ZIRP was bigger than we thought / AI might be less disruptive than we think
Gil Dibner

I want to share two interesting thoughts that crossed my desk this week. The first relates to the impact of ZIRP (bigger than we thought). The second relates to the impact of AI on industry (perhaps less disruptive than we think).

ZIRP. Over the past two years, the term “ZIRP” (zero interest rate policy) became popular as people looked for a way to describe the impact of loose monetary policy on decision-making of all types. We’ve all come to use the phrase “ZIRP phenomenon” to describe inflated VC rounds, inflated salaries, generous workplace perks, gravity-defying valuations, and logic-defying business models (NFT-based applicant tracking system? hotel on the moon? Sounds good!). But a recent chat with a very wise founder (whom I’m not yet authorized to identify) left me thinking that the impact of ZIRP may have been far wider and more insidious than I thought. “SDRs,” he said, “are a ZIRP phenomenon.” Wow.

Now, I don’t think he meant that literally all SDRs (sales development representatives) were erroneously employed due to low rates. But here’s his point: SDR success rates are very low. Suppose an SDR is paid $50K per year, makes 100 calls a day for 260 workdays per year and manages to connect and sell once per week (52 per year) at a $1000 ARR ACV. That generates $52K in incremental ARR. Even at very high gross margins, that’s probably not enough to justify the cost. However, now imagine if (due to ZIRP) the market is paying 100x ARR multiples for startups. Under that assumption, the incremental $52K of ARR increases the company’s perceived valuation by $5.2M. The example might have some holes in it, but the broader point holds water: ZIRP jacked up the multiples on revenue so high that lots of things that might not make much rational sense suddenly felt entirely rational. An interesting implication of this is that a lot of the operational startup playbook we are currently running with was actually written during ZIRP days. Consequently, much of the conventional wisdom about the sorts of activities founders should conduct and investments founders should make may actually be mistaken in today’s more normal environment.

Disrupting AI. The term “disruptive” as it relates to technology is older than “ZIRP” and dates back to 1995, when Clayton Christensen coined it. It came up today in a discussion with my partner David when he highlighted how so many large public tech companies are emphasizing their “AI strategies” on their quarterly earnings calls and have been doing so since 1Q if not before. This intense focus on AI from companies like Google, Microsoft, Amazon, and many many others — coupled with their unique access to large volumes of proprietary data — suggest that they will be able to execute effectively on the AI imperative. Tech incumbents clearly know about AI (they will not be taken by surprise). They can clearly implement the technology (it is mostly open source, widely available on a commercial basis, or developed by them). Finally, they clearly have an edge over challengers due to their access to large volumes of training data from years of prior activity, something no challenger will have. If so, many (most?) of the likely winners in the AI battle are incumbents, wouldn’t that suggest that today’s wave of AI enthusiasm is actually the precise opposite of disruption innovation? This has a name, by the way. Christensen called it a “sustaining innovation,” innovation that strengthens the hand of incumbents. The more applied AI startups I see in this present wave, the more concerned I become that we are collectively misapprehending a sustaining innovation with a disruptive innovation. I want to give David the credit for saying it first and succinctly: “Given the alertness and advantages of incumbents, AI may not be that disruptive after all.”

As a venture firm, we can invest successfully in both sustaining innovations and disruptive innovation, but it’s important to know which is which. Our view is that a wave of sustaining innovation (such as AI/LLMs/GenAI) creates strong opportunities for infrastructure players, but less so for application-level players. When it comes to application-level companies, we are laser-focused on making sure that we identify truly disruptive innovation. Once we recognize the AI wave as sustaining versus disruptive, the flurry of entrepreneurial and venture capital activity we are witnessing today appears more bubble-like and less compelling.

If you have plans to do something truly disruptive in any industry, we want to hear from you.

Regards,
Gil

PODCAST

US Immigration Best Practices
Jennifer Schear, Founding Partner, Schear Immigration Law Firm

Learnings from Co-Founding Peakon, Podio, and Future Five
Kasper Hulthin, Co-Founder, Peakon, Podio, Future Five

The Evolution of Collibra’s Product Positioning & How They Created a Category
Stijn “Stan” Christiaens, Co-Founder & Chief Data Citizen, Collibra

Lessons Learned From Investing Early in Over a Dozen SaaS Unicorns
Jason Green, Founder & General Partner, Emergence Capital

FROM THE BLOG

Kafka Survivors of the World, Unite!
Why we backed Memphis.dev.

The Problem with Startup Advice
The best founders and investors know the rules, but also know when to break them.

Looking Back to Move Forward
How to survive this extraordinarily exciting and wildly disconcerting age of generative AI.

LLMs and the Future of Customer-built Software Design
How will LLMs change software development and design?

WORTH READING

ENTERPRISE/TECH NEWS

Is LK-99 for real? A team of researchers, led by Hyun-Tak Kim at the College of William & Mary in Virginia, have claimed to have created the first superconducting material that conducts electricity without resistance at room temperature and pressure. Named LK-99, the material was created by mixing powdered compounds containing lead, oxygen, sulfur, and phosphorus, then heating them to form a dark grey solid. The scientific community remains largely skeptical, pointing out that certain necessary signatures of superconductivity aren’t demonstrated in the data provided. (It appears that the relevant papers were a bit hastily drafted). Labs around the world have been seeking to replicate Kim’s results themselves and posting videos on Twitter and BilliBilli (e.g. this recent unverified video from a Chinese lab), but so far, attempts to replicate have fallen short.

The bots are coming. On the heels of open sourcing Llama2, Meta announced that it is planning to launch AI-powered chatbots with diverse personas on its social media platforms as early as next month. These chatbots, designed to engage in human-like conversations with users, will have different personas, including a travel advisor with a surfer’s persona and another speaking like Abraham Lincoln. Google and Amazon are joining in as well. Google plans to revamp its Assistant to be LLM-powered, similar to Bard. And Amazon hopes that some LLM superpowers might jolt its Alexa business back to life.

The weird world of LLMs. The best rundown I’ve come across of the 35 year history of large-language models comes from Simon Willison, British programmer, and current director of architecture at Eventbrite. It’s not exactly a quick read, but it’s deeply informative, well structured, and worth your time.

Big tech is the big AI winner (so far). US tech giants Meta, Microsoft, and Google recently reported strong quarterly earnings, driven by their respective AI-focused businesses, despite significant capital expenditure in AI and technology. Meta reported revenue of $32 billion, up 11% compared to last year, with daily active users increasing from 2.04 billion the previous quarter to 2.06 billion, defying analyst expectations. Microsoft and Google also reported positive earnings, with Microsoft’s shares slipping after a lower than expected revenue forecast, and Google’s shares rising following strong results.

HOW TO STARTUP

Shamir secret sharing. A fascinating story from Max Levchin, Paypal co-founder, about a bug that almost took down Paypal in the middle of the night. I would hate to give away the details, so I’ll let Max tell the story (please give it a read!), but it serves as a useful reminder that startups are often teetering on the edge internally, no matter how successful they look from the outside.

Down rounds on the rise. Turntide Technologies, a manufacturer of electric motor systems backed by Bill Gates’ Breakthrough Energy Ventures, is expected to slash its valuation by over 80% in a new capital deal. Quora, the internet answer site, and Nuro, the autonomous delivery startup, are planning similar moves. Companies slashed burn to extend their runways as long as possible, but their time is running out. The down rounds are coming. As a result, many former employees will have overpriced, almost worthless stock options, while some investors may have to infuse new cash into somewhat distressed companies to protect their stakes.

HOW TO VENTURE

A unicorn mass extinction event. Hopin, which raised $1B and was last valued at $7.8B, sold off its core events business and the CEO stepped down. Rapid-delivery company Gorillas, which similarly raised more than $1B, sold to Getir earlier this year. Sequoia-backed Clutter, a storage startup which had raised $300M, ran out of cash last month and sold for pennies on the dollar. These are three examples of many. These “emergency” M&A deals are becoming the new normal, as investors seek to capture whatever value they can from these companies which previously bolstered the TVPI they reported to their LPs.

The value of the check-in. Great post from Charles Hudson of Precursor on the value of the “check-in” meeting with a portfolio company. I thought this was a particularly good point:

“As an investor, I don’t know, ex-ante, which of the 10–12 meetings per year with a given company will be impactful. My experience is that neither I nor the company can evaluate the value of any of those meetings in real time; the value is often only known after the fact with the benefit of hindsight….But the more frequent the touch points, the less pressure there is on any conversation or check-in.”

Founders don’t owe their investors a check-in or an update. But Hudson does a good job of articulating why a check-in can be valuable to both parties.

PORTFOLIO NEWS

LightSolver’s CEO, Ruti Ben Shlomi was interviewed on CBS News where she discussed LightSolver’s ambitions to build the next generation of computers.

Sisense appointed Ayala Michelson Chief Product Officer and General Manager of its Israeli operation.