What VCs Really Want: A Dialogue on Service Scalability

Transcript

SPEAKER_1: We landed on reusable software modules as the bridge between a services firm and a more scalable model. I want to get into the numbers—what does a VC actually look at? SPEAKER_2: Gross margin is the opening filter. A typical agency runs thirty to forty percent. A scalable AI firm should be pushing toward sixty or seventy, because that gap is where compounding happens. It tells investors immediately whether the business looks like software or staffing. SPEAKER_1: So gross margin is necessary—but a consultancy could have strong margins and still be completely stuck, right? SPEAKER_2: Exactly. The key idea is whether revenue can grow faster than costs. If every new client requires a proportional headcount increase, the margin stays flat and the business struggles to escape the staffing trap. That is what headcount-linked revenue means—it is a structural ceiling, not a temporary problem. SPEAKER_1: Think of a firm landing five new enterprise clients—each one needing a dedicated engineer for six months. That is not scaling, that is just getting busier. SPEAKER_2: That is the exact scenario investors fear. A later client should cost meaningfully less to serve than an earlier one. When it does not, capital efficiency collapses. Investors are not buying growth—they are buying a machine that gets cheaper to run over time. SPEAKER_1: So how does a firm actually prove that the machine is getting cheaper? What is the operational evidence? SPEAKER_2: Standardized workflows and reusable delivery templates are the clearest signal. For example, if implementation time dropped from twelve weeks to four across successive clients using the same internal tooling, that is measurable. Operational metrics like project throughput and utilization rates tell that story concretely. SPEAKER_1: That connects to the Service-as-a-Product framework. What does that actually mean in practice? SPEAKER_2: It means packaging delivery into fixed-scope offers rather than bespoke engagements. Instead of scoping every project from scratch, the firm sells a defined outcome—say, a compliance automation module—with a known timeline and known deliverables. That reduces sales friction and makes performance comparable across customers. SPEAKER_1: And repeatability is what VCs are actually underwriting—not the technology itself. SPEAKER_2: Repeatability is the whole game. Investors want evidence the same offer can be sold and delivered across many customers with limited customization. A repeatable niche—automating compliance workflows for regional banks, for instance—is a wedge they can underwrite. A vague promise to serve any enterprise is not a business model. SPEAKER_1: What about revenue structure—does project billing versus recurring revenue actually move the needle in diligence? SPEAKER_2: Significantly. Recurring revenue improves predictability, and predictability reduces investor risk. But the signal they really want alongside that is expansion—existing customers increasing spend over time. Retention plus expansion is the compounding engine that makes a services business start to look like a platform. SPEAKER_1: Customer concentration seems like another fast way to stall a VC conversation. SPEAKER_2: One of the fastest. If two clients represent sixty percent of revenue, one lost contract could crater the model. Investors want a distributed customer base. Now, that is also why the go-to-market motion matters as much as the technology—because efficient customer acquisition is what builds that distribution. SPEAKER_1: There is also the founder dependency problem. The more the business runs on one person's expertise, the harder it is to pitch as a platform. SPEAKER_2: That is key-person risk, and it is a valuation killer. The more a company's value depends on the founder's personal judgment or relationships, the less it resembles a scalable asset. Documenting processes, building internal tools, distributing decision-making—those are all visible ways to reduce that dependency before a diligence conversation. SPEAKER_1: So the takeaway for everyone building in this space—it is not just about impressive technology. It is about proving the machine can run without you. SPEAKER_2: That is it. Not every AI automation consultancy needs to become a pure software company. But every one needs a believable scaling mechanism. Show the gross margin trajectory, show the repeatable niche, show that delivery gets cheaper with each client—and remember, the path from services to productized offerings is what converts a strong consultancy into a VC-grade investment.