Metal sells founders a faster way to raise money, and claims quarterly growth of 30 to 80 percent

Metal says it has grown between 30 and 80 percent in each of its last six quarters. Compounded, that range means the thing it measures is now between about 4.8 and 34 times its size six quarters ago (1.3 and 1.8, each multiplied by itself six times). The spread is a factor of seven, and it exists because the company has given only the range. It has not said whether the figure is revenue, customers or something else.

The claim comes from Usman Gul, the Airlift cofounder who started Metal, in a social media post announcing a $4.5 million seed round in early September 2026. He also said Metal was on track to close its 2026 financial year with multi million dollar annual revenue. None of these figures has been independently checked, and the company has not named customers or published prices.

What the product does

Metal calls itself an operating system for fundraising. Its website offers to help founders find investors, build a pipeline of them and automate outreach. On a podcast episode released on 17 December 2025, the company said its tool draws on data from more than 500,000 venture rounds, and that it helps founders avoid wasting 80 percent of their investor meetings. The 80 percent figure is a company claim with no method attached.

The technical idea is matching. A venture investor typically backs particular stages, sectors, cheque sizes and regions, and a founder who pitches outside that range loses a meeting that was never going to lead anywhere. Past rounds are the public evidence of where each investor actually writes cheques. A tool that ranks investors against a company's profile is, in effect, a filter over that history, followed by software that sends and tracks the outreach. The saving is time, which is what a founder with a few months of cash left has least of.

The shape of the round

Andreessen Horowitz's Speedrun programme and Y Combinator led the round. Pioneer Fund, Rebel Fund, Gaingels, Indus Valley Capital, Team Ignite Ventures and Phaze Ventures also took part, according to Gul. Rebel Fund, which focuses on Y Combinator companies, had announced a strategic investment in June. Speedrun's own listing puts Metal in its fifth cohort, founded in 2023 and with 22 employees.

Speedrun's listing gives Los Angeles as the company's base. Gul's earlier company, Airlift, was a Pakistani quick commerce startup, so Metal is a case of a Pakistani founder raising from two American programmes. It is evidence that outside investors will back such a founder, and it says less about how much of Metal's work happens in Pakistan, which the company has not stated.

Fundraise Insider, in its report of 15 September, said a16z Speedrun admits well under 1 percent of applicants and called it uncommon for Speedrun and Y Combinator to lead a round together. Read plainly, the lead investors are a signal about the founder and the team, since neither has published what it saw in Metal's revenue.

A growth rate says little without its starting point. Six quarters ago Metal was a small company, and a young business can post high percentages on a small base. The statement that matters for the year is the one about revenue: multi million dollars of annual revenue is a figure that can be checked later, through customers who renew.

Where the business is exposed

In our reading, the main risk is the nature of the customer. A fundraising tool is needed in bursts. A founder who closes a round has less reason to keep paying until the next one, which for most startups is a year or more away. Subscription software with a built in end date has to be cheap enough to keep, or give founders a second job to do. Metal's pricing is unpublished, so it is not known whether it charges per user, per round or per month, and that choice decides how well revenue holds up between raises.

Metal's own answer is to grow beyond the single raise. The ProPakistani report on the seed round says Gul wants the software founders use to raise capital to become infrastructure they keep using across the wider process of capital formation. That direction would address the problem above, though it also means the product Metal sells today and the one it describes are not yet the same thing.

The second risk is the data. Records of past venture rounds are widely available, so what a buyer pays for is the ranking: which investors fit which company, and which of them will reply. Whether those rankings beat a founder's own research is a testable claim, and the answer depends on how current the data is. A wrong match costs the founder time that cannot be recovered. A tool used by many founders at once could in principle learn which investors answer and how quickly, which a spreadsheet cannot. If Metal collects that signal, its data improves with every customer, which is the usual route to a lasting advantage in software of this kind.

Automated outreach also has a cost on the receiving end. Investors already receive more cold messages than they can read, and a tool that makes sending easier raises the volume further. In our reading, the gain from sending more messages shrinks as investors learn to ignore templated ones, so Metal's value rests on sending fewer, better aimed requests rather than more. That is consistent with the company's own claim about avoiding wasted meetings, and it is the part of the pitch a founder can test in a single raise.

What would settle it

The facts that would settle the question are not public: Metal's exact revenue, its paying customers, its price list and its retention. The test is simple and can be run from outside. If the founders who signed up in 2026 are still paying in 2027, the company has found a way past the end date built into its product. Until then, the six quarters of growth are a claim about direction, and the seed round is the only independent evidence of interest.