Hacker News Reader

Comment by ThalesX | Hacker News Reader

I've started my career in enterprise environments, and recently on my 3rd startup. As the author mentions, startups to seem to be broken, going after everything and finding product market fit and others...

While this is true, I can't help but shake the feeling that adding corporate-planning to a startup-environment would actually calm down things and would still hit targets. I do get a bit frustrated when startup leadership don't think about brokenness because 'that's the way startups are'. I have a friend, big corporate leader; I suspect he'd restructure any startup into something more performant in terms of delivery in less than a quarter.

View on HN

It's built into the startup concept that the people who make them are mostly going to be driven, action-oriented "doers" who stick things into the gears of organizational machinery at random to try to make them turn.

A smarter founder will know to check assumptions, but it's a case of putting armor where the bullet holes are: sheer persistence is way more important, and where new companies are most likely to fail is in giving up on one or more types of roadblocks and trying to choose only certain fights when they need to persist and come up with some immediate thing that addresses every challenge and change, even if it's not a real or permanent solution or the specific thing they want to be focused on.

The other way to go is, well, the slow way: study the problem, build consensus, look before you leap. While older founders who are bootstrapping will tend to act more conservatively like this, it's not favored by startups of the VC money sort, not because it's actually a worse way to go, but because it isn't playing the game - it doesn't result in huge investment opportunities where a giant market share is captured in a short time frame. If you go in aiming to build consensus you may actually benefit society, but you aren't going to "capture" things. After all, capturing means "I win, you lose".

Thus, a startup that does get to the point where they are rapidly soaking up market share is likely to have a dysfunctional quality of being dissatisfied with the thing the business is currently about: there always has to be a next thing to conquer. Thus another re-org, new hires, repeat until you've attained tech giant status.

With my own business I’ve experienced just about everything you’ve mentioned here. We are explicitly not seeking hyper growth, and we are fine with the “traditional” approach that says “businesses take {three,five,seven} years to become profitable.” From a financial perspective my cofounders and I aren’t doing this to get the 45,000 square foot mansion in the hills overlooking Veil. We just want to use our skills to do something that helps other businesses as a means to building a tidy retirement nest egg. Several family members have done that same thing in each of our families in the past.

And yet we’ve still fallen into some of the same traps any startup in technology falls into. We recently decided to exit the Atlassian ecosystem (prior to their recent issues) and have been doing so while we’re still under the grandfathered pricing for their cloud offerings. We didn’t want to wait until the last minute. We bought ourselves some time so we didn’t want to rush it either.

Just deciding to step back and replace tooling can be difficult. It means you’re tying up a LOT of cycles in internal work, but sometimes that’s a good call to make. We had been using Jira for anything “taskable” but we have both internal/code work and external/billable work, and we weren’t managing either very well in Jira. It’s easy to say “we will lose a lot of history and velocity” but you can export data, and the velocity you lose continuously by not using tooling that works for you, can be much more harmful.

It can be difficult to SEE that you need to step back and reevaluate something as big as your tasking or documentation management. For us, we decided to use this as an opportunity to look with fresh eyes at everything and ended up replacing virtually all of our tools and infrastructure. We even replaced all but one core AWS service in our application and moved everything else to Linode. We replaced our billing system too. We moved from Chargebee, which was vastly over complicated for our current needs to the point where we hated using it, to using pure Stripe and project management with ActiveCollab.

We said no to some business while we did that, but from the inside perspective our lives are much easier, our revenue cycle is more consistent and manageable (and actually -happens-) now that our billables are in a revenue-aware tasking system (ActiveCollab) instead of a pure-task tracking system (Jira).

We’re OK with taking our time because slowly we’ve been seeing just how to build this business effectively, and for us that’s far more important than scale or quarterly returns right now.

Deutsche Telekom has this "innovation arm" called T-Labs, where they spawn startups that could have the best of both worlds: they have someone from senior leadership on board mostly to help with the operational aspect, the executive side is free to do pretty much anything they need/want.

I joined one of their companies because I thought exactly like you, that without the chaos the startup could probably be more efficient.

After 3 years in 3 different projects there, I had changed my mind. Without embracing the chaotic nature of a startup, leadership gets too attached to their original plans and ends up taking longer to find product fit. Hiring was fast because they would bring consultants for anything at first, but because their pay was not tied to company growth, they only cared about milking DT cash as long as possible, and so on.

Some of the people that worked there went on to create their own startups, but I am yet to see any big breakout story from T-Labs.

They're simply not going hungry if it fails which probably correlates with the eventual failure. This is one reason why I believe that too much capital too soon is bad for start-up founders. They lose a part of their drive and on top of that they are now much more diluted than before. effectively the failure will be theirs and the success will be someone else's who didn't do much other than to write a check.

Could you share what parts of corporate planning would work well in a start-up environment?

I find year long plans in a start-up environment difficult mainly as the plan needs to adapt more than in a corporate environment.

Definitely introducing some sane planning and prioritization helps. A lot of startups fail because they end up reacting to events around them and are constantly fire fighting. That usually means a lot of work gets done, people are very stressed, the goal posts keep on moving, and things don't actually progress a lot and are possibly getting worse.

Ultimately, the amount of work you do does not matter if it doesn't get you closer to where you need to be. So, the most valuable thing you can do in a startup is figuring out the big picture. You can't fix all the issues. You are starved for resources and time. So you need to be super picky about which things you actually do and when. In fact, mostly this is about deciding which things you definitely aren't going to waste brain cycles on. Perfect is the enemy of good here. It's a given that you are going to have to take short cuts on some topics.

In practical terms:

- Identify what is really blocking the business. Start removing those obstacles. This requires that you understand both the business and the tech. That's the classical role for a CTO. A good CTO needs to be all over product management, business development, and sales so he/she understands what is being sold and how the product aligns with that. The tech is a means to an end and needs to serve business goals. Why are sales deals falling through? What's missing in the product that makes it not fit the market? Etc. Ask those questions and figure it out. Understand why this is not working out.

- Then write down the ideal end state of your company some years down the line. Basically, define the strategic goals and translate that into short term tactics to get to those goals. Shoot for something that is actually doable and realistic. At least on paper.

- Break it down into things you do not yet have in the product that would work towards that vision. Think in terms of themes that you can strive to improve rather than a long feature list. This isn't a backlog. It's gaps in your product where your lack of progress is holding back the business in some way.

- Prioritize these themes by quarter and year and argue why these are the most important things to tackle next. Basically, you plan a few quarters ahead and you front load the most critical things. This is your roadmap. Realistically, cap the number of themes per quarter to just a handful. "I want it all, I want it now!" is not a plan but a great line from a song by Queen.

- Organize your backlog of features, nice to haves, tech debt, enhancements, etc. by value and difficulty and align with your roadmap themes. Only do those things that are aligned, valuable, and doable and in order of priority. You will never get to the bottom of that backlog. But if keep chipping away at the most valuable things first, you should be making the company better over time. This is what a good product manager should do together with the CTO. In some companies this is one person: a technical product manager.

- Revisit the plan regularly and divide work you are doing day to day into "unplanned things that just have to happen" and "things that make progress on our roadmap". About a 30/70 split of time is ballpark a realistic goal. There's always shit to deal with but you want to keep on making progress on your real goals most of the time. Anything else, you don't waste time and energy on. The benchmark of success is clear visible progress on the themes that were identified as the most critical. If you have extra time, start working on the next most important thing. No reason to sit on your hands.

Get consensus from anybody that matters in the company that this is what the company does. You'd be surprised how many startups don't even have that consensus and are just operating day to day on the whims of some erratic C-level types with short attention spans. Basically, insist on some realism. You're not looking for wishful thinking here but something that is doable and valuable. Beware of CEOs that suffer from magical thinking here that just gloss over all the obvious issues that will ensure the goals are perpetually unachievable. I've seen this in action a few times where they just assume a knight in shiny armour will come along and wave their machine learning sword around a bit and will turn them into the next Steve Jobs. Never happens without a real plan.

Stop reacting, start planning and strategizing. Figure out the hard issues. This is how I've been running my current startup. It's hard. There are so many things to react to and almost all of them are horrible time sinks.

Reply on news.ycombinator.com