How we launched our MVP in 3 weeks
adamthewan.com
adamthewan.com
Is this meme the single greatest reason why the developed world has so few successful moonshots? Imagine if Elon Musk took this approach with Tesla/SpaceX/Boring/Neuralink, or Steve Jobs took it with the iPhone.
The best class of startups is where the key risk is not "is there a market?" but "can this be technologically pulled off right now?" For example, if in 2002 you had asked a random smart person "suppose I build a reusable rocket at 1/100th the cost of a normal rocket; do you think there's a market for this?" The answer would be "of course, but could you really build that??"
(With all due respect to the founder of MeetButter; some companies naturally fit the quick MVP approach, but they tend to not be technological moonshots).
Could you elaborate on that? I believe it is a very interesting hypothesis, but I don’t fully understand it
No I'm not. Suppose I told you my business idea was this: build a time machine. There's no time machine market right now. But do you think if I built one, people would be interested in it? Answer: hell to the yes. There is zero market risk in this idea.
The problem, of course, is that I have no feasible way of building a time machine. But thinking along these dimensions is better for entrepreneurs/VCs/the world. It forces you to work on things that actually matter/will move the needle. When they work, the VCs will appreciate it quite a bit and the world will appreciate it quite a bit. You also get to transfer your time from working on marketing problems/science experiments to working on engineering problems/real science experiments. I think this is probably a more fulfilling way to live as an entrepreneur than otherwise.
Not necessarily. Interest alone does not create a market. You need potential customers who have the means to buy what you're offering under conditions that make it worth selling to them.
For example, I might be interested in a time machine, but if it's going to cost you a trillion dollars to make one and I'd need 5kg of unobtainium to power each jump, I do not represent a (viable) market.
Your premise in this discussion seems to be that the best startups (in some sense that isn't entirely clear yet) are the ones that are aiming to be game-changers. However, I expect that in practice that vast majority of successful startups are not doing something truly profound, but rather something much more mundane that either has not been done before or can be done significantly better in some way by the startup.
Not "rockets" but "rockets at 1/100th of the price" which sort of takes into account your point.
That right there is my goal. At the moment I'm solidly the first. Any tips for moving the needle?
What limitations will you end up with? Once it is built, what will the user experience be? Can you change history? Can you travel to the past or only the future? Can you return? It is interactive or view-only?
Without actually spending the money to do the R&D and see what you can achieve, you have no idea what price people would be willing to pay for the experience, because you cannot tell us what the experience will be. So you have no idea whether the price point would be sufficient to recover the R&D costs. Sure, it gets you thinking of how to "move the needle". But... what needle? What problem are you trying to solve? Who has that problem? Can they afford the solution?
Your example falls into the exact trap that people try to avoid by creating a quick MVP - jumping to conclusions about ideas that the founders are 100% sure people will buy, and assuming that people will pay for it, without actually doing the work to prove it.
Then you can eliminate him, and take his time machine, and reverse engineer it to figure out how it works. Then you can travel further into the past, to meet your younger self, and give your younger self the theory and the blueprints to build such a time machine. Then, your younger self can patent it, and claim all legal intellectual property rights to it for the next 100 years.
And you can also tell your younger self to buy certain stocks in specific companies, that will enable your younger self with the financial capital to build such a time machine.
Thus, in the end, who invented the time machine? You did.
First I developed a box. I used existing box designs as inspiration. No frills on the first go, just a normal cubic shape. I started with cardboard but plan to iterate further on the material selection. Currently the time machine only goes forward at 1x speed but that's the challenge of being first to market. I tested with users and they said the box is uncomfortable so I made a bigger box and put a cushion in there. Now, users report a more comfortable experience. You have to understand as a founder, any new technology is going to have some rough edges.
After this 3 week period, I learned a lot about time machine prototyping and the desired user experience. My next steps are to take this to VCs, funded by pensions of dock workers, and raise a $100M round to further development. I'm confident with this funding, we'll be able to disrupt the time machine market entirely.
How about a series of weekly podcasts,videos or blog which does exactly that: take a 'real' situation (in this case meetbutter) and then simply go full throttle and pushing the reasoning to it's outer limits, until LOL occurs.
I would subscribe to that. Let me know if that becomes a thing.
I also love Monty Python, so this is just the type of humor I truly enjoy!
First I spent my whole life building the time machine.
After this life long period, when I finally finished it at a really really old age, I went back in time, to 10 seconds after I had decided to build a time machine, and I handed the time machine to my past self and then I died, but the past me lived on, with a time machine that from his point of view was built by him in 10 seconds flat. It was a great success and all rejoiced, including past me, future me and present me.
I worked for a group of assholes in 2011 or so who basically had Zoom'ish host-guest videochat working by 2012 and they wound up petering out. These were big-exit veterans making a kind of moonshot for the time, but for whatever reason ran out of gas (it certainly wasn't a money issue given the history of the founders).
The market can't be counted on to rise to the occasion.
Both the technology and the market side of this project seems to be well-understood.
This seems to be focused on design & UX innovation and uses a tech stack which is very front-end leaning.
With this in mind it absolutely makes sense that they used a rapid-prototyping and MVP approach. UX heavy projects are very feedback and iteration oriented.
I would say that both their tech and methodology choices are spot on.
Big companies are interested in high tech too. It’s not so black and white. But it is still all about defeating incumbents.
Or another way to put it is, do what your competitors won’t do, not can’t do.
The issue for the really tough things, as in "things that cannot be solved by VC-backed marketing", is to ask random smart people. There is a lot of domain knowledge and industry experience behind them. Still no guarantee tfor success, so. It also has the potential to block these experts from trying something "risky".
I think you are experiencing a bias. The developed world takes a lot of moonshots. Very few succeed. Many last decades. Here are some currently active moonshots.
* Novel Coronavirus vaccine
* Self driving cars
* Delivery drones
* SETI
* Commercial fusion
* General Purpose AI
* Quantum Computers
* "Quiet" supersonic airplanes.
* The search for a Theory of Everything
In a best-case scenario, you have something like SpaceX that was founded 18 years ago and raised $3.2B, including the first $100M from Elon Musk himself. Almost a decade in, they came very close to failing after the first three Falcon 1 launches didn't reach orbit.
In a less good scenario, Magic Leap spent $3B and 10 years trying to make their technology work, but ultimately failed to deliver on their original technical promise.
Most founders don't have $100M lying around or the option of walking into a VC firm and saying, "Hey, there's a market here. I don't know if I can do it, but if you give me a few billion, I'll try."
As a result, the vast majority of founders have to take the MVP approach. There's nothing wrong with that. If founders insisted on only taking moonshots, the outcome wouldn't be more moonshots. There would just be fewer incremental improvements.
3 weeks is the reason that innovative projects usually die on the vine at established companies. PM's and bean counters never want to spend more than a few weeks on a "sprint task" that might not work out. The exact opposite of the mindset you want in a startup
It took us 3 weeks in total from inception, prototyping, to setting up a full project like you said (linting, CI CD, AWS containers, database).
To be more accurate, the breakdown is more like this:
• 5 days of discussing, investigating the problem, and design thinking. We took our sweet time here :p
• 2 days of creating a prototype (in the article I mentioned a weekend, using Firebase and React)
• 5 days of reiterating prototype and design within a closed group of testers (family and friends)
• 5 days of MVP design (landing page, branding, etc) and proper project and infra set up (migrating to AWS, database, CI CD, staging and production environments, etc).
So that's about 3 weeks (minus two weekends).
Hope that made sense!
1976: Wozniak built Apple I in 16 months, but he actually started selling the schematics for the computer before that.
1981: Gates bought QDOS for $75k and hired Paterson to modify it into MS-DOS; that’s what he licensed to IBM within months.
1995: Musk built Zip2 in less than a year, selling it for $305M just 4 years later.
2004: Zuck built v1 of Facebook in 2 weeks.
Revenue generating products live and die by their name/branding, this one for me would be dead in the water for professional use.
I also thought "Meat Butter", but my next associations after that are semen and lube. I often wonder whether more marketing departments should hire at least a couple of 13 year olds to pitch names to and then see if they laugh or not.
Oh no! Our branding was totally tongue-in-cheek, we felt that it was fun. But if it's conjuring negative images when you see it, that's totally bad! I'll feedback this to the rest of the team.
I don't believe you can slap together something in 3 weeks and call it a day.
We did it to gather market sentiment and feedback for an idea, any signals or signs to show that we were in the right direction.
Your MVP is the start of a conversation with your users / target market. Reiteration and pivoting your initial MVP based on user feedback will slowly inch you towards product-market fit. That is the long, hard slog.
In regards to building low-hanging fruits (hackathon timeframe ideas), some low-hanging fruits have deeper roots. You might find deeper problems that give you better insights on how to build something that people really want.
These kind of general statements about how an mvp should take x weeks are pointless because each project is different. Which is an obvious statement. So is the suggestion that you should launch as fast as possible.
Launching quickly, failing, is a really good way of understanding time management on a macro scale for an entire software project/startup.
And its not just about pacing or knowing how long a feature should take to develop, but many many things like which feature to develop at which point because it might be much more difficult later on.
I agree, every project is different. I don't believe that my experiences and advice should be a catch-all for all projects and teams everywhere, there are too many permutations.
Without defining the variable X (for X amount of weeks) that you should take for an MVP, the general advice for MVP is to build as little as possible, launch fast, fail fast, and reiterate.
There are industries where the barriers of building an MVP are much higher. For example, in ML you need to train your algo huge amounts of data. In the automotive industry, you need to meet a lot of minimum requirements to even enter the market.
Thanks for the feedback! I'll send this over to the team. Branding is super important, and we'll fine tuning it to make sure we get it right.
"Zoom-bombing" became a thing because 9 digit room codes without an additional passcode was fine, before millions of new people started using it and suddenly the probability of guessing a valid room code shot up.
Nowhere near enough entropy with a 4 alphanumeric character room code to prevent this kind of attack if the app takes off.
If we never launched and got feedback from you guys, we'd never see the potential downfalls of the tech we created!
Yup, we're listening to all the feedback. Working on a few ideas on how to fix any room-bombing issue.
We hear you! We're tinkering on a few ideas (room knocking, private rooms, password locked rooms) that should make this more secure.
Thanks!
Thanks!
I built the blog over a weekend using Gatsby. It's super neat! Gatsby is pretty flexible, I'm even able to make things like filters and queries, add tags. SEO is super easy, and their image processing is top notch. Highly recommend it.
I connected it to Netlify so that it auto-updates every time I make a git commit. If you're not familiar with the term Jamstack, you should explore it. There's so much cool stuff being done these days with static websites.
I'm not saying it's impossible to launch a successful product by investing a couple of weeks, but it sounds as probable as winning the lottery.
I agree, I wouldn't invest a couple of weeks and just call it a day.
An MVP is the start of a conversation with your target market. That's how we're approaching it. It'll take more than 3 weeks to reach product-market fit, we're just super excited to kickstart the journey!
Oh. So...not an app for buying butter or a dating site for people with unattractive faces?
Thanks!
The product team is pretty small. We have a UI UX product designer, myself (mid-level full stack developer), and a jr front end developer.
> We just launched video functionality powered by Jitsi!
We're using Jitsi (which is open source) but are also testing out various third party providers such as Daily.co and Agora.io
We want to focus on improving the meeting flow experience and interactions, which is why we leverage open source or third party providers for the video conferencing tech.
Are those chat bubbles based on any open-source CSS / inspired by others?
It was built using CSS!
I made a GitHub Gist of it, just for you: https://gist.github.com/adamthewan/11af76e766308c46b11b6203c...
The gist is using styled-components, but you can easily transfer it to a CSS class :)
Yes, I love the Design Thinking process. I ran my own dev agency before, and would always recommend clients to do a Design Thinking workshop with my team. It's really the best way to break down a big 6+ month project into bite-sized pieces.