Hacker News new | past | comments | ask | show | jobs | submit login
Reid Hoffman interviews Michael Seibel [pdf] (mastersofscale.com)
48 points by breck 28 days ago | hide | past | favorite | 7 comments



Found the following quote interesting:

"The number one killer of startups is company building. It's the number one killer of startups. And people are 10x more afraid of scaling too late than scaling too early, which always blows my mind. Scaling too late isn't fatal. It could be bad, but it's not fatal. Scaling too early is almost always fatal."


The way I see it this is part of the execution, although often overlooked in favour of product-market fit.

Trying to scale an organisation by hiring when the organization is not fit for that scale is like trying to deploy a product at scale when it is not ready. Very easily you get into exponential costs due to the amount of handholding, spiraling firefighting and potentially immanent implosion in staffing caused by growing dissent.

There may be times when it is unavoidable and maybe the right choice. But it is not sustainable if not fixed and much better avoided if possible.


Well said. Grow headcount as slowly as possible (but no slower) and fire mistakes quickly is a great rule of thumb. At the same time, put loads of effort into hiring.

A lot of this comes down to resisting anything that's tempting to the ego: an expensive office in a nice building, a large headcount under you, talking to investors or journalists/receiving awards/attending seminars ... instead of simply building and trying to find the top talent that'll make the difference.


There is one thing where engineering and organisation building differ though, even at the high-level: available sources and trusted guides, and iteration speed.

For something that is such a common problem, that basically every company has to solve, there is astonishingly little factual guidance out there on how to scale an organisation and what scale requires what type of infrastructure. And all entrepreneurs have to reinvent the wheel to a large extend.

There is this saying that farming is one of the hardest jobs, because you only have 30-40 tries to get the perfect harvest. But being an entrepreneur cuts it to 6-8 tries to get the organization right.


My take on this is, recognize that there is a cycle to your business process, identify what they are, focus on just one at a time, scaling it as you go, until the others become the bottleneck, then shift your focus to the next one. Rinse, repeat. 9 out of 10 times, premature scaling or trying to scale everything simultaneously is a recipe for disaster due to wasted/misguided allocation of resources.

For a SaaS company, that cycle could be building -> selling -> support -> building. To kick off that cycle, you have some idea of a problem to solve, and you start building (to an MVP). This doesn't mean you don't start selling/support until you're done building, of course. It simply means you devote more resources to building. When selling becomes a bottleneck, you stop scaling product building, and start growing your sales team. As more clients get on-boarded onto your product, support becomes the next bottleneck. You start to see common feature requests and pain points. Now building becomes a bottleneck. Hopefully you sell/raise enough to sustain the next round of growth, and this time, you focus on building.


Thank heavens for this transcript.

I tried listening to this episode earlier today on my podcast app and I just couldn't sit through because of all the narration and dramatic music. Why can't they stick to just two people talking? It sounds more like a podcast for kindergarten and less like one for startups.


Personally not a huge fan of the question “do you see yourself doing this 10 years from now”. The only thing stopping your imagination is answering this question with a binary no.




Applications are open for YC Winter 2022

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: