This is excellent advice. One of the challenges is that really big things can become overwhelming and intimidating, but small things are quite doable. When I was writing a column for JavaWorld I remember how 'hard' it was to write a 1500 - 2000 word column, but 'easy' to write a 2000 word email. Michael O'Connell, my editor there, suggested that I just write an email and let the editors make it into the column. I decided that was stepping too far back, but I could effectively break my writer's block by emailing myself the question I wanted to answer and then approaching the first draft as an email.
I would expect that much the same happens between 'startups' and 'weekend projects.' If you build a small project for the weekend you don't really care that you will be done on Sunday (or Monday perhaps :-) and that will be that. A 'startup' on the other hand that is the first step and then you have to drive traffic, and get buzz, raise funding etc etc. So it stops being a 'weekend project' and becomes the start of huge pile of work. Yuck!
This is so true. Just last week I was struggling with writing a press release and couldn't get started. While staring at my screen an email came through from an old friend asking what the scoop was on my new project. I immediately dove into the email and began writing what would later become our press release. At the time though I was just responding to a friend. Funny how that works.
You're spot on with weekend projects vs startups. I was fighting that mindset when I developed "nbaschedulesapp": drive traffic, get buzz. But lately I managed to escape that thought and the weekend project slowly becomes just that :).
This is a solution to a problem I have. I am a procrastinator.
No matter how much I "want" to do something I always rationalize why it could be better left till later. The way I combat this is by making the tasks I need to complete in a day small and easy so that doing them becomes insignificant. I normally have nothing against the actual tasks I need to complete. In fact I usually enjoy them. But the idea of tackling a giant problem gives me anxiety. If I start with only working for 5 to 10 minutes in mind it will usually extend to hours. If you are a procrastinator I suggest you try this strategy.
This is exactly how I overcame a tendency to procrastinate: decompose tasks to trivially small bits that are easily taken care of right now, at this moment. That gets you started, and that's really all you need.
That's the essence of GTD. For everything you have to do, you write down the precise "next action" that needs to be accomplished to move towards that goal. If you stick to it, it's very effective.
Though I am the greatest breaker of this rule, I agree so much with it.
Anytime I see a "Facebook killer" that has all the current features of Facebook I smile and tell the builder(if he/she is a friend) to observe that the sites after FB that ever got so big (Twitter, Instagram, Foursquare) had only one real feature before building out over time.
I am sure even Pinterest will become more complex, but for now, it is all about Pinning.
Nice article Joel, this point cannot be emphasized enough.
Absolutely. It's impossible to gain traction unless you simply your product and start with one feature. The problem that then arises is how to choose the "killer app" out of your entire model, something that is groundbreaking, evaluates the market opportunity, and leaves you with a cliffhanger to prepare for the next step in increasing complexity.
Nice article. I like how you pulled in some examples of major brands that started small, like Virgin.
As an entrepreneur, you're unfortunately constantly confronted by the push-and-pull of needing to think big, and then to build small. Investors (seem to) want to hear about big markets and big ideas, and at the same time want to see real product and traction - things that are hard to achieve unless you start small. Thanks for sharing.
The kids character, Thomas the tank engine, is a great example of this. It all started in 1942, with a dad telling his sick kid stories to cheer him up. The kid liked it so much, that his Dad turned it into short stories, and then books. Today there are 42 books, with the last one released in 2011, a TV show that has been running for 27 years, and kids all over the world know and love Thomas. And the little kid, now 72, is still involved in the franchise.
I don't like to envision big because I would be too intimidated by my big vision to start small. Starting with a modest idea helps me keep things concrete. "Envision big" causes me to start thinking abstractly prematurely. I tend to not write much code at all when I do that.
I can't stress that enough. I've been in the startup world for a couple years now, had my own success and failures, and saw lots of success and failures from others.. starting small and focusing on making it great is such a good advice.
If someone asked how he'd learn to code, the answer would probably turn around a small tutorial or a trivial program to write, not a WoW clone to start with. Oddly enough, first-time entrepreneurs often try to start with the Facebook or Instagram killer needing a massive growth and a huge funding.
It seems from the comments that some people tend to understand this advice as praising of the reductionist approach to problem solving, which is only one of the many methods of 'building' or 'engineering'. But I think the article talks about how intentional staying away from 'thinking big' doesn't prevent your product from becoming big - eventually, which is a different thing. Article doesn't try to explain why that is the case, actually, and reads more like a statement of an important fact with some prominent examples. And we know that there are examples to everything.
Also, it is easy to mistake 'starting small and growing big overtime' for 'starting small and staying small, but having a great impact on whatever side of life/business we choose' - kind of what folks from 37signals do (an example, yes).
That said, I think this is a very good advise: applied right and regularly, it helps you escape the procrastination period (uncertainty of 'big' may be too strong to overcome), skip needless research and preparations, avoid paralysis by analysis and other friends of 'thinking big'. The next logical step is reducing 'start something small' to just 'start something'.
This is all on the topic of gradual improvement vs doing a leap forward. Most of the things can be gradually improved from something small. Take as an example an issue tracker. You can start as small as you like (a file with a list of issues to deal with) and go towards a reasonable solution.
The problem is not every single problem in the world can be broken down into such pieces. Take for example a fashionable electric car from Tesla. There is no reasonable way how you can start an electric car from a gasoline car (put an extra electrict motor? you end up with hybrid. take away the gasoline motor? you end up with a horse cart) and tesla did not even have a gasoline car to start with.
There are both kinds of problem and I agree most startups, especially in the social world can be implemented in gradual steps that make sense. The problem with other problems is that first you have to understand the problem area really well. A very good start is to try the gradual way and fail at it, later to start a leapfrog approach, knowing the problem area.
The Tesla Roadster was a modified Lotus Elise with off the shelf components combined with an off the shelf electric engine and a container of common laptop batteries all coordinated with on board software. They used this base to design the Model S. Giant leaps or explosive growth are only called as such by naive people who weren't watching the thing being painstakingly set up over many years. Everything is a slow burn - there are no explosions - only ignorance and inattention.
Furthermore the design occurred mostly in simulations in a gradual, agglomerative, iterative fashion just like most software. It's why they can bring out cars so quickly nowadays.
Your argument is the same as the creationists. Everything is iterative evolution. There are no leaps of faith, miracles or gods - just directed random evolution.
"The Tesla Roadster was a modified Lotus Elise with off the shelf components combined with an off the shelf electric engine and a container of common laptop batteries all coordinated with on board software"
You call it evolutionary I call it a leap, IMO it still took quite a bit of effort. The point is nothing in between was working good enough to show as the famous MVP.
Giant leaps are also sometimes called by people who did the leap itself, even if it did not look a leap from the outside.
I remember there was an article a while back* about a valley vc lamenting that most founders these days are only interested in doing little "dipshit companies".
This is very, very good advice. Many of my most successful projects also started small-ish - including my first feature film, my 500k visitors/month website and my current project starring some Pretty Famous Actors.
Having said that, in the case of film I tend to have more of a problem keeping the damn project small long enough to release something - hence why my average time-between-releases is about 3 years. This seems to be something of an endemic problem in the pretty-pictures-moving-on-a-screen industry, as James Cameron and Joss Whedon, for example, demonstrate.
I wonder - does anyone have any good ideas for how to create an MVF (Minimum Viable Film)? Would be interested to have a discussion on this.
This is closely related to another story that's on the front page right now called "Nobody’s going to steal your idea" http://news.ycombinator.com/item?id=4732684
All my previous projects were big and they all failed. The reason I started them big and loaded with functionality is that I was afraid that if I start small someone else would steal the idea and leverage their network to get traction faster than me. That was a big time consuming mistake.
This is a subject that I am constantly challenged with while working on my ideas/projects etc. Every time I think I have the next facebook killer idea, I tell myself to go back to the drawing board because I am thinking way too much ahead and not starting small.
This topic is never going to be too cliched or repetitive. Very hard to get it right.
To help with me with starting small, I also have started a small habit and not to add new one until a few weeks later. Now I have habits in walking 10K steps a day, writing 500 words a day, as well habits that help me measure my well being in various ways such as BP, weight, pulse rate, blood sugar level, and so on.
Great insight. My "first project" was something way too big for a first attempt and crashed and burned before it even "launched". It was such a bad experience that I didn't really try anything for 3 or 4 years. Then I decided to try a couple of really small projects and it's been a much more manageable experience.
When you start doing something small, your minimum viable product is your most valuable product. It is all you have. If you don't make your MVP, you are missing something of a highest value.
A complex system that works is invariably found to have evolved from a simple system
that worked.
I'm not sure if it always works for a MVP; I think some technical/scientific ideas (as opposed to market opportunity ideas) need quite a bit of work before a user/customer can appreciate them. That is, ideas that are cool/intriguing, but don't meet a need.
Uh, I guess you could argue that a MVP at this stage is seeking collaborators (not users). e.g. Codd's first relational paper - peers were fascinated, but a deliverable took 10 more years.
But the relational concept itself was fully formed, and without that, he wouldn't have intrigued people. One could argue that this is to do with the "Minimum" threshold in "MVP". I'm just saying that an arbitrarily small MVP won't always work - there is a "minimum".
esr in CatB suggested similar to MVP for open source - but again, I think this also needs to connect with a need, or something the audience appreciates i.e. "something really neat". Historically, it seems hard for open source to start revolutionary new projects (revolutionary academic projects are often open source, but remain academic). It's also hard for big business - it often seems to take a startup to do it. Here's what esr said: http://www.catb.org/esr/writings/homesteading/cathedral-baza... (3rd paragraph):
Your program doesn't have to work particularly well. It can be crude, buggy,
incomplete, and poorly documented. What it must not fail to do is (a) run, and
(b) convince potential co-developers that it can be evolved into something really
neat in the foreseeable future.
Steve Jobs talked about people not knowing that they want something until you show it to them - so it has to be complete enough for them to recognize it. And "The Innovator's Dilemma" mentions many cases where the revolutionary new thing wasn't wanted by mainstream customers... however, to be fair, it did find a market with other customers - it started small, in a small, niche market.
Perhaps one of Carnegie’s most successful marketing moves was to change the spelling
of his last name from “Carnagey” to Carnegie, at a time when Andrew Carnegie
(unrelated) was a widely revered and recognized name.
I would expect that much the same happens between 'startups' and 'weekend projects.' If you build a small project for the weekend you don't really care that you will be done on Sunday (or Monday perhaps :-) and that will be that. A 'startup' on the other hand that is the first step and then you have to drive traffic, and get buzz, raise funding etc etc. So it stops being a 'weekend project' and becomes the start of huge pile of work. Yuck!