# The real cost of building in-house *The engineering bill is the smallest line item — what founders actually lose when they try to hire their way to a product.* **Authors:** [Vadim Zolotokrylin](/c/people/vadim-zolotokrylin) --- Most founders do the same calculation when deciding whether to build in-house or bring in an external team. They add up salaries, equity, and recruiting fees on one side. They look at a partner's rate on the other. The one with the lower number wins. The problem is that the calculation leaves out the most expensive variable: founder time. Time spent hiring is time not spent on customers. Time spent in architecture reviews is time not spent closing partnerships. Every hour a founder puts into managing a team is an hour not spent on the decisions that only a founder can make. That cost doesn't show up on any invoice. But it compounds. ## The numbers most founders don't run Assembling a senior team that can ship end to end — engineering, design, the judgment to know when to cut scope — typically takes a year. The first three months go to sourcing and hiring. Two more to onboarding. Then someone leaves, or the team never quite gels around the product vision, or the pace of good work turns out to be slower than expected. By the time the team is genuinely functional, the market has often moved. A competitor shipped. A regulatory window closed. An investor who was warm went cold. The in-house model also carries a hidden first-time-mistake premium. Every senior external team has already paid for most of the expensive errors — the architectural decisions that look obvious in hindsight, the scope choices that save six months, the early warning signs of a product drifting from its users. A team assembled from scratch will make those mistakes fresh, on your timeline, at your cost. ## What co-ownership looks like in practice We use the word co-owner deliberately. It is not a pitch — it describes a specific operating model. When we work with a founder, the same senior people who were in the pitch are the people building the product. There is no PM layer in the middle. There are no surprise handoffs to a team the founder has never met. A typical engagement is two to four senior engineers and designers plus a co-founder or partner from Holdex — a group small enough that everyone holds the full context and senior enough that nobody needs to be managed through basic decisions. The distinction matters because it changes who absorbs the coordination cost. In a typical outsourcing arrangement, that coordination falls back on the founder — managing an external team the same way they'd manage an internal one, without the institutional context that makes management efficient. In a co-ownership model, the partner absorbs that cost. The founder's job is to make the decisions that require their judgment. Everything else moves without them needing to be in the room. We run every engagement this way — same tools, same visibility, same source of truth the engineering team uses. The operating model is covered in more depth in [GitHub as the source of truth for client engagements](./github-customer-ops.md). ## The actual comparison When founders run the real calculation — including their own time — the math changes. Six months of in-house hiring and onboarding, plus six months of slowed execution while the team ramps, adds up to a year of delayed shipping. Against that, we are not the cheapest option. We are the fastest, with the fewest expensive first-time mistakes already paid for by someone else. The founders who get the most out of working with Holdex are the ones who've already done this math. They've tried to hire their way to a product, seen how much of their own time it consumed, and made a different call. They're not looking for the lowest rate. They're looking for a partner who can move as fast as they think, without needing to be explained to. The question is never really "how much does this cost?" It's "how much does the alternative actually cost when you count everything?"