YC Visiting Partner Max Kolysh: Fall Batch Kicks Off Next Week, $1 Million ARR Possible in 12 Weeks
Y Combinator visiting partner Max Kolysh announced that the fall batch will kick off next week. He has just set demo day goals for the companies he leads, including notable targets of $1 million in annual recurring revenue, $3 million in revenue, and 1 million daily active users. Most companies start close to zero, and he believes a lot can be achieved in 12 weeks.
Kolysh is a two-time YC founder: Zinc, an e-commerce interface company from W14, and Dover, a recruitment platform from S19, which serves over 1,500 companies including Stripe, Scale AI, and Vanta. He became a visiting partner in January 2026, partnering with group partner Pete Koomen, starting from the spring class. In the Startup School course, he emphasized that the first ten customers should come from former colleagues, classmates, and referrals, rather than building growth channels first.
YC's recent batches have turned the "$1 million ARR by demo day" from an exception into a statistical phenomenon. President Garry Tan stated that the number of companies reaching $1 million in annual revenue in W26 is about three times that of W25, with nearly 200 companies in the batch averaging about 14% weekly growth, setting a record for the accelerator. Additionally, public statistics indicate that over a dozen companies reached $1 million ARR before demo day without institutional venture capital funding. At the same time, some companies that have already reached $1 million ARR still choose to give up about 7% equity to join the batch and continue to double their revenue.
The demo day goals compress three different business models into the same time frame: subscription software looks at ARR, transaction or contract businesses look at cumulative revenue, and consumer products look at daily active users. A million daily active users and a million ARR are not on the same balance sheet, yet they are used as the same score line that "investors can understand in 12 weeks."
In market mechanics, this is the accelerator selling a time box to founders and LPs. The buyers want to present a repeatable metric on demo day that surpasses nearly 200 competing teams; the sellers are YC's batch discipline and alumni checks. The beneficiaries are AI workflow companies that can turn zero into an annualized number in 12 weeks; the pressured parties are teams still refining their products and unable to present a weekly growth curve. Funding flows into projects with "existing scores" from alumni and professional seed funds around demo day, rather than into the narrative itself.
Kolysh also reminds to first build a warm network for customer acquisition. The average high growth across the W26 batch does not mean every company can claim a million ARR or a million daily active users.
Source: Public Information
ABAB AI Insight
YC has shifted the three-month period from "learning how to build a company" to "submitting an annualized report card." Kolysh's goals for the fall class are to turn the extreme values already seen in W26 into assignments for the next class. $1 million ARR, $3 million in revenue, and 1 million daily active users correspond to three different machines but share the same deadline. This makes the accelerator resemble a short-term hedge fund's performance evaluation rather than a school.
The capital path is an auction on demo day. The average 14% weekly growth turns "having revenue before securing institutional funding" into a comparable snapshot. Alumni funds and co-investors like Pioneer take the first shot, creating momentum, and then let slower institutions take over. Dover proves he knows how to go from zero to paying customers; his role as a visiting partner turns his customer acquisition course into a batch operating system. AI tools have reduced the time to create the first billable version, thus raising the score line rather than eliminating it.
Analogous to sports preseason roster decisions and hedge fund monthly drawdown lines: it’s not about whose story is better, but who delivers auditable numbers within a fixed window. The industry phase is that generative tools compress "from zero to demonstrable revenue" into a quarter, and seed valuations begin to be priced based on pre-demo day curves rather than founders' resumes. The fact that teams with $1 million ARR still come to YC indicates they are buying batch liquidity and controlled experiments, not just startup checks.
Structural judgment shifts pricing power from narrative to observable growth within the time box. The mechanism is: nearly 200 companies present on the same day, and investors can only rank them using a few integers; those who make weekly growth a habit can explain less in the auction. A lot can be achieved in 12 weeks, provided the product already has a billable slot. Going from zero to a million is the new norm after AI has disrupted production costs, not a universal physical law.
ABAB News · Cognitive Law
- The accelerator sells deadlines, not classrooms.
- How much can grow in three months depends on whether billable slots are already in place.
- The same score line will force out three completely different companies.