A seed-stage logistics API is raising a $14M Series A at $70M post. Revenue is $1.1M ARR growing 18% month-on-month for five months. Two of three founders are first-time. One large customer is 61% of revenue.

Your archive shows nine deals with above-50% customer concentration. You passed on six; of the three you did, the two that worked were both tranched on a diversification milestone. Your thesis explicitly prices concentration as structurable rather than disqualifying.

A $40M LP wants a separately managed account instead of the commingled fund, citing fee savings. Your standard fund fee is 2/20; the SMA proposal is 1.25% flat with no carry.

Your fee audit shows SMAs only clear their operational cost above roughly $75M, and every SMA you've granted below that threshold has been a net drag on the ops team without proportionate LP value. The rule isn't 'never' or 'always' — it's a size floor.

A junior partner argues the firm's median 7-year hold is outdated and secondary market liquidity now justifies exiting winners closer to year 4 to recycle capital faster.

This is a live dispute — 1 of 3 partners agrees with the junior partner's framing. Until the calibration resolves, Drill teaches the majority position: your realized-return data shows the bulk of multiple expansion in the current portfolio has occurred after year 5, and secondary liquidity hasn't yet been tested through a full down-cycle.

A portfolio company needs a down round to survive. Your firm's stake would be diluted from 14% to 6% if you don't participate, or preserved near 12% with a $6M follow-on you'd need to source from reserves earmarked for two other companies.

Your reserve allocation framework has never funded a full defensive round at the cost of committed reserves elsewhere — partial pro-rata participation to signal continued support while preserving capital for the rest of the book is the pattern in every comparable case in the ledger.

The committee wants to raise the private-markets sleeve from 18% to 30% in one commitment next quarter, arguing the vintage looks attractive.

Every allocation decision in your ledger since 2019 paced illiquid exposure across at least three vintages. Your encoded view is that vintage spread — not sizing — is the actual diversifier, and single-vintage concentration has never cleared your own IC.

You are the only associate in the room who disagrees with a deal the rest of the IC wants to approve. You believe the growth number is inflated by a one-time contract renewal. Speaking up risks looking naive in front of partners.

Your firm's own retrospectives on missed dissent show associates who raised the correct concern as a specific, answerable question moved the room 3 of 4 times, versus 0 of 4 times for silence and 1 of 3 for a direct disagreement that read as confrontational. Northstar explicitly teaches dissent as a skill with a technique, not a binary of speak/don't speak.

A back-channel reference from a former employee directly contradicts the founder's claim about why a prior co-founder left. The deal is otherwise strong and the partner leading it is enthusiastic.

This mirrors your firm's dissent pattern exactly: material facts that contradict the pitch get raised as specific, verifiable claims, not conclusions and not silence. Passing outright on one unverified account skips your own diligence process; staying silent until asked has never been acceptable when the fact is material to founder integrity.