<p>When should you start with process and architecture in a startup?<br>That was the question that we've asked ourselves (as if there is a real-life example currently happening :-) and then we thought, “Why not ask someone who is living in this space?”</p><p>Vidar co-founded his first tech startup at 19 because he didn't know enough to know how hard it would be. 30 years on, he has gone mostly from startup to startup, usually as the first technical hire or a co-founder. He has both bootstrapped and raised VC capital and recently spent 3 years working at a VC fund. He now runs a tech consultancy focusing on the intersection of DevOps and AI while working on his next big thing.</p><p><br></p><p>In this episode of the <a href="https://www.whatsyourbaseline.com/podcast/">podcast,</a> we talk about:</p><ul><li>A startup is defined by pre-revenue or pre-profit status combined with rapid growth ambition — a chip shop is operational from day one, not a startup. Once you're profitable and growing modestly, you're a lifestyle business.</li><li>Early-stage capital is the most expensive capital you'll ever spend, because you're paying in equity. The earlier you are, the larger the slice of the company you trade away for the same dollar amount.</li><li>VC investors expect most of their portfolio to fail and are only looking for the 10x–100x outlier. If you can't raise your next round within about 18 months, their interest moves on — so failing fast and validating quickly is the entire game.</li><li>AI has dramatically lowered the barrier to building a working prototype, letting founders show investors something tangible and compelling much faster than ever before.</li><li>But the industry is swinging back hard toward upfront specs and documentation, because AI coding agents can't infer your unique business context. Writing a truly good spec turns out to be one of the hardest parts of the entire software development process — and most teams have been skimping on it for 25 years.</li><li>Your truly proprietary assets are your ideas, processes, use cases, and customer segments. The generated code is a commodity — everyone building on the same AI tools has access to the same output.</li><li>Lightweight processes pay off quickly once a startup begins to scale. The absence of basic QA, project management, and clearly written tickets is almost always what causes delivery to break down first — not the process itself.</li><li>A common and costly mistake is over-engineering: developers building for 10 million users when the total addressable market is 10,000. This happens when engineering teams are never told what the product actually is or who it's actually for.</li><li>Process and architecture function as a communication layer — aligning engineering, sales, and leadership around a shared vision, the customer they're serving, and the strategy connecting the two. It builds buy-in, and buy-in produces better work than a paycheck alone.</li><li>Vision must be communicated continuously from the hiring process onward, not just stated once and assumed to stick. Developers detach from the "why" quickly when daily work becomes purely tactical.</li><li>Both one-on-one check-ins and group meetings are essential to a healthy team. People rarely surface real blockers, interpersonal tensions, or technical concerns in group settings — individual conversations build the trust that makes those things visible before they become crises.</li></ul><p><br></p><p>Vidar can be reached on LinkedIn <a href="https://www.linkedin.com/in/vhokstad/">here</a> and also has a website: <a href="https://hockstadconsultng.com/" target="_blank" rel="noreferrer noopener">hockstadconsultng.com</a>.</p><p><br></p><p>Reach out by emailing <a href="mailto:
[email protected]">
[email protected]</a> or subscribe to our newsletter and articles on Substack at <a href="whatsyourbaseline.substack.com" target="_blank" rel="noopener noreferer">whatsyourbaseline.substack.com</a>.</p>