The problem
Stillborn projects: denial at six months
Some startups do not die in a day; they fade while the team keeps a busy pace. By six months, signals should talk: frequent iterations without verifiable customer learning, repeated refusal to choose a target segment, a roadmap driven by engineering rather than paid commitments. The core problem is missing revisable proof: every sprint ships work, but rarely falsifies a hypothesis or tests pricing seriously. Activity indicators (commits, features, posts) replace truth indicators (contracts, renewals, recurring usage). The six-month window is not magic: it is when runway becomes visible, first sales cycles should have closed, and early-adopter retention should be readable. Without explicit failure criteria, the team reframes each weak signal as “we are still early” and postpones market confrontation. The cost is twofold: wasted money and attention, and missed opportunity to pivot or stop cleanly.
Behavioral signals stack: longer meetings, micromanagement, a roadmap that swells without external proof. Early adopters stop answering surveys or drift to competitors while the team avoids naming why. Funding can temporarily mask the issue: more runway without criteria only postpones confrontation. At six months, a stillborn project often leaves a paper trail of many “next steps” and few conclusions of “we were wrong about X.” Clarity means turning legitimate early ambiguity into testable hypotheses with dates—and honoring those dates as seriously as a customer commitment.
Why it fails
Why detection is late
Founder identity ties to the project: admitting early failure threatens self-image. Sunk investments (time, reputation, funding) create commitment bias: you “justify” the past by continuing. Internal dashboards mix effort and outcome: a busy team looks legitimate even without traction. Startup culture celebrates perseverance; saying “stop” is socially costlier than “we persist.” External stakeholders may lack granularity to challenge assumptions. Finally, legitimate early uncertainty makes it hard to separate healthy exploration from stagnation: without dated, measurable milestones, every delay feels acceptable. These psychological and structural forces explain why many projects zombie on six to twelve months past the point where proof is missing.
Advisor incentives can be asymmetric: cheering perseverance is socially rewarding, saying “stop” risks the relationship. Internal data is filtered before it reaches decision-makers: nobody wants to be the bearer of bad news. Calendar density substitutes for depth: being busy becomes proof of seriousness. Co-founders may mutually avoid “damaging morale.” Without written scaffolding, every milestone slip is negotiated verbally and forgotten. Breaking the loop requires durable artifacts: public milestones, shared failure criteria, and someone empowered to say “we have not learned fast enough.”
A concrete method
Method: truth milestones at 30 / 90 / 180 days
Before day one, set three milestones with binary criteria—for example “ten validated problem interviews,” “three LOIs with numbers,” “first paid contract or equivalent.” Write what invalidates the project if a milestone is missed (segment change, offer reset, stop). Keep a hypothesis log: statement, expected evidence, date, outcome. Measure learning velocity, not deliverable volume: how many hypotheses tested per month with market feedback? Run quarterly kill/continue reviews with someone without emotional debt (advisor, second co-founder, board if mature). Track pipeline quality: deals stuck for X weeks, recurring loss reasons. Compare remaining runway to time needed to reach the next proof threshold. If slack is below two realistic iteration cycles, the stillborn-risk signal is red.
Add coded qualitative indicators: interviews where the customer spontaneously proposes a paid next step, real budget presence, quantified pain. Separate “interest” from “commitment”: signature, deposit, dated LOI. For product, measure whether users return without marketing nudges—if organic retention is nil, no feature saves weak positioning. Keep a log of pivots refused: often the team sees the signal but fears change. Finally, test your pitch with an external anonymized panel: dissonance between your story and their feedback is an early warning. The aim is not pessimism but adjustment velocity before denial becomes structural cost.
Example
Example: internal platform that wins no budget
A B2B team builds an “insights platform” for enterprises. At six months, two free pilots, no budget line migrated, and internal sponsors rotate without engaged successors. Demos impress, but no champion signs a mandate. Founders count dev hours and “very interested” feedback as traction. After strict milestones, they see the problem is not a client priority and the purchase cycle exceeds their runway. They narrow scope to a “audit plus deliverable” sold in thirty days or pivot to mid-market with shorter cycles. Without dated failure criteria, they would have continued six more months on the same path. Early detection preserved capital and credibility for a second, structured attempt.
Cross-cutting lessons: a pilot without budget is not proof; a single sponsor without backup is risk; a purchase cycle longer than runway is a strategic veto. The team documented why enterprise was a dead end at their stage—that note prevents repeating the error under a new codename. The mid-market pivot forces a commercial reset: shorter scripts, paid trials, standardized onboarding. Investors often prefer a reasoned stop to cash burn on an unvalidated thesis. The human signal: when strong contributors ask “where is traction?” it is not always cynicism—it is an early warning worth taking seriously.
What to do now
What to do now
Write a one-pager with your 180-day milestone and minimum acceptable proof. Add a kill/continue review to the shared calendar. List three weak signals you usually excuse; for each, define a red numeric threshold. Reach five customers or prospects this week with a closed question on budget or timing, not only interest. Compute runway over burn and how many full iterations remain before cash gets critical. Name a devil’s advocate for the next product meeting. If you cannot state what would make you stop in under two sentences, you still lack failure criteria: fix that before adding features. Clarity at six months beats a heroic narrative at eighteen months without cash.
Block a “raw truth” slot this week with no marketing slides. Ask each lead: what proof are we missing to continue? Write the answer in a shared doc, not only verbally. If the missing-proof list is long and unchanged for thirty days, you are not iterating fast enough on the right plan. Prepare two paths: continue under conditions, and stop or pivot with criteria. Signal early to key stakeholders if a milestone is at risk—transparency preserves trust. Use a timer: at most thirty minutes to pick the next market test, then execute. Founder dignity also shows in naming a blind spot early instead of extending a traction theater.
Related reading
Lumor puts your idea in front of 13 AI roles to stress-test assumptions, surface blind spots, and deliver a verdict, scores, and an execution plan.