Leadership1 publisher2 min readPublished Updated
Entrepreneur column tells most founders of small-business AI agents to skip venture capital
Most AI agent startups serving small businesses should skip venture capital, an Entrepreneur column argues, citing weak retention under $50 a month. Its own vertical example took $330 million from an outside investor, so going vertical may still mean answering to someone else's return target.
The Board Room · Leadership desk
What happened
- An Entrepreneur contributor column argues that most founders building AI agents for small and mid-sized businesses should not raise venture capital.
- It cites a 2025 study in which AI-native software priced under $50 a month kept a fraction of normal B2B SaaS retention, while products priced above $250 retained at normal rates.
- Jasper raised $125 million at a $1.5 billion valuation, then cut its revenue forecast within a year as ChatGPT and Google's own tools took over its copywriting task.
- JobNimbus, software built for roofing and contracting businesses, took a $330 million investment from Sumeru Equity Partners.
Compiled by The Board RoomSomething wrong?How this is made
Why it matters
- decision An agent priced under $50 a month would have to charge five times as much to reach the $250 line, and the column links that price to depth in one industry's workflow.
- exposure A small-business agent whose core task a general assistant can add for free faces churn that product quality does not prevent. Jasper's forecast cut is an example of that risk.
- constraint A venture round on a horizontal agent commits the company to a growth target it has to hit while larger platforms give the same feature away.
The column reduces the funding question to one test: "what happens to net revenue retention the day the core feature turns free" [5]. If retention goes negative, it argues, there is no defensible business, "only a temporary arbitrage on a capability gap that closes faster every quarter" [14]. On Jasper, the column writes: "The product didn't get worse. The workflow it automated became available for free one layer up." [15]
Its evidence is thinner than the conclusion it draws. Entrepreneur runs the piece as a contributor's opinion [10]. The column cites a 2025 retention study without naming it, and it gives the study's price lines of $50 and $250 a month but not the retention rates on either side of them [3]. Jasper's forecast cut is sourced to "one account of the reset" [4]. For private equity, the column relies on "deal data on private equity buyers in vertical software" [9].
The venture half of the argument is the precise one. Venture capital wants revenue to grow 10x in three years off a thin horizontal wedge, the column says, "because that's the only math that returns a venture fund" [7]. Growing 10x in three years means compounding at about 115% a year, because the cube root of 10 is about 2.15 [11].
The vertical half describes a much smaller company. In the column's version of success, the business has a few hundred customers [8]. At the study's $250 monthly line, one customer pays $3,000 a year, so 300 customers bring in $900,000 [13]. A vertical business "doesn't need to grow that fast to be a great outcome," the column says [16].
JobNimbus complicates the self-funding advice. The column holds it up as the vertical model, and it took its investment from Sumeru Equity Partners [6]. The column's deal data describes private equity buyers acquiring premium vertical software [9]. It says a vertical business funds its own growth "instead of someone else's return target" [8]. I think the evidence it cites points somewhere else. Vertical companies that keep their customers attract outside capital, and that capital has return targets too. On that data, a founder who skips the venture round this quarter is setting up a sale to a private equity buyer as the likely exit [9].
The column rests each side of its case on one company: Jasper for the horizontal risk, JobNimbus for the vertical reward [4][6]. Two companies and one study [3] are enough to set out the choice. How often each outcome actually happens is not known yet.
What to watch
- Publication of the 2025 retention study's actual rates would show how wide the gap below $50 a month is and whether the $250 line holds.
- A general assistant shipping free website, booking or customer-service agents for small businesses would put the column's retention test to work in the market.
- Retention or revenue disclosures from JobNimbus or Jasper would show whether the column's two examples hold up.