HN Debrief

Halvar's Guide to Entrepreneurship

  • Startups
  • Fundraising
  • Product
  • B2B
  • Leadership

Thomas Dullien posted a long, experience-heavy entrepreneurship guide aimed at technical founders who feel misfit inside normal company hierarchies and want a realistic playbook for building a software business. The piece covers founder motivation, picking markets, separating user and buyer needs, working with investors, and staying sane through the grind. People responded well because it reads like scar tissue rather than startup theater. The parts that landed hardest were the practical ones: get clear on who actually buys versus who actually uses, do not confuse one enthusiastic customer with a product, and do not let investor ambiguity or founder martyrdom wreck your decisions.

If you are building a company, push every fuzzy commitment into something explicit, pay founders enough to stay stable, and treat “design partner” work as consulting until several customers want the same thing. Product discovery also gets better fast when you anchor it in named users and buyers instead of abstract personas.

Discussion mood

Strongly positive. People saw the guide as candid, experience-backed, and more grounded than typical startup writing. The main friction was around a few specifics, especially coaching advice, persona language, and whether any guide can say much beyond “build something hard to copy that people want.”

Key insights

  1. 01

    Handshake beats investor ambiguity

    Pushing an investor to state a specific amount and valuation cap turns vague enthusiasm into a real signal. The value is not legal magic. It is removing the founder-hostile gray zone where an investor sounds interested while keeping full optionality. A follow-up email then freezes what was actually agreed.

    Use this in every high-stakes negotiation where delay hides non-commitment. If you do not have a written yes on concrete terms, operate as if you have a no.

      Attribution:
    • tptacek #1 #2
    • xyzzy_plugh #1
  2. 02

    Named customers are better than personas

    Abstract personas are just believable enough to let teams fool themselves. Product work gets sharper when you can point to an actual person at an actual company, describe their job and buying role, and talk to them repeatedly. That makes room for surprising experiments that a tidy persona model would filter out too early.

    Require every target segment to map to real people you can name and contact. If your team cannot do that, slow down roadmap decisions until it can.

      Attribution:
    • jph #1
    • josephmosby #1
    • tdullien #1
  3. 03

    Design partners become consulting by default

    Early customer intimacy is not the problem. Pretending it is already a product business is. One customer will always validate some custom build. The shift happens only when several customers want the same core thing, and you can refuse requests that bend the product into a bespoke service. Examples from email analytics and Basecamp's pricing discipline made the point concrete.

    Track how many customer requests generalize cleanly across accounts. If too much revenue depends on one client's special needs, treat that work as services and price or limit it accordingly.

      Attribution:
    • softwaredoug #1
    • tdullien #1
    • lubujackson #1
    • hylaride #1
    • satvikpendem #1
  4. 04

    Founder salary should remove survival anxiety

    The useful bar for founder pay is not sacrifice theater. It is enough cash to keep the household steady so the company gets your full focus. Several people treated this as standard modern practice once funding is raised, and a sign that an investor is unserious if they pressure founders into insecurity. One practical rule was capping founder pay at the top salary level you are willing to offer other hires.

    Set founder compensation deliberately before it becomes awkward. Make it high enough to prevent distraction, then tie increases to company stage and internal pay norms.

      Attribution:
    • tdullien #1 #2
    • softwaredoug #1
    • pixel_popping #1
    • tptacek #1
  5. 05

    Coaching works only with real operators

    Helpful support came from people with actual business experience or psychology training, not from generic “startup coach” branding. The point was less about the title and more about whether the person can offer grounded advice or emotional support without exploiting a funded founder's insecurity.

    Vet coaches the way you would vet an executive hire. Ask what they have actually done, how they work, and what problem they are solving for you.

      Attribution:
    • faeyanpiraat #1
    • tdullien #1

Against the grain

  1. 01

    Sometimes leaders cannot unpack every decision

    The complaint about bosses who want things “their way” ran into a credible objection from people who have spent years building judgment that does not compress into a meeting. In that framing, insisting on a full rationale for every choice can destroy momentum. Some decisions really do need trust in a single decider, with deeper explanation happening off cycle if at all.

    Do not mistake every top-down call for ego or bureaucracy. In fast-moving technical work, decide in advance which areas are open for debate and which belong to an owner with final say.

      Attribution:
    • feoren #1
    • avmich #1
  2. 02

    Startup guides can overstate controllable process

    A skeptical view held that if you have something hard to replicate that people want, most tactical advice is secondary, and if you do not, no playbook saves you. That pushes back on the idea that entrepreneurship can be systematized into a reliable checklist rather than heavily shaped by rare product advantage and luck.

    Use tactical startup advice as support, not as proof that execution mechanics can substitute for a strong product edge. Recheck whether your company has a real moat before polishing process.

      Attribution:
    • alexashka #1 #2
    • orphereus #1

In plain english

secondary ↩
A sale of existing shares by founders or employees, allowing them to take some money out without the company itself raising new capital.
valuation cap ↩
The maximum company valuation used to determine the price an early investor gets in a future financing, often in convertible instruments like SAFEs.

Reference links

Fundraising norms

  • Y Combinator Handshake Protocol
    Referenced as the clearest version of the advice to turn investor interest into an explicit commitment on amount and valuation.

Customer discipline and product strategy