: we pick the ugliest outcome (fire risk, downtime, failed inspection, angry homeowner), then list the first 5 causes in plain language. I force it into buckets: access (can we physically work clean?), load/capacity (are we guessing vs calculating?), hidden heat (what will thermal imaging reveal?), and communication (what will the customer say we didn't explain?).
One early risk that changed my plan: on a panel upgrade, the pre-mortem called out "we'll open it and find heat damage/overloaded circuits and now we're stuck mid-job." That pushed me to do the load calc + thermal imaging as a gate before scheduling, and to pre-stage the right breakers/parts so we weren't improvising with the house half-powered.
It's boring, but it saves you: write down the failure points, assign one owner per risk, and decide what you'll do *before* you're standing there with a buzzing panel and a customer asking "is this normal?"
Favor High-Intent Actions Over Busyness
Before tackling a risky goal, pause and ask yourself, "If this fails six months from now, what likely caused it?" Keep it quick, honest, and concrete. You're identifying weak points before they hit you.
Assuming more activity automatically produced progress. A quick review revealed the danger; focusing on unqualified leads would waste time and energy. That insight changed our plan; prioritize high-intent actions, track results that count, and align every step with the outcome we actually wanted.
A few minutes of reflection can save months of wasted effort. Anticipating problems turns uncertainty into clear steps, and clear steps accelerate big goals.
Fix Misalignment Before Momentum Builds
When evaluating a risky effort, I always run through this question in the very beginning to pressure test the initiative: if the initiative fails 6 months from now, what went wrong? This immediately alters the conversation and challenges the execution, timing, resources, and whether we have the infrastructure in place to support the idea. In one instance, we identified an issue of internal misalignment early by recognizing that although the opportunity was present, we were advancing beyond the team's ability to support us.
Roles weren't clearly defined, some processes were still being defined, and our map to success wasn't clearly built out. Identifying this early allowed us to address the issue of misalignment prior to it becoming a bigger problem. Subsequently, we slowed the rollout of the initiative, clarified our ownership around it, and developed a stronger foundation. In my experience, I don't consider a pre-mortem to be negative. Thinking about how to solve an issue as early as possible provides you with an opportunity to be honest and make informed decisions, ultimately increasing your chances for success.
Avoid Platforms And Own The Entire Stack
I do something like a pre-mortem before I spend real money on anything. Years ago I spent over a million dollars on Google Ads PPC over the life of a domain. Got almost nothing to show for it long term. No brand equity, no organic traffic and just a massive hole in my budget. I kept spending. The leads were coming in. But I never stopped to ask what happens when I turn this off. The answer was everything disappears. That was the failure point I should have identified on day one. I was building on rented land and didn't see it.
When I rebranded to QuicLoans and focused on organic content, I asked that question upfront. What kills this strategy? The answer was obvious: if I build on someone else's platform I'm in the same trap. I built the entire site myself with my own code, my own hosting. If any single vendor disappears tomorrow nothing breaks. That one risk assessment shaped every build decision I've made since.
Target The Likeliest Failure Mechanism
Before I 410'd 15,000+ product pages on WhatAreTheBest.com to pivot toward SaaS-only, I didn't run a pre-mortem — and it nearly killed the site. Google interpreted the mass page removal as a quality signal problem and algorithmically suppressed the entire domain. The risk I should have surfaced beforehand: "What if search engines interpret removing 90% of our indexed pages as a site-wide quality collapse?" Now I run pre-mortems on every major change by asking one question: "If this goes wrong, what's the most likely mechanism of failure?" Not "what could go wrong" broadly — that generates anxiety, not insight. Identifying the specific mechanism forces you to build a mitigation plan for the highest-probability failure. The 410 pre-mortem I never ran would have saved six months of recovery.
Albert Richer, Founder, WhatAreTheBest.com
Surface Hidden Decision-Makers On Day One
I don't call it a pre-mortem, but I do the same thing every time before taking on a big project or a new client relationship. I ask one question: "What would make me regret saying yes to this in three months?"
That question surfaces things a pros-and-cons list never will. It forces you to skip past the excitement of the opportunity and sit with the version of the future where it went wrong. Usually, the answer isn't about the work itself. It's about misaligned expectations, unclear decision-makers, or a timeline that only works if nothing goes sideways.
The specific risk that changed how I operate came early in my career. We took on a large website project with a main contact who was enthusiastic, responsive, and ready to move fast. What I didn't ask was who else had a say. Three weeks in, a board member saw the design for the first time and wanted to start over. Our timeline was built around a single decision-maker, but the actual approval process involved five. We absorbed weeks of rework because I never asked, "Who else needs to sign off on this?"
Now that's the first thing we surface before any project starts. Not just who's our point of contact, but who can derail this, and when will they see it? If the answer is vague, we build a review checkpoint into the first two weeks specifically to flush out those voices before real work begins.
Most risks aren't surprises. There are things you could have seen if you'd asked the uncomfortable question before the contract was signed, instead of after.
Diversify Away From One Provider
Most founders skip the pre-mortem because it feels like pessimism. It's actually the opposite. Before any major launch I get the team together, assume the thing already failed spectacularly, and reverse-engineer why. That mindset shift unlocks honesty you never get in normal brainstorming. People stop cheerleading and start truth-telling.
One risk we caught before it wrecked us: we were building core features of memelord.com on top of Twitter's API as our primary data source. The pre-mortem surfaced it in under five minutes. "Twitter kills the API or prices us out" was on the board before anyone touched a keyboard. We diversified across platforms immediately. When Elon nuked the API pricing in 2023 and killed dozens of social tools, we didn't skip a beat. Pre-mortem paid for itself ten times over that week.
Add Price Locks To Preserve Trust
The way we run pre-mortems at GpuPerHour is deliberately low-ceremony. We get three or four people in a room, set a 30-minute timer, and start with one sentence: assume this initiative has already failed, and write down the single most likely reason why. Everyone writes independently for five minutes, then we read them out loud without commentary. The constraint of picking just one reason forces people to name the thing they are actually worried about instead of listing every theoretical risk.
We did this before launching our spot pricing feature last year. The initiative was risky because we were moving from flat hourly rates to dynamic pricing that fluctuated based on real-time GPU availability. The pre-mortem surfaced something none of us had put in the launch plan. Two people independently wrote the same failure scenario: customers would see a price spike during a long-running training job that started at a low rate, feel tricked, and churn.
That one finding changed the entire design. We added a price lock mechanism where any job that begins at a quoted rate stays at that rate for the full run, regardless of what spot prices do after the job starts. It added about three weeks of engineering work. But if we had launched without it, the first customer to see a mid-job price jump would have posted about it publicly, and the trust damage would have taken months to repair.
The pre-mortem pattern that works for us has three rules. First, individual writing before group discussion. If you start with open conversation, the most senior person's worry dominates and nobody says their real fear. Second, force a single prediction per person instead of a risk list. Lists diffuse responsibility. A single prediction creates ownership. Third, end the session by picking the top two failure modes and assigning someone to design the mitigation before launch, not after.
The whole exercise takes under an hour and has saved us from at least three launches that would have gone badly.
Focus On Death Risks Not Friction
I'm Runbo Li, Co-founder & CEO at Magic Hour.
Most people skip the pre-mortem because they're too excited about the idea. That's exactly when you need it most. My approach is simple: I assume the project already failed, then I work backwards and ask, "What killed it?" I give myself ten minutes, a blank doc, and one rule. I'm not allowed to be optimistic. Every instinct to say "that probably won't happen" gets overridden. You write down every way this thing dies, no matter how uncomfortable.
The key is separating the fatal risks from the annoying ones. I call it "death vs. friction." Friction you can push through. Death means the whole thesis is wrong. You want to find the death risks in ten minutes, not ten months.
Here's a real example. Early on with Magic Hour, we were building custom AI video models from scratch. The pre-mortem on that path was brutal. I sat down and asked, "If this fails in six months, why?" The answer was immediate: open-source models were improving so fast that anything we built custom would be obsolete before we shipped it. We'd burn months of engineering time building something the open-source community would surpass in weeks. That was a death risk, not a friction risk.
That single insight changed our entire technical strategy. Instead of building models, we built a platform layer on top of the best open-source models. We made it so we could swap in new models as they dropped, sometimes within days. That decision is a huge part of why two people were able to build a product serving millions of users. If we'd gone the custom model route, we'd still be training our first model while the market moved on without us.
The pre-mortem took maybe fifteen minutes. It saved us six months of wasted work and probably saved the company.
People treat pre-mortems like a formal corporate exercise with sticky notes and facilitators. Forget that. Just sit alone, assume you already lost, and be honest about why. The risks you don't want to write down are the ones that matter most.
Prevent Vehicle Chaos With Fit Checks
I run operations for two self-storage locations on Aquidneck Island, so "risky initiatives" for me are anything that touches access, security, move-ins, and customer trust. My quick pre-mortem is a 10-minute whiteboard: "It's 30 days after launch and customers are mad--why?" Then I force answers into four buckets: access/traffic flow, security & liability, customer comms, and staff workload (including what breaks on a Saturday when the office is closed).
I keep it fast by doing a walkthrough like a customer: arrive, unload, find the unit, use a cart, exit, pay online later. Anywhere I hesitate, that's a failure point; I write the fix as a one-line "guardrail" (signage, script, rule, checklist) and assign an owner before we launch.
One early risk that changed my plan: expanding vehicle/boat/RV parking ("store your toys") and assuming it would behave like a regular unit rental. The pre-mortem surfaced two ugly endings: oversized rigs blocking lanes at peak times and disputes over what "fits" in a space, which turns into safety and neighbor headaches fast.
We changed the rollout to require a quick fit/clearance check and a simple traffic rule set (where to stage, how to turn, what hours to move large vehicles), plus clearer wording in our reservation and on-site signage. It slowed the initial conversion a bit, but it prevented the "parking lot Tetris" problem that would've wrecked day-to-day operations.
Preempt Costly Finish Swaps
I often pre-mortem client decision-making by imagining last-minute design changes that could derail budgets. In one luxury hotel project, we identified that clients frequently requested floor finishes mid-install, risking delays and cost overruns. By building a pre-approved finish sample library and requiring formal change sign-off, we cut unexpected overruns by more than half.
Standardize Prompts To Stabilize Outputs
I run a quick pre-mortem for AI work by gathering a small team with different roles and asking one focused question: what single failure would stop this workflow? We spend five minutes listing likely failure points, then vote to prioritize the top one or two that threaten reliability. Early on we uncovered the risk that the team would rely on repetitive prompts, which produced inconsistent outputs and would have broken the workflow. That finding led us to pause the rollout and prioritize building AI skills to minimize unnecessary re-prompts and improve reliability before proceeding.
Insert Proactive Check-Ins To De-Escalate
Before a risky initiative, I run a quick pre-mortem by gathering the key people for 10 minutes and asking one question: "Assume this fails, what most likely caused it?" I listen without debating, then I restate the concerns in plain language to confirm we all mean the same thing, and we agree on the top few failure points to address first. In my setting, I uncovered early that a plan could fail because a family felt unheard and would escalate the issue before we could correct it. That risk changed the plan by building in a short, proactive check-in where we clarify expectations, acknowledge emotions, and confirm what will happen next before moving forward.
Split Averages To Unmask Weak Segments
One risk I uncovered early was false confidence from blended performance data. On paper the initiative looked strong because the average metrics were healthy. I ran a pre mortem and asked the team how the plan could fail across audience segments traffic sources and time. That exercise showed that a small group of high intent users was hiding weaker results from the broader audience.
We adjusted the plan by setting clear goals for each segment instead of relying on one average. I also rolled out spending in smaller waves to reduce risk. We added faster reporting that focused on early signals instead of overall results. This approach slowed the launch a bit but helped protect margin and improved decision making over time.
Prioritize Execution Over Slide-Deck Optics
One of the most useful pre mortem questions we use is simple. We ask what would make this look successful in a slide deck but fail in real life. This shifts the focus from big ideas to how the work will actually happen. We bring a small cross functional group together and give them a short time to think through risks.
We ask everyone to list possible failure points based on their role. Sales may point to incentives while finance may note margin pressure and operations may highlight timing. When these views match the risk becomes clear and harder to ignore. In one case we saw we were relying on executive excitement instead of team behavior so we changed the plan.
Design Around Missed Client Kickoffs
Before we kick off anything that carries real risk, whether that's a new service offering, a significant hire, or a major process change, I run a short pre-mortem with whoever is closest to the work. The prompt is simple. Assume this failed six months from now. What went wrong?
It's a twenty-minute conversation, not a formal exercise. The goal is just to get people out of planning mode and into honest mode for a few minutes before we commit.
The most useful one we ran was before rolling out a restructured client onboarding process a couple years ago. One of our project leads immediately said "we haven't thought about what happens when a client misses the kickoff call."
She was right. We had built the whole thing assuming clients would show up and engage on schedule and almost none of them do consistently. We rebuilt the process with that constraint in mind before we ever launched it.
That one conversation probably saved us three months of rework. The risk was obvious in hindsight but nobody had said it out loud yet. That's what the pre-mortem is actually for.
Secure Native Research And Localized Reports
Chris here -- I run Visionary Marketing, specialist SEO and Google Ads agency.
A pre-mortem is just structured paranoia. You imagine the project failed spectacularly, then you work backwards to figure out what went wrong. It's not pessimism--it's a conversation about failure points that people won't mention in a normal planning meeting because nobody wants to be the person who kills the vibe.
We used one recently for a client's international expansion into German and Dutch markets. The brief was straightforward enough: SEO and Google Ads from scratch in two new languages and regulatory environments. But in the pre-mortem, something came up that would've tanked us: keyword research quality.
See, we assumed we could contract that work to a freelancer in Germany who'd handle translation and regional nuance. Sensible on the surface. But the pre-mortem revealed a buried risk nobody had actually voiced: our freelancer's deadline wasn't aligned with our campaign launch, and there was zero buffer for revision cycles. If the initial keyword research came back weak--say, too much focus on volume over intent, or missing critical regulatory language--we'd be scrambling two weeks into the campaign. At that point, pivoting is expensive and demoralising.
That insight changed the plan entirely. We brought keyword research in-house and hired a native speaker as an advisor rather than outsourcing the whole thing. Added about £3,600 to the project cost but eliminated that critical failure point. The campaign launched clean, and we caught a language-specific nuance in healthcare claims regulations that the freelancer wouldn't have flagged.
The pre-mortem surfaced something else too: our reporting template wasn't designed for client handoff in a non-English market. That was a second-order failure point we would've discovered on day fifteen when the client asked for a dashboard they could actually read.
The sharp takeaway? The failures nobody wants to name in planning are the ones that'll actually kill you. Make the space to name them early.
Counter Shame Drop-Off With Clear Privacy
I run a fast pre mortem by asking three people from different teams to imagine a headline six months later that explains why the idea failed. Each person then lists the hidden assumptions behind that headline and ranks them by how fast they could cause damage instead of how likely they are. This helps us focus on early warning signs that are easy to miss.
One early risk we found was shame based drop off. In sensitive areas people may click with interest but hesitate when identity or timing feels uncertain. This changed our plan to focus on reducing friction first with clear privacy signals and simple messaging. We also removed steps that could make people pause or feel unsure.
Seek External Counsel Before Novel Moves
Getting an outside opinion is crucial. I always consult with internal stakeholders, starting with my C-suite, when planning new ventures, but especially when we're trying something radically different, I'll run it by someone completely outside of the business first. This has been especially helpful as we work our way into the healthcare space. There's a lot of room for advancement with AI tools, but there are also a lot of regulatory hurdles to overcome, and talking to doctors, hospital managers, and insurance leaders helped me get a much better sense of what's possible in this space.
Reduce Dusk Liabilities With Safer Format
Before launching anything risky at Doggie Park Near Me, we run what I call a "what kills this" session. The whole team spends 20 minutes assuming the initiative has already failed, and everyone writes down the single most likely reason why. No filtering, no optimism allowed. We group the answers, and whatever shows up more than once becomes a mandatory solve-for before we move forward.
The risk that changed our plan most was when we were about to launch an off-leash evening event series during summer. The pre-mortem surfaced something we had completely overlooked: liability exposure from reduced visibility at dusk combined with unfamiliar dogs mixing for the first time. Multiple team members independently flagged that a dog altercation in low light would be both more likely and harder to manage.
That single insight changed the entire plan. We moved the events to late afternoon, added a mandatory temperament screening at check-in, and invested in additional portable lighting for the park areas. We also capped attendance at half of what we originally planned for the first three sessions so staff could manage the flow.
The events ended up being one of our most popular offerings that summer, but they would have been a disaster in their original form. The pre-mortem took 20 minutes and probably saved us from a serious incident and the insurance nightmare that would have followed. Now we never skip it, even for small initiatives.
