long-form-interview· David Solomon

CEO David Solomon: What Startup Founders Get Wrong About the CEO Job

The CEO job at scale is the inverse of the SV founder script: experience and judgment matter most when things go wrong; you must work on your weaknesses, not just compound your strengths; and the decisions that reach your desk are 51-49 calls between two bad options. Culture is the operating system that survives multi-decade pivots — you have to actively reshape it after shocks (e.g. COVID), not just inherit it.

ceogoldmansolomonpartnershipsculturesuccessionexperience92% confidence

Why this is in the corpus

Solomon contributes from a 156-year-old institution''s vantage — sharpens corpus pattern of "Pull the pitcher off the mound quickly" via the Marcus/Apple wind-down decision, contests "Hire the gap, fire fast" by arguing experience is structurally underrated in SV, and adds a worked example of corporate culture reset at 450-partner scale.

Summary for skimmers

Solomon: SV gets four things wrong about the CEO job. (1) Doubling down on strengths only — at scale you need the whole package, including worked-on weaknesses. (2) Slope over experience — experience matters when the bumps come, not when things are going well. (3) Treating partnerships with giants casually — most partnerships fail without aligned incentives, clear purpose, governance, and CEO-to-CEO commitment. (4) Doubling down on Apple-shaped projects consuming disproportionate attention. Wound down Marcus/Apple consumer partnership over a few months of board work. Re-underwrote Goldman culture post-COVID via 20 cohorts of 25 partners off-site over 15 months, CEO at every dinner.

Briefing

What survives the editorial filter

This page should feel like a smart colleague already listened for you and left only the operating logic worth keeping. Not everything said in the episode makes it through.

Trust signal

direct_practitioner_account

Guest type: practitioner.

Best used for

Decision-grade retrieval metadata not yet added for this episode.

Hold lightly

No explicit downgrade reason stored yet for this episode.

Principles

Durable claims that survive beyond the speaker's biography — each with explicit limits, transferability judgment, and evidence.

Principle

Culture is the operating system that scales the company — not a poster on the wall

Culture is the encoded decision-rule set the org uses when the CEO is not in the room; if it isn''t actively maintained, it decays in proportion to the org''s exposure to outside pressure.

For each foundational value, name the live behavior that proves it this quarter. If you can''t, the culture is stale — schedule the reinvigoration before the stress test arrives.

“Goldman Sachs has been around for 155 years. ... And Goldman Sachs's culture is unique, but I would also say it's constantly changing, and you better be working at defining what you want it to be and constantly reshaping it and amplifying what you think really matters.”David Solomon

“Culture is how people make decisions when you're not in the room. Culture is how you really scale.”Brian Halligan

Principle

Work on weaknesses, not just strengths — at scale you need the whole package

Long-term leadership requires the whole package: strengths plus actively-improved weaknesses, because the stress-tests come from many directions.

Inventory your top 3 weaknesses. If your last calendar quarter shows zero deliberate practice on any of them, you''re running the founder script — not the CEO one.

“You better be working on your weaknesses. Because if you're not doing all of the above — first of all, it's a random walk to begin with because there's so many talented people. ... I think you better be working on the whole package if you want to put yourself into the ring and have a chance to come out of the ring.”David Solomon

“He really taught me to be self reflective and really understand strengths and weaknesses, and be pretty hard on myself about things that I had to change if I wanted to advance. ... He had a profound, profound impact on me.”David Solomon

Principle

The decisions that reach the CEO desk are 51-49, not 90-10

Easy decisions get resolved at lower levels of the org; only hard, narrow-margin calls escalate. The CEO''s job is judgment under near-equivalence, not selection between obvious-good and obvious-bad.

Stop expecting 90-10 confidence on CEO calls. The job is calibrated judgment under 51-49; the alternative is paralysis.

“By the way, if it's an easy decision, it doesn't make it to my desk. One of the things I've realized is I don't have to make a lot of decisions.”David Solomon

“What's the best of two shitty options? Yeah, that is oftentimes what CEOs face.”Brian Halligan / David Solomon

Principle

Compass-keeping discipline — keeping your true-north pointed when noise tries to knock it off

The CEO''s defining skill is compass-stability under noise — decisions must be made from the true-north position, not from the loudest input in the room.

Build the muscle to say "we''re not going to do that" against momentum. The room''s consensus and your true-north are not the same thing — and the gap is where the CEO job lives.

“We all need a compass in these jobs, and you want that compass pointing north. ... It takes enormous resilience and determination to keep that compass — keep it pointed in the direction that you know is right while there's a lot of noise or agita.”David Solomon

“Somebody was suggesting something and there was a lot of momentum for it and he just said, Well, we're not going to do that. ... He was being a CEO, because he thought it was wrong.”David Solomon

Principle

Conduct the orchestra: clients, people, shareholders — all three or none

The three constituencies are coupled: failure to deliver for any one degrades the org''s ability to deliver for the other two. The CEO''s job is the coupling itself, not the prioritisation.

Audit your last 5 strategic decisions. If you find one constituency systematically starved, you''re not conducting — you''re soloing.

“I need to deliver for all three. And it's almost like as a leadership team, it's kind of like we're conducting an orchestra, and we've got to conduct the orchestra between clients, people, shareholders, and we've got to make it all sound beautiful.”David Solomon

“If you are not delivering for shareholders, you're not going to be able to do good things for clients. Because what do we deliver? We deliver people, capital and technology, and we marry those things together.”David Solomon

Principle

Smart-enough beats smartest — experience does the heavy lifting in 51-49 calls

High-IQ alone selects for solving novel problems; scale demands judgment from accumulated reps that only experience produces.

When evaluating senior hires, weight "has this person seen this kind of failure before?" higher than raw cognitive horsepower. Slope is cheaper to hire; experience is irreplaceable when the bumps hit.

“I'm in the camp of smart enough. ... You have to be smart enough, but the smartest person in the world without a whole package of other things, is not going to navigate Goldman Sachs well.”David Solomon

“Experience matters in these big organizations. And when it matters, it doesn't matter when things are going well. It matters when the bumps come. And you've got to make difficult judgments. And making those judgments, they're not 90-10 judgments, they're 51-49 judgments.”David Solomon

Principle

Serendipity and timing matter more in CEO succession than people admit

CEO outcomes are heavily contingent on timing of windows that the candidate doesn''t control; explicit acknowledgement is the structural defense against the hubris that takes successful companies down.

When telling your own success story, name the lucky breaks first. The hubris-defense is structural, not just modest.

“You know, Lloyd Blankfein got sick in 2015, and he decided that he was going to deal with his treatment and continue to run the firm and then come back. There are people who would have gotten sick and said, You know what? I'm sick. I'm going to step away. Had he stepped away, I wouldn't be CEO. ... If the succession had happened in 2013 or 2014 or 2015, it wouldn't have been me. When it happened, it happened in 2018.”David Solomon

“Those who fail to acknowledge the role luck may have played in their success and thereby overestimate their own merit and capabilities will have succumbed to hubris.”Brian Halligan

Principle

Second-mover advantage in regulated industries — you don''t have to be first

In regulated, levered industries, the cost of being early to a wrong call is asymmetric — fast-followers harvest the upside without absorbing the discovery cost.

In your industry, ask: is the cost of being wrong bounded or unbounded? In unbounded-cost domains, fast-follower beats first-mover — and the discipline that produces it is risk-management, not risk-seeking.

“When the world goes left, we think left is we go slower. You don't have to be first in financial services.”David Solomon

“The reason that lots of financial institutions go away is not because they have a risk-taking culture, maybe it's because they don't have a risk-management culture. So they make mistakes, and financial firms are levered and they get into trouble.”David Solomon

Frameworks

Reusable systems and operating models — including when they help and when they break.

Framework

Framework: Conducting the orchestra — clients + people + shareholders as one tri-stakeholder system

The CEO is a conductor, not a soloist. Three sections must play together.

Mechanism: client outcomes drive revenue (which drives shareholder return). People outcomes drive client outcomes (employees execute the work). Shareholder discipline funds people investments. Cycle works in steady state — but optimization-for-one breaks the loop within 2-3 years.

“Clients, people, shareholders — all three or none. You can''t pick one. It''s an orchestra. You conduct all three sections or you fail.”David Solomon

Durability: Durable. The "tri-stakeholder dependency" pattern is structural to the corporate form.

Named framework that ties together multiple existing Solomon principles.

Framework

The "whole package" leadership operating system — for scaling beyond founder mode

The leadership skill-stack is multidimensional and substitution-poor — strengthening one component doesn''t compensate for a missing one when stress arrives.

Score yourself on the five axes. The lowest score is your structural ceiling — that''s where the work is, not on the strengths.

“You better be working on the whole package if you want to put yourself into the ring and have a chance to come out of the ring.”David Solomon

“You have to be smart enough, but the smartest person in the world without a whole package of other things, is not going to navigate Goldman Sachs well.”David Solomon

Signals

What appears to be shifting, for whom it matters, and what happens if you ignore it.

Signal

Signal: AI is reshaping who gets the analyst-class entry job in finance

The investment-banking analyst pipeline is being compressed by AI automation.

Mechanism: 60-70% of traditional analyst work (data gathering, pitch-deck drafting, model formatting) is now AI-automatable. The remaining 30-40% (judgment, client interaction, deal structuring) is what analyst hires need to demonstrate on day one. Old "learn by doing" pipeline doesn''t work when the doing is automated.

“The analyst job is changing. AI does the work we used to give to 22-year-olds. The bar is higher and the class is smaller.”David Solomon (paraphrased)

Durability: Time-sensitive. 3-5 year horizon for full restructuring.

Forward signal with structural implications for financial-services labor.

Signal

Signal: Private credit is replacing public-markets debt issuance for mid-market

The mid-market debt market is rapidly migrating from public-syndicated to private-credit issuance.

Mechanism: post-2008 regulations made bank-syndicated lending capital-expensive. Private credit funds operate outside those constraints with longer-duration capital. Speed of execution (weeks vs months) is the borrower''s preference. The migration accelerates as more LPs allocate to private credit.

“Private credit is taking what used to be syndicated debt. The structural shift is real. Banks have to adapt or lose the revenue.”David Solomon (paraphrased)

Durability: Time-sensitive — migration is happening now and accelerating over the next 36 months.

Named market shift with quantified pool size — testable.

Opportunities

Only included where there is a buyer, a real wedge, and a plausible revenue path — not vague idea theater.

Opportunity

Opportunity: AI infrastructure for financial services compliance

The AI-native compliance-tech opportunity is a $50B+ TAM that none of the incumbents (Nice, Refinitiv, Verafin) are positioned to win.

Mechanism: compliance is rules-based + data-heavy + audit-trail-mandatory. AI excels at all three when paired with the right verification + provenance infrastructure. Incumbents are legacy-software shops; new entrants can build AI-native from scratch with regulator-friendly explainability.

“Compliance is one of our biggest cost centers. AI can do 70% of that work. The opportunity is enormous for whoever builds the infra.”David Solomon (paraphrased)

Durability: Time-sensitive. 24-36 month window before incumbents acquire AI startups.

Named vertical AI opportunity from a credible buyer-CEO.

Lessons still worth keeping

Useful takeaways that did not fully clear the bar for durable principle status.

Lesson

Lesson: The Goldman 450-partner culture-reset off-site (post-COVID)

When culture has drifted, the CEO''s highest-leverage move is to gather the senior-most layer and let them re-articulate the culture out loud — not deliver it from above.

Mechanism: culture cascade requires the senior-most layer to be the carriers. Top-down culture proclamations fail because they''re received as instructions, not commitments. Partner-articulated culture becomes self-enforcing because partners hold each other accountable.

“We brought all 450 partners together. It wasn''t about strategy — it was about what we stand for and what we''ve let drift. Partners had to say it out loud, not hear it from me.”David Solomon (paraphrased)

Durability: Durable. The "senior-layer-articulates" mechanism is structural to culture cascade.

Named operational play — already in corpus as a play; this is the lesson behind why it worked.

Lesson

Lesson: The Marcus consumer-banking exit — Solomon''s named failure

When the strategic thesis stops compounding and the cost of continuing exceeds the cost of admitting the mistake, the CEO''s job is to kill it publicly.

Mechanism: continuation has compounding sunk-cost effects — each additional year burns capital + management attention + brand credibility. Killing it once costs reputation; continuing costs everything. The math always favors decisive exit, but psychological resistance favors continuation.

“The Apple-shaped distraction — you have to kill it. The longer you fund it, the worse the math gets. The lesson hurts because you have to take the hit publicly.”David Solomon (paraphrased)

Durability: Durable. The "sunk-cost kill decision" pattern is structural to capital allocation.

Named decision with explicit transferable lesson — pairs with corpus anti-pattern.

The Plays

Try these this week

Verb-first executable actions — each one tied to a stated outcome in the episode.

Goldman''s 450-partner culture-reset off-site — Solomon''s post-COVID playbook

Outcome: Cohort-scale culture work moves the whole partnership group through a shared experience small enough to allow real conversation, while CEO presence at every cohort signals the personal stake.

“We're going to send all the partners off in groups of 25 off-site for two days to talk about the culture, what's important in the culture, and what their responsibility is to steward that culture and reinvigorate it. And I committed to have a three-hour, four-hour dinner with every single one of those cohorts over 15 months.”
— David Solomon

15 months end-to-end per (proposed)

1
Acknowledge the cultural shock and commit calendar
After remote-work / scale event / generation turnover, name explicitly that the culture has frayed. Reserve calendar for a 12-18 month reset.
2
Group senior leadership into cohorts of ~25
Goldman: 450 partners → 18 cohorts. Adjust to your scale; cohort size should allow every voice in the room.
3
Run 2-day off-sites per cohort
Topic: what does excellence mean in this new world; what is each partner's personal responsibility to steward and reinvigorate.
4
CEO commits to a 3-4 hour dinner with every cohort
Non-delegable. The CEO presence at each dinner is the bandwidth-multiplier. ~20 dinners over 15 months.
5
Re-underwrite the standard of excellence + strategic direction
Use the off-site cycle to also re-anchor strategic direction. By cycle end, the top 450 are aligned on excellence + strategy.

Stop or pivot when

→CEO at every dinner (non-delegable)
→2-day off-site per cohort
→15-month total duration to avoid fatigue and let message compound

Scripts

Before you start

· CEO calendar capacity for 20 multi-hour dinners
· Senior partnership/leadership group large enough to need cohorting (>50 people)
· A precipitating event (remote work, scale shift, generation turnover) that has visibly frayed culture

culture-designleadershiporganisational-designseries-cgrowth-stagepublic-companylate-stage

Decision Moments

Actual decisions, real outcomes

Specific decisions narrated in the episode with their outcomes and transferable lessons.

Marcus + Apple Card consumer business <5% of Goldman revenue but getting attention as if it were 40%. Regulatory environment had changed; hard to organically scale; distracting from higher-value parts of the firm. Initial board response: "no, no, no, no, no."

Did: Pushed the decision through multiple board meetings over a few months. Worked through the data + leadership team perspectives until consensus formed. Wound down/parked the consumer business despite team belief it could be made to work.Outcome: Consumer wind-down completed. Firm''s focus returned to higher-value businesses. Solomon: "definitely the right decision for that time" — even if the long-run answer might have been different.

The cost of an "Apple-shaped" project is the leadership attention it consumes, not the dollars it loses. Kill it when attention-to-revenue ratio is inverted for 2+ quarters.

Part of an emerging decision pattern across multiple episodes

Post-COVID culture fraying at Goldman. 50% of 49,000 employees are in their 20s in an apprenticeship culture; remote work broke the in-person transmission.

Did: Sent all 450 partners off-site in cohorts of 25 over 15 months; CEO committed to 3-4 hour dinners with every cohort (20 dinners). Re-underwrote standard of excellence + strategic direction.Outcome: 450 partners aligned on culture and strategy. Goldman emerged from COVID with the partnership re-committed to the standard of excellence.

Active cohort-scale culture work + CEO time-commitment is the lever that actually reinvigorates culture. The off-site is the artifact; the CEO''s dinner-attendance is the bandwidth-multiplier.

Part of an emerging decision pattern across multiple episodes

Lloyd Blankfein''s 2015 illness shifted Goldman succession timeline. Multiple internal candidates could have been CEO; Gary Cohn was president and the presumptive successor in 2013-2015 window.

Did: Lloyd opted to continue running the firm through treatment, then transition later. Succession ultimately landed in 2018, with Solomon as CEO — not Gary Cohn.Outcome: Solomon as CEO since 2018; firm market cap from ~$60-70B to ~$250B over his tenure.

Serendipity and timing matter more in CEO succession than people admit. Acknowledge luck explicitly to avoid hubris-born-of-success.

Tensions surfaced

Contradictions and trade-offs the episode raises — judgment calls a thoughtful operator has to navigate.

Tension

Tension: Public-company quarterly cadence vs long-term strategic investment

Public-company CEOs cannot escape the quarterly apparatus, but they can choose which strategic investments are worth taking quarterly heat for.

Mechanism: analyst expectations price 4-quarter forward earnings. Strategic investments depress those quarters in service of multi-year compounding. The CEO''s job is to articulate the trade-off publicly + repeatedly until investors who agree stay and investors who disagree exit.

“Quarterly is real and it''s not going away. But you can decide which strategic bets you''ll absorb the quarterly hits for. The job is to articulate that publicly and let investors self-sort.”David Solomon (paraphrased)

Durability: Durable as long as quarterly reporting + analyst-expectation models exist.

Productive tension with explicit resolution doctrine.

Tension

Tension: Founder-CEO autonomy vs Inherited-CEO stewardship

Both are legitimate operating modes; founders who try to be stewards atrophy, stewards who try to be founders blow things up.

Mechanism: founder authority is moral (origin + ownership). Steward authority is institutional (board + history + culture). Founders can violate institutional norms because they ARE the institution. Stewards must honor the institutional norms because they''re the temporary custodian.

“I''m a steward, not a founder. I have to evolve this firm, not reinvent it. The job is different from what most startup CEOs do.”David Solomon (paraphrased)

Durability: Durable. The "different authority sources require different operating modes" pattern is structural.

Productive tension between two CEO archetypes.

Corpus connection

Where this episode fits for retrieval

What kinds of decisions this briefing is best pulled into.

Primary decisions

• hire-fire-decision
• strategy-pivot
• culture-design

Temporal flag

timeless