Leadership1 publisher3 min readPublished
Rechtman's 30-100x market claim restates software's 1-3% revenue capture
Yoni Rechtman of Slow Ventures says AI lets vendors sell the outcome instead of the workflow. His 30-100x market expansion is software's 1-3% revenue share inverted, and the one priced case in the framework moves 10x.
The Board Room · Leadership desk
What happened
- Yoni Rechtman, a partner at Slow Ventures, previewed a five-model taxonomy of post-SaaS software businesses, with forward-deployed engineering first and neo firms second.
- His starting point is that software has historically captured 1 to 3 percent of a business's revenue while services captured the rest.
- He argues AI lets software deliver the outcome itself, and puts the resulting market expansion at 30 to 100 times.
- The forward-deployed model prices annual contracts at $500,000 where the product-only sale was $50,000.
- The post says high gross margin, passive retention and R&D amortized over long product cycles break down along basically every line of the income statement.
Compiled by The Board RoomSomething wrong?How this is made
Why it matters
- decision Owning the implementation means hiring experts before the revenue arrives, and a delivery organisation is much harder to unwind next year than a price list is.
- exposure A vendor that guarantees the result carries the performance risk the customer used to carry, and absorbs the cost when the work does not land.
- constraint Growth then depends on a transformation budget approved by a different set of people than a software renewal, so the vendor's pipeline moves with a discretionary line.
The two headline numbers in this framework are one number. Software historically took 1 to 3 percent of a customer's revenue and services took the rest [2], because software could only orchestrate the work and not do it [3]. Capture the whole budget instead and the multiple is 100 divided by 3 at one end and 100 divided by 1 at the other, roughly 33x to 100x [17]. Rechtman's 30-100x market expansion is that arithmetic [6]. It is a ceiling, and it holds only for a vendor taking every dollar the customer spends on the function.
Rechtman, whose fund manages more than $1B [1], prices the one worked example, the forward-deployed contract, at $500,000 against $50,000 for the product sale, which is 10x [18]. That captures between a tenth and a third of the ceiling [19]. What the buyer is paying for, he said, is "the thing doing what I said it's going to do" [13].
Forward-deployed firms carry heavier delivery costs and harder scaling curves, according to the post [12]. The post does not disclose gross margin or a delivery headcount for one of them [20].
"No one has ever woken up and said, I want to buy workflow software," Rechtman said [4]. The higher price gets signed on the transformation line item instead of the software line item [10]. "Rather than saying you change your business process around my software," he said, "I'm going to change my software around your business process, and I'm going to guarantee that it's going to do that" [9]. Enterprises want to buy AI transformation now, per the post [11]. A vendor who arrives with experts and owns the implementation joins the buyer's future-proofing story; one who arrives with a web app competes on features with every other web app [11].
Software takes 1 to 3 percent because a vendor sells into one function, and much of the remaining 99 percent is payroll, rent and materials that no vendor can invoice for. Rechtman's own examples of selling outcomes at software margins are Google, Visa, Uber and Facebook, which he describes as network effects businesses [5]. Forward-deployed engineering does not have a network effect. He states the risk in the post: whether a vendor can graduate from the wedge to something with software-like defensibility before the arbitrage window closes [14].
For a vendor deciding this quarter, the choice is narrower than the taxonomy. Price an outcome and hire the people who deliver it, or hold the product at $50,000 and compete on features [12]. Slow's own bets are structural ones: it seeded Metropolis, which acquired its way into the parking industry and is now a $5B physical AI company, and led a $4M seed round for Ando in restaurant staffing AI [16].
What to watch
- The full deck: whether the remaining three models arrive with pricing and margin figures, or only labels.
- Any forward-deployed vendor disclosing gross margin at $500,000 contract values would settle what the model costs to run.
- Whether enterprise AI transformation budgets survive the next planning cycle or fold back into the software line.