VIDA Went Public at $330,000 of Quarterly Revenue to Sell AI Employees
The labor-budget pitch is genuinely differentiated. The 2% sequential growth, the withheld ARR line, and the empty CFO chair are the parts to argue with.
VIDA · Earnings Call · 2026-11-16
A startup's pitch delivered in a public company's wrapper
VIDA's debut call is an odd artifact: a company that cleared roughly $330,000 of revenue in the quarter, explaining — patiently, from first principles, with a Humane Society wildfire anecdote — why investors should fund its attempt to grab a slice of the American payroll. “Thank you for joining us for VIDA's first earnings call as a public company. Because this is our first call, I want to take some time to introduce VIDA and the opportunity we're building toward” — Lyle Pratt, Founder and Chief Executive Officer · 2026-11-16 is how Lyle Pratt opened. There are no prior quarters to hold him to, which is precisely why the framing choices matter so much.
The numbers, laid out by co-founder and principal financial officer Brandon Robinson: revenue of $330,000, up 88% year over year from $180,000 — and up just 2% sequentially. Gross margin of 44%, versus 35% a year ago but below the 49% posted in Q2, dinged by infrastructure stood up for the new computer agents. Total costs and expenses of $2.1 million against $0.9 million a year ago. Operating loss $1.77 million, net loss $1.47 million.
Revenue for the third quarter was $330,000, up 88% from $180,000 a year ago and up 2% from the prior quarter. Our partners are still onboarding and building the repeatable processes they need to sell and deploy VIDA across their own customer bases, and we are building the business for that ramp.
The year-over-year comparison is doing enormous rhetorical work here. A business with a $1.3 million annualized run rate that grew 2% quarter over quarter while spending $613,000 on sales and marketing is not yet a growth story you can underwrite; it's an early-stage company that happens to have a ticker. Management is unusually candid about that, at least indirectly.
Raiding the labor budget instead of the IT budget
This is the part of the call that is genuinely company-unique and worth taking seriously. VIDA's product is an AI employee — an agent with a defined job, guardrails, an audit log, and a handoff to a human — and its commercial argument is that software has historically competed for the IT budget, while agents can compete for the payroll. “In the U.S. alone, office and administrative work is estimated to cost about $915 billion a year in wages based on the Bureau of Labor Statistics data. That is the budget VIDA aims to tap into.” — Lyle Pratt, Founder and Chief Executive Officer · 2026-11-16
That framing produces a real product decision: outcome based billing, launched publicly in October, where customers pay for a verified result such as a qualified lead or a booked appointment on supported jobs rather than per seat. It's a wedge that most agent vendors have not committed to, and it's the sort of thing that only works if the reliability claim holds — which is why management keeps returning to the harness, the deterministic process graph that turns a job into repeatable steps and lets the company substitute smaller models or plain code where SOTA intelligence isn't needed.
The market wave is real but crowded. Personal AI agents show up in the global keyword tape this quarter, and "AI agent" appears in a striking number of unrelated reporters' calls — from a Japanese internet company to a space tourism outfit. VIDA is riding that wave; its differentiation has to come from the labor-budget framing and the billing model, not from the noun.
Tokens are a commodity — and that is the margin story
Pressed by an investor on inference costs, Pratt gave the sharpest answer of the call, and it doubled as a margin thesis. “Ultimately, we think that the business model of charging for tokens is a bad business model. Ultimately, tokens are a commodity.” — Lyle Pratt, Founder and Chief Executive Officer · 2026-11-16 He went on to describe routing: SOTA models for one step of a workflow, open weight models — possibly self-hosted — for others, and fully deterministic code where neither is needed. That is squarely in line with a broader market conversation: open weight models surfaced as a top-25 global theme last quarter, and the cost of inference is now the central swing factor in every agent-economics debate.
Still, 44% gross margin on a services-like delivery model is thin for software, and it fell from 49%. The company's answer is scale: fixed infrastructure costs spread over a larger base. With $330,000 of revenue, everything currently spreads over nothing.
Distribution as the moat, and the metrics withheld
The channel architecture is more developed than the revenue line suggests. ALLO Fiber resells VIDA under its own brand to roughly 15,000 businesses across 50-plus communities in six states; AVANT, a technology distributor with more than 3,000 advisers, signed on as a co sell partner in September. Management claims its signed partners reach more than 70,000 businesses and adviser opportunities a year. Pratt went into the unit economics unprompted-ish: co-sell deals land on VIDA's paper and skew larger, with a fee to the partner; resell partners pay roughly retail less volume discounts, which is economically similar once you account for avoided sales and marketing spend. The new MCP server lets a business owner describe a job in plain English inside an assistant they already use — a genuinely low-friction entry point, and one that turns every agentic platform into a possible front door.
The disclosed operating metrics are the interesting tell: nearly 1,400 AI employees, more than a million billed voice minutes (+161%), about 60,000 messages (+73%), roughly 40% of billings coming from usage overages, and revenue per agent of about $79 across the installed base versus about $167 for agents created in Q3. If that doubling holds, it's the single most important number in the deck — evidence that the shift to computer agents is monetizing richer work. But it is an estimate from recent monthly billings, not recognized revenue.
Which brings up what isn't there. Asked directly about ARR and MRR, Pratt deferred: “We want to see a few quarters with this new sales team and with these new partners before we start including that in our earnings materials.” — Lyle Pratt, Founder and Chief Executive Officer · 2026-11-16 Reasonable, and also convenient. There is no CFO in seat — the search is "underway" with a firm retained and no timeline. And on spending, the guidance is effectively flat: “I wouldn't expect us to grow OpEx substantially until we grow revenues.” — Lyle Pratt, Founder and Chief Executive Officer · 2026-11-16 That is the right discipline for a company that burned from $14.5 million to $12.7 million in cash and short-term investments in one quarter — about seven quarters of runway, or a bit more if you count the $977,000 of Bitcoin that rounds the holdings to $13.6 million.
The part that feels like 2026 in a capsule
One detail deserves more attention than it got: VIDA listed early on purpose, arguing the median company now waits twice as long as it did in 1980, and it issued equity in tokenized form — VIDAx, via xStocks on Kraken, for eligible non-U.S. investors. A sub-$2 million-revenue company with Bitcoin on the balance sheet and a tokenized share class is precisely the kind of creature this market is producing, and it sits right on top of a live global theme in tokenized stocks.
What changed at VIDA this quarter is that the pitch now has public arithmetic attached to it. The thesis — reliability plus distribution as the moat, agents billed against the labor budget — is coherent and unusual. The financials say the company is one good channel ramp away from mattering and one slow quarter away from needing to raise again. Management asked investors to judge it on deployments, work performed, and revenue per agent. Three metrics, one quarter of history. The next two prints, and whether a CFO arrives to explain them, are the whole story.