VC Renata Quintini, co-founder of early-stage VC Renegade Partners, said on stage at TechCrunch Disrupt last week that 2025 funding will continue to be a “tale of two cities.”
“While some companies do have the promise of going after large, rapidly growing markets, [will be] It's gaining a lot of funding and momentum. But on the other hand, Quintini warned that “companies that need to build real businesses and efficient businesses” will struggle to raise capital.
She is referring to the tight funding market faced by startups in an era of high interest rates. In 2023, around 3,200 startups died due to easy funding during the 2021 party. In 2024, venture capitalists rushed to raise money for AI companies, many other early-stage startups struggled to get funding, and many more deaths occurred. In 2024, fintech was particularly cruel.
As 2025 rolls around, your best chance of attracting venture funding will be with a solid business foundation. (Unless you're a world-renowned AI scientist, for example.) This means selling products and services that serve a sizable customer base at a profitable price.
But wait! VC Colin Reilly, a partner at Greylock, warned on stage that it's not just about acquiring paying customers.
Riley said there are “no clear milestones” for sales or growth rates that will attract VCs' attention next year. “There's no key number that says if you hit that you'll have a successful Series A. What we're looking for is really the quality of ARR.” [annual recurring revenue]Not the amount of ARR. ”
In other words, once a customer gets on board, will they never leave? Do they tend to increase their spending on startups over time? This startup may actually have fewer customers and less revenue than its competitors, but if the customers who sign up continue to stick around, they will invest more money. The house would cut the check.
“What we're looking for is a quality customer base that will come back again and again as the funds grow,” Riley says.
“When VCs talk about 'we need to build a moat,' this is what they mean,” Hustle Fund co-founder VC Elizabeth Yin said on stage. This is a way to explain how to lock in your customers so they don't try to leave. “The more unique things you can do that others can't do, the more it will help you.”
Riley cited as an example a Greylock portfolio company called Braintrust, which helps developers build AI apps and evaluate their performance. Greylock was confident of leading the $5 million seed deal after founder Ankur Goyal secured early customers who were “industry tastemakers.” His clients included Zapier, Coda, Airtable, and Instacart, which she described as “people on the forefront of building products with AI.”
If powerful customers with strong connections are satisfied with the service, they will attract other similar customers, creating a cycle. Braintrust has since acquired other well-known technology names, including Stripe and Notion, the company said. And in October, Andreessen Horowitz's Martin Casado led a $36 million Series A, with Greylock also participating.
This “quality of the customer” is always important to investors, which is why it is important to name the customer. But in 2025, after the AI gold rush, this metric will become even more important as much of the revenue generated by many AI startups will be one-time revenue.
Another panel at Disrupt was about bouncing back after a down round, with VC Elliott Robinson, a partner at Bessemer Venture Partners, explaining the situation. In early 2024, nearly every large company's board of directors was struggling with AI and allocating large exploration budgets to CIOs to buy and research products.
“Now, we have been buying AI products for 18 and 24 months and have seen companies grow from 0 to 4. [million dollars in revenue] and from 4:00 to 20:00 [million]” he said. “The question now is what will be renewed? Because CIO budgets are starting to dry up.”
CIOs will only continue to buy things that make a measurable difference. Therefore, the sheer number of AI startup revenues and customers (which may also include customers in AI-adjacent fields such as application monitoring) means that a particular startup is a good bet for the future. It's not a thing.
Or as Quintini explains: “At the end of the day, what you're trying to do is, one, build something complex. And two, get into a business that you run faster than others, or that others can't imitate.” Either you do something.”