What I don't understand is why every single startup has to be around some web 2.0 software idea. There are thousands of industries outside of software for MBA students to pursue. If you're a good businessman, you shouldn't pigeonhole yourself into a position where you have no expertise, that's just bad judgment.
I think in general the distinction is that a startup can scale more readily than a small business. That's easier to do if you're selling software than, say, haircuts which is why I think the term "startup" is used so much around here.
I think there's also an element of the entrepreneurial mindset on HN. The term "small business" seems to be considered implicitly defeatist here, unless you're (merely) building a "lifestyle business" which excuses your lack of ambition!
I've found that the best definition for "startup" that I've come up with is that a startup is a business that seeks a future liquidity event. A web company that is one guy's lifestyle business selling tv stands isn't a startup any more than a small family owned restaurant. But a restaurant chain like Chipotle is a startup -- at least, until they were bought by McDonalds.
Thus there are startups in and outside the tech world.
Not every MBA is doing this. A number of my MBA classmates are getting into things like bio-tech, micro-sensors and other kinds of industries, where their background matches research developments in the University.
But if you have an interest and passion for a particular software idea (that meets a strong customer need), then I say go for it.
That's what I'm doing. Though I'll admit to having a decent bit of programming experience a number of years back, it's been a while and I'm rusty enough that I'd prefer to partner with a real coder who will be much more productive while I concentrate on the business aspects.
Agreed. Its a lot easier to make another Google where it reaches billions of people compared to a car manufacturer who only can reach to places he can ship.
This is terrible advice, even for people who want to do web startups. The job of the entrepreneur is to pull everyone from their social network to make the thing work. They're responsible for finding the web guys, the sales guys, the marketing team, the science advisers, the customer support people, etc.
Learning to code for the CEO only makes sense if your ambitions are limited to creating a two person startup in a garage. Don't get me wrong, plenty of great products have been created that way. But most of the best businesses are just slightly different versions of things that have already been done, and by focusing on learning to code instead of building your social network you are eliminating 95% of your business options before you even start.
I don't think the advice is "learn to code", as such; rather, the author's saying "be prepared to do whatever your business needs to be successful." Obviously the CEO shouldn't be writing a lot of code in the long run, but at a startup, everyone needs to be prepared to do whatever needs to be done, not necessarily what their job description defines. Hiring "web guys, a marketing team, science advisors, customer support people", etc. makes sense in the long run, but a CEO who only views his role as managing others + "the vision thing" and isn't prepared to get his hands dirty, probably isn't a great early-stage startup CEO.
Spot-on. The whole premise of learning to code as a "preparation" to launch a startup is just idiotic.
It takes years to become a half-decent programmer, so if you have this hot internet-idea right now but no idea about programming then better go find someone who can implement it for you instead of trying to learn it yourself...
The problem with this is that non-technical people have no idea who can implement their brilliant idea for them. They don't know how to tell good programmers from bad ones, so they'll just hire somebody with an impressive resume, with varying results.
I'm not saying that a 28 year old MBA who has never written a line of code should spend 2 years becoming a semi-competent hacker instead of hiring somebody, but they probably would have been better off spending some time on it in high school instead of pursuing straight As or doing application-polishing ECs.
The problem with this is that non-technical people have no idea who can implement their brilliant idea for them
That's why you need to build your network.
I'm an electrical engineer: I have no idea how to identify a brilliant biochemist. But if I need to hire one for my new alternative energy startup, hopefully I'll already know someone who has expertise in that field and can either recommend one to me, or help in the hiring process.
Yes but the GPs point still stands. You're an EE, after you've found the brilliant biochemist you have the knowlege framework to relate to her. A non-technical MBA may very well be unable to relate and without someone communicating properly to him, he'll be lost.
See the movie startup.com for an example. A Harvard MBA hires his friend, who brags that he's been programming since he was 6 IIRC, as his CTO. But this CTO doesn't really appear to know anything, and the company ends up burning through $60 million of other people's money without making a usable product.
I'd like someone who can do a lot of things, such as properly form the corporation, manage the books, do some branding, get us some PR and has an eye for great visual design. Oh, that and get us some funding. Really good funding. At least I need someone who has the network or drive to find those people to do it for us.
I'll write the SQL, do the SEO and find great coders to help us build the darn thing. I'll even manage the projects as needed. In my situation CEO that codes would just be another programmer. I really do need the business skills.
Oh, and one more thing. The catalog site he built, hiscatalog.com is neat looking, but just looks like another affiliate catalog site. I'm not sure (aside from some clean design) what it offer beyond that.
I agree. I am programming for my own startup but I actually think that works to my disadvantage at times. The reason being that it limits my perspective - I tend to look at problems and want to solve them by writing code. Someone who is unable to do so will have to solve them through delegation which is often much better for business.
if you don't mind, i'd like to throw in a few desirable traits of a business person.
i've rubbed shoulders with business students as a co-founder of a student entrepreneur group. the problem usually starts when the business guy runs out of money to pay people in india on elance for his project. so he poaches other people's teams for programmers, exaggerates the heck out of his project (we built this in 2 months! [read: this has been sitting inactive for 4+ years]), treats the programmers like s*, gives ridiculous deadlines, and yells a lot.
but that's not to say the business guy is useless at all. here are a few things they could help the coders with:
[1] term sheets, make sure they're good.
[2] go do the fundraising (be really good at presenting in front of a lot of people too)
[3] talk to as many of the users as possible and let me know what the top complaints and compliments are. the rest i'll figure out from analytics. [4] get us partnerships that will help everyone immensely, the user, the startup, and the partner.
Knowing code has made it a lot easier to hire people to work for my company, as they don't care that I don't know code as well as they do, they just care that I understand their world enough not to place unreasonable demands on them.
Totally agree. I'm a product manager, but I taught myself RoR and MySQL last year mostly because I wanted to better understand how databases work with web apps. In the process I also learned how to use CSS, SVN, nginx, and search engine optimization techniques.
I don't ever expect to use these skills in a major web app (I'm nowhere near good enough to do it for real), but they've helped me immeasurably in communicating effectively with engineers. Wish I had done it years earlier.
It was totally random. I was an English major / mass communication minor planning to be a reporter until I started playing around with HTML while editing my college newspaper (way back in 1995). I snagged a grunt job at Cooking Light magazine putting their content online, and just went from there. I don't know of anyone who set out to become a product manager, but it's a great career for certain people. Having an engineering, UI design, or even customer service background helps a lot. For consumer-facing web sites, just being a very heavy web user is an absolute requirement. Some MBA's are good in the role, but too often they see the role as a stepping stone to other opportunities. Hope this helps.
Thank you for your answer. I'm 25 and considering what I should next in my career. I've been reading about product management lately and it's seem to be a good fit to my personnality and aspirations.
It's interesting how "start-up" has somehow become synonymous with software or web start-up company, but I think it applies to all start-ups in the sense that competency in the skills needed to create the core business are primary when starting out.
Only on HN is startup synonymous with software startup. In other place startup can apply to many different kinds of firms: biotech, cleantech, nanotech, MEMS, semiconductor, medical device, to name a few technologies/markets that get much less attention on HN.
This site was Startup News when I registered. It was about software startups. Most every person who registered found this by googling for PG's name from lisp books, slashdot, and ycombinator. When the discussion became about web startup and business advice from blogs, pg increased the scope.
Nowadays, such articles about web startups are actually much rarer and do appeal to many users.
As a word, "startup" has it's origins in VC. In the last decades VCs focused primarily on software/IT. Because the industry was growing the most the fastest. Normal stuff. In a decade or so the conversation goes like "What do you do?" "Startup" "Oh, biomedical or..."
People have a tough time thinking in sets. This applies to developers as well. This is why people constantly fight the relational paradigm; you end up with things like CouchDB or OODBs because they make more sense to developers. It's easier to reason about an object than it is about a set of data.
I've been working on writing a bunch of "SQL from the ground up" postings, but I keep getting side-tracked with another project I'm working on and I haven't had a blog in years to post it on. Besides that, 99% of all the SQL work I've done has been in Sybase, MS-SQL and Oracle; so I'm not sure how relevant T-SQL & PL/SQL are to MySQL or Postgres. I'm assuming the basics are all the same though (which is to say ANSI compliant).
Why is it that every time a database gets mentioned it turns into PostgreSQL vs MySQL ?
I've worked with both, they both have their learning curve, whatever you already know is going to be 'easier' to you because you don't have to unlearn any habits, I wouldn't advice anybody to use either until I know more of the intended application.
Yes, but the other way around (Oracle's implementation of SQL lines up closer to Postgres). Also, Postgres has a procedural language based off Oracle's PL/SQL language called plpgsql, so if you have experience in one, learning the other is pretty simple.
I have seen terrible ceos who can code and great ceos who can't. The only common factor amongst the great ones is that they are smart and interested in learning.
When you are just starting out with a shoestring budget, developing the product is the most important thing on the to-do list. Business skills are certainly desirable. But since the company is operating on limited budget, those skills cannot be procured at the expense of actual product development. However, when a company has grown beyond a certain level, the ability to manage, market, and sell the products becomes a crucial requirement for growth.
This is really just so true. I've seen so many would be "CEOs" getting distanced from the code far too early, they just want to run a business and make a bunch of cash. All the great companies have coders at their heart for the longest time, listen to their users, build a dedicated community and most importantly create a _great_ product.
Of course, this depends on the stage of the start-up. If you want to start a start-up, yeah, for the first 9 months or so, all that matters is product development, so if you can't help with that you're no good.
After the product gets to a sufficient state, though, there are other things than coding to do for the company. Sales, marketing, deals, etc... Those can use a business CEO.
Actually, I disagree. If you have a CEO that does not fully understand the company's product to a sufficient technical degree, then you have someone who is leading the company with no idea of what is possible. What nearly always happens in companies when this is the case is the development group starts making outlandish claims ("that feature will take 30 man-years and $9m to develop", etc.) with no one to call BS on them.
Personally being able to code (even if I'm not the sharpest at it) has led me to be able to drive teams a lot faster because I can say "well, if I can develop this in a week and I suck at programming, then you should be able to get it done in a day - come back tomorrow and let me know how it went".
That said, you're right - after a product is on its way it's all about sales and doing deals. But you still have to be able to whip the parts of the organization that need whipping (mind you, that's not always the development team either).
Surely that's the job of the CTO rather than the CEO? If the CTO is supporting those 30-man-year estimates, the problem is not with the "development group" - and by that stage it's very reasonable for the CTO and the CEO to be two different people.
Great book. Read it. Two-sentence description: product development is essential, of course, but "build it and then market it" is way too risky. Instead, do customer development alongside product development, from day 1.
Absolutely - That's exactly what we've done on my second business, and exactly what we failed to do on my first business. I won't be making that mistake again.
However, the level of customer development that you need to do while you're building the product is less than what you need to do while you're in the second start-up phase of selling it.
This might be true for a class of startups, but not all. For the first 4 months or so of my latest (CAPEX intensive) venture, 'all that mattered' was developing the business case and raising funds.
I agree though; the merits of a technically capable CEO can't be overstated.
The key is that he knows code. Depending on the size of the organization you run you don't have to be a regular coder, but a coder you must be.
I've seen far too many CEOs and managers who are technically inept, and as such completely gloss over great technologies and business opportunities, while at the same time pursuing impossible goals with no understanding of the technical underpinnings of their business.
I'm not so sure that a CEO/Founder needs to know code, but I do think that a Founder of a startup company in the software realm should understand where programmers are coming from.
This post seems to be written from the perspective that if you are developing a product that utilizes code, and you have no money to spend, then clearly, you must write the code. However, some of the comments about this article aren't talking about this aspect, but rather that people who aren't coders, don't understand coders.
This, I don't think, is necessarily true. I believe that by keeping up with what coding/technology can provide, while also understand a programmer's philosophy, a noncoder can facilitate an environment that supports constructive collaboration between both the coder and the noncoder.
In short: I don't think you necessarily need to have deep coding skills to start a tech company utilizing programmers. I do think you need to have a great idea, an open mind, and a willingness to step inside the shoes and take the perspective of the coder.
