Episode Summary
Executive Summary: Keith Rabois argued 2023 would be a normalization year after the 2021 bubble: inflation, labor costs, and recession risk will determine the macro path, while startups must cut burn, plan multiple scenarios, and raise on far more disciplined terms. He also discussed hiring dynamics, VC entitlement, founder coaching, AR/VR, ChatGPT’s promise and limits, IP/legal risks, and a bearish view of Trump’s political future.
Main Topics: 2023 Macro Outlook and Scenario Planning: Rabois said inflation and labor costs are the key macro variables, making 2023 hard to predict. Founders should plan for several possible environments rather than single-point forecasts, and keep burn and hiring flexible. Startup Valuations Return to Historical Norms: He argued the market is not uniquely bad; it is reverting to long-term averages after irrational 2021 pricing. Seed, Series A, and Series B valuations should be much lower and more disciplined. Big Tech Hiring Freezes and Startup Talent: Layoffs and freezes at Google, Meta, Salesforce, Microsoft, Apple, and Twitter reduce employee entitlement and strengthen startups’ recruiting position. He prefers founders hire people who would not choose big tech. VC Entitlement and the Harder Job of Investing: Rabois criticized the surge of new venture firms and said many were built on easy markets, paper markups, and low discipline. He emphasized that VC success requires post-investment support, judgment, and multiple market cycles. Founder Management: Focus vs. Distraction and Conflict: He advised founders to simplify, concentrate resources, and avoid spreading talent across too many projects. He also noted that abrasive founders can succeed if friction is intentional and tied to mission-critical priorities. AI, ChatGPT, and Legal/IP Concerns: Rabois saw ChatGPT as impressive but constrained by cost and uneven quality. He warned that AI products raise fair use, derivative works, and creator compensation issues, especially when commercialized. Consumer and Enterprise Tech Bets: AR/VR, Google Search, OpenDoor, and Media: He was skeptical of Meta’s VR strategy, more open to AR’s industrial uses, and said Bing+ChatGPT is more of a marketing threat than a substantive Google threat. He also discussed OpenDoor’s operational resilience and the need for founders to communicate directly amid adversarial media.
Key Arguments: Inflation is the central macro variable; if it stays high, the Fed may tighten further, and if it falls, recession risk still remains. Founders should stop assuming perpetual growth and instead build 2-4 scenario plans with editable hiring and burn decisions. 2023 fundraising terms should reflect normalized historical tech valuations, not 2021 bubble pricing. Big Tech freezes and layoffs reduce salary-comparison leverage and make startup recruiting easier for the right kind of candidate. The best startup hires are people who would not be attractive or willing to work at Google/Facebook/Amazon/Apple/Netflix. Most new VC firms lack the ability to return capital, survive rebalancing from LPs, or do the operational work required after the check is written. Abrasive founders can work if the conflict is deliberate and tied to important company objectives rather than collateral damage. ChatGPT is impressive but will not become broadly productized until costs fall by roughly an order of magnitude and outputs become more reliable. AI-generated content raises fair use and derivative-work questions, especially when commercialized or when it replaces paid creative labor. Meta’s core culture is better at iteration and mimicry than inventing new paradigms, making VR a poor fit and AR a better, though still difficult, opportunity. Bing embedding ChatGPT may help consumer perception, but it does not solve Google’s substantive search advantage on most queries. OpenDoor’s challenge is as much communication/optics as execution; in many markets it is already gross-margin positive. Founders and companies must communicate directly, but not rely on owned channels alone; they still need earned media and engagement with skeptical audiences.
Data Points: Seed valuation guidance: $10 million - Rabois cited this as roughly the normal maximum for a seed round in the current market. Series A valuation guidance: $20 million - He said this is the normalized benchmark for a strong Series A. Series B valuation guidance: $40-$50 million - He described this as the upper range for a good B round in normalized tech investing. SOC 2 compliance speed with Vanta: 2-4 weeks - Podcast ad copy contrasted Vanta’s timeline with manual compliance. SOC 2 compliance speed without Vanta: 3-5 months - Used as a comparison to emphasize Vanta’s efficiency. Acquire marketplace monthly messages: 35,000+ - Described buyer-seller activity on Acquire.com. Average 12-month revenue of listed startups on Acquire: almost $600,000 - Shown as evidence of marketplace quality and activity. Buyers on Acquire: 120,000+ - Indicates the scale of the acquisition marketplace. OpenDoor markets with gross margin positivity: 48 of 52 markets - Rabois said only four markets were not gross-margin positive. U.S. home transactions in a normal year: ~5 million - He used this to explain OpenDoor’s dependence on transaction volume. U.S. home transactions during the GFC: ~4 million - Provided as a stress-case comparison for real estate activity. ChatGPT query cost (fully loaded): 10-12 cents - Rabois estimated current cost as too high for broad productization. Needed AI cost improvement: ~10x reduction - He said the cost structure must fall by at least an order of magnitude. Meta layoff scale: 10,000 people - Referenced as a major first step in Meta’s restructuring. Opendoor stock price reference: 94 - He said he bought the stock when Meta announced the 10,000-person cut; used as a trading example. OpenDoor market count: 52 markets - Referenced while discussing the company’s operational footprint. Big Tech market share context: Google ~90%, Bing ~8-9% - He used this to frame search competition and Bing’s challenge.
Pivotal Quotes: "You have to plan for two to three or four different environments and have a burn rate, have a fixed cost structure, have hiring plans that can be edited on a quarterly basis at a minimum." — Keith Rabois: Advice to founders on scenario planning in an uncertain 2023 macro environment. "If you can't return capital... in the hottest market in the history of technology, there's a question of when the hell you're ever going to return capital." — Keith Rabois: His critique of many newly formed VC firms and their lack of durability. "You want to ensure that the friction you're causing is not collateral damage, but it's intentional." — Keith Rabois: Guidance for forceful founders on how to channel intensity productively.
Implications: Founders should operate with capital discipline, flexible hiring, and honest pricing expectations. VCs must prove real judgment and returns. AI, AR, and search will create opportunities, but legal, cost, and product-market constraints will shape who wins.
From the Episode
Easiest way to find for yourself whether you know we've stabilized inflation or not. Even if you do stabilize inflation, then you have to worry about a consumer recession that's like severe, which we haven't encountered yet. Consumers are still spending, they're spending a little bit more with debt than with savings, but they're spending and spending more on some kind of in some categories like travel. They're spending more than they really ever have. So, if we do stabilize inflation and consumers stop spending, that actually is a bad thing, too. So, nobody really knows. What's going to happen in 2023? So it's pretty hard to predict. The asset bubble stuff was very easy to see. It was very easy to call. You know, there's really only one tool left in the Federal Reserve's toolkit, and they started using it. They may have to use it a little bit more. You know, the employment data is still pretty hot, which is typically drives inflation. The cost of labor is a major input to most things we buy. And so if the cost of labor is going up, inflation is probably going to continue for a while.
Almost every one of those venture firms started the last five years is going to fail. There's maybe two or three that might survive. The reason why is you just, there's two reasons, three fundamental drivers, actually. First is if you can't return capital, I don't mean mark up things on paper, return capital in the hottest market in the history of technology, there's a question of when the hell you're ever going to return capital. So anybody who didn't put literally points on the board and distribute money to LPs should not be raising future venture runs. And I think LPs are going to be out. About that. Like, if you literally couldn't mint money between 2019 and 2021, you're never going to mint money in technology because things aren't going to be that good and that easy. Second, a lot of LPs have lost money in the public markets and they have a capital allocation balance of how much money can be in private assets, period, venture as a private asset class. And so they have to, quote unquote, rebalance their portfolio, which means they have to be more selective about which VCs they fund. So that's going to drive more selectivity and more scarcity in venture. And then third,
Successful with that kind of DNA because everybody's opting into your cult. However, the one piece of feedback I would give somebody who is of that ilk, and I actually work with founders like that, I've worked with founders like that for 23 years. I've worked for founders like that. I've worked with founders, I've funded founders on the board of founders like that. The one piece of feedback is: you want to ensure that the friction you're causing is not collateral damage, is not collateral damage, but it's intentional. And sometimes those two things get mixed. So, one of the founders I work with once asked Mike Moritz, who I consider to be an amazing investor, what's the most consistent, what's the most important trait for a successful founder in his entire career? And he said, they relentlessly apply force. So that's very similar to being quote unquote toxic in the world. Relentless application of force people do not like, but that's what makes founders unreasonably successful.
About This Week in Startups
Jason Calacanis covers startups, tech, markets, media, and all the hottest topics in business and technology. He also interviews the world’s greatest founders, operators, investors, and innovators.