I met a set of founders recently that were solving a real problem for a known market. So far so good.
The solution the team implemented was technically trivial. It was a simple CRUD app that, if built on top of a web framework like Rails or Django, could probably be implemented in 7-10 days by a…
By Bruce Sterling’s definition, Design Fiction is “the deliberate use of diegetic prototypes to suspend disbelief about change.” Experiencing the future through “speculative objects and services” in films or science fiction is part of his rhetorics.
This reminds me of the cyberpunk genre that I have been reading a lot, a science fiction movement that Sterling is at the center of: books from William Gibson, Rudy Rucker, etc. It also reminds me of films like Minority Report, Google Glass’s advertisements, Cloud Atlas, Blade Runner, Iron Man.
These design/science fictions help us visualize a future as a species. While some of them have dark, post-apocalyptic themes, I also believe that they are pushing us, as a human society, forward – by giving all of us a chance to dream.
Some of these dreams are already a reality. Companies like Oblong Industries (G-Speak) and my friend Jason Sosa’s IMRSV (Cara) are at bleeding edge of the next wave of infrastructure evolution. Not to mention, G-Speak powers all the screens in Minority Report.
In recent months, I have repeatedly returned to the theme of intertwining human societal evolution and information evolution. A quick survey of the recognizable systemic evolutions helps me see this cycle: establishment of infrastructure follows by emerging ecosystem. Oblong’s CEO Kwindla Kramer states, “The next next thing will give us the platform for the multi-device era.” I think he is spot on.
Information Evolution
All technologies go through an S-Curve. Bruce Sterling identifies each stage as R&D, startup phase, maturity, decline. I can’t contain my excitement to be witnessing and to be a part of the beginning of the next evolution.
The time to intervene is, now.
The other day we talked about First Mover Advantage with some teams. I’m not a big believer in First Mover. After probing a bit further, I agree that First Mover does exist in certain cases, and in others, it may actually be a disadvantage.
Here is how I see it breaks down. Based on Christensen’s Disruptive Innovation model, First Mover may apply in sustaining innovation as opposed to disruptive.
Disruptive innovation by nature implies a steep education curve. True disruptive innovation creates new sets of behaviors. This is why the businesses have to spend a lot of resources in teaching and converting users.
Sustaining innovation on the other hand captures current behaviors and extends it by technology. The model I have in mind is BJ Fogg’s Behavior Model.
A good example of how First Mover works in sustaining innovation is AirBnB. AirBnB relies on the necessity and our understanding of looking for housing over the decades. That behavior has already established itself in the intraweb through the vast popularity of Craigslist. Andrew Parker’s observation of the Spawning of Craigslist aptly describes this. Offering temporary lodging is the flip side of the coin, and AirBnB captures it through its momentum as a First Mover.
On the other hand, scenarios where Wesabe lost to Mint, Marc Hedlund concluded that “Changing people’s behavior is really hard.” Other cases include Yahoo! losing the search war to Google, Friendster & Myspace to Facebook. These company out executed their predecessors.
Some people call this the Second Mover Advantage, and the link includes some other examples.
Yesterday I met with a group of entrepreneurs and held Startup Clinic workshop at HUBBA for the whole day. It was a crash course of JFDI’s 100 days. We had great time. I really enjoyed the topics of conversation, and shared some of the oft discussed ideas at JFDI.Asia.
The guys were extremely smart and passionate about the verticals they worked in, and before the morning was over, the group was helping each other out in constructing one-liner pitch of their businesses.
We started with a round of introductions, and I asked the group to pitch us who they were and what they were working on. We then spent the whole day working on breaking down how to articulate their businesses, and the problem & customers they were targeting. All of these were things I did everyday with the JFDI teams.
Some of the entrepreneurs started with product pitches or mission statements instead of explaining what their businesses did. They confused the “customer pitch” with the “investor pitch”.
I saw the same mistakes while on the panel the day before at Thumbs Up’s conference. We were listening to a lot of product features as opposed to what the businesses did. I pointed this out at the conference then, and yesterday again at the workshop. We worked hard to change that.
I also shared with the guys how we constructed 6 minutes investor pitches. I used examples from current JFDI teams: the flow of explaining the business through Problem (Customer), Solution, Traction, Market, Business Model, and Vision. This helped with understanding how to construct the one-liner.
In the afternoon, we helped each entrepreneur fill out Lean Canvases as a group. I loved the energy and bond in that room. Everyone was asking each other hard questions and giving great constructive advices.
Through all of our conversations, and answering questions, we ended up with an amazing list of blog posts, videos, and tools (see below) to help the guys carry on the momentum from the workshop. I was very proud of how much ground we covered; it was also very gratifying to see the progress the guys made just in one day.
Most of our discussions surrounded Lean and Customer Development. We focused on: how to ask customer development questions, what kind of hypotheses to come up with, customer segmentations, identifying early adopters, unfair advantage, theories on innovations, behavior models, metrics, etc.
Thanks to Mimee (Thumbs Up) and Amarit (HUBBA) for making this workshop possible. It was extremely cool. I’d love to lead another workshop in Bangkok!
Here is the entire curriculum and notes from our session:
Our goal of the day – Start Now: http://www.youtube.com/watch?v=1h6RsLxro9s by Derek Sivers
Agenda
Introductions:
Ray @raywu
Link @jomzup
Pound @teeravee
Seth @gnosisadvisory
Numfone @bongboeaw
Of @ojazzy
Lim @csengathubba
Py @pmuenpra
Aim @aimamarit
Who are you pitching to?
Notes:
Homework:
Blogs to check out:
Smoochy & me during the workshop. Thanks Numfone for the photo!
This is a talk that I gave in Bangkok this week at Thumbs Up’s Start It Up, Power It Up #4
Lean Startup Customer Development Level: Tomb Raider
At JFDI.Asia, we help over 60% of our teams raise on average 500,000 dollars, including a Thai startup you may know, ShopSpot. Everyday I work with our teams and help them design weekly experiments.
It is not always easy. Sometimes the results from these experiments run against the teams’ assumptions, and even challenge their product visions. When that happens, we spend hours arguing about local maximums or sample bias, and deciding what to do next.
Over time, and especially in the last 80 days of our current bootcamp, I have come to appreciate some observations I made from working with our teams.
I started to think that running a Lean Startup is as if you are playing the video game — Tomb Raider — and before long, I will tell you what I mean.
At the end of this talk, I hope to leave you with three steps that we go over weekly with the JFDI founders:
This is what we do week in, week out: to become Tomb Raiders.
Lara Croft: Ultimate Domain Expert
Through working with hundreds of entrepreneurs, I really like the analogy that Lean Startup is a game of inference.
I imagine it’s like being Lara Croft in Tomb Raider, where as the game progress you become better and better at reading clues and unlocking secrets.
Similarly, Lean Startup helps you become a domain expert. The goal is to be Lara Croft, the ultimate Tomb Raider.
And your clues? They are from the conversations you have with your customers. It’s about getting to know your customers really, really well.
Believe me, this gets me really excited. I felt like my 12 year old self playing Tomb Raider II.
Tomb Raider Unfair Advantage: domain expertise
As the saying goes, ideas are worthless, and it’s the execution that counts. We believe that domain expertise is what set successful teams apart from others.
Your early adopters and customers will tell you what you need and what to build strategically – and these are the clues. They will provide you with insights that help you recognize an opportunity to revolutionize a market with the technology you are excited about.
By repeating the three aforementioned steps and continuously running experiments and collecting clues, you are able to become domain experts systematically.
Your Goal: collect clues
The only way to succeed is to get your foot in the door. Your beach head is where your early adopters are.
As a Tomb Raider, you are collecting clues to solve a puzzle. After you have done a lot of these data collection, you start to recognize a pattern.
Your goal is to become the Lara Croft of tomb raiding.
Step 1: prioritize customer segment
This is the first of the three steps to become Lara Croft, first you must identify the tomb you want to raid: form a hypothesis and prioritize your customer segment.
JFDI Mentors like Kevin Dewalt and Justin Wilcox are two of the most diligent customer development practitioners. The first thing they did with our teams was to help them segment their target market.
Focus on three key factors: Market Size, Accessibility, and Will they pay for value?
One of our teams, Duable, when they first started at the program, were not sure which tomb to raid. Duable is revolutionizing the Chinese Learning market: there are 120 million Chinese learners as of today. It is a big market.
This came with plenty of ancillary markets, too: including the teachers, tutors, the students, parents of students, online learners; Duable needed to decide.
Justin Wilcox taught them to assign a number between 1-3 to Market Size, Accessibility (channel), and how much are they willing to pay for value to each segment. You tally the score up by multiplying them. He has a blog post about this on CustomerDevelopment Labs which you should check out.
This helped Duable prioritize their customer segment.
Step 2: map your risks
The next step to become Lara Croft is to understand your risks:
When you have a target customer segment — basically your tomb — you need to minimize the traps by looking at a map.
Your map is the Lean Canvas. It helps you visualize your entire business model.
At the early stage, you may only have assumptions about a few items: the most important ones are:
When our team Collabspot joined JFDI, it was growing at a steady single digit pace. They were making CRMs accessible in Gmail. We asked them to go through this exercise of mapping out their business model.
What kept Collabspot up at night, was his channel. The Lean Canvas helped him mapped this out.
Step 3: design your experiment
The last step in this three steps process to become Lara Croft: design your plan of attack.
A tool I’d recommend here is the Validation Board, it helps you zoom in on your risk from the map.
To pull all three steps together, our team Krake did very well. Their risk involved whether or not they could identify behaviors that suggested their value proposition was appreciated: they were going after recruiters.
They believed that recruiters were wasting a lot of time qualifying candidates. It was a statement most recruiters resonated with. But most recruiters weren’t doing any thing about it.
Only after interviewing 50 recruiters, did they find out that the independent IT recruiters were hacking boolean strings on LinkedIn. They did this because they only got paid when they placed a candidate fast – their motivation was the time constraint. These savvy recruiters even had a LinkedIn group to share boolean tricks!
In the end Krake pivoted away from this market, but it was because through learning about their risks, they realized that their expertise did not match the requirements of this particular problem.
As a bonus, Krake moved away from recruiting industry and took what they learned about behaviors of IT recruiters and applied to startups and SMBs – they are now growing their user base at double digit weekly and helping to bring big data to individuals.
Conclusion: as long as you are learning
I don’t want to give you the cliché Lean Startup conclusion where I tell you to embrace “fail fast.” One key thing to remember: you do need to evaluate what you learn from the experiments.
What I want you to walk away with is this: Lean Startup can be fun and rewarding, just like when you completed a stage in Tomb Raider. You can become a domain expert while you look for treasure in a tomb: eventually your clues will solve the puzzle.
Be Lara Croft.
Yesterday I had dinner at a friend’s office with his whole team. My friend was an experienced entrepreneur, and his current startup launched a few months ago. He made amazing food for us.
After everyone was done, casually he offered to show us some numbers he had been looking at. As a leader of the pack he invited the team (CTO and four developers) to take a trip down the metrics lane. I was along for the ride.
He made word clouds to help the engineers get excited about trends they spotted in their user base. It was funny for the guys when he pointed out the obvious; some ethnic last names were well represented. As he drilled down the keywords on the word cloud on users’ activities, the team jumped in and surmised different premises as to why they were seeing these trends.
He showed nice graphs of user traction and adoption. In cohorts, he tried to help the team understand how their key metrics were performing per week. They were doing well. One guy pointed out that there was a bump in acquisition one week because of a push that they had done. Another asked questions about the discrepancy in retention rates week over week. Together they concluded it was because of the differences in the sample size. He patiently helped everyone up to speed; people were asking and answering each other’s questions.
The atmosphere around the table was fantastic, and it was a great treat to see my friend in his element. I admired how transparent and honest he was with everyone.
This was a reason why working with experienced CEO was so fun. The team was learning and being challenged to look at numbers together through product manager’s lenses.
Developers all want to work with star CTOs to hone their skills. The same principle applies here. Work with a great product manager and you will learn a great deal, too.
Meng has introduced me to many sci-fi books. He’s an avid reader.
A theme kept on resurfacing in these books. Both Terry Bisson’s They Are Made Out of Meat and Rudy Rucker’s Software examined the disconnect between our mind and body.
“The weak are meat the strong do eat.” — Dr Henry Goose, Cloud Atlas
This was an apt description of how Rucker and his peers viewed where the world was headed with technology.
It was also fascinating to trace Danny Hillis’s description of evolution reflected in Rucker’s depiction of bopper anarchy. Everything came back to Conway’s Game of Life. Amazing parallel with Darwinism.
All of them argued that artificial intelligence could not be designed, but rather evolved.The Game of Life demonstrated that petri dish of sub-routines could exist, and Rucker had a wonderful description of in Software.
What happened to this line of software building? Danny Hillis said they used to evolve software this way at Thinking Machines in the 90’s.
AI movement started in the 1960s at MIT and quickly spawned to Stanford. The history of AI was chronicled in Steven Levy’s Hackers, but since reading the book I lost track of the path the industry took.
I never really understood it from a technology stand point. It seemed daunting.
Maybe now is the time to look into it.
We have been hosting Startup Clinics to meet with entrepreneurs that we haven’t a chance to work with during JFDI.Asia Bootcamp.
Quite a number of people came through. We asked them to give us presentations, with or without slides on their businesses. As a group we would try to get a good picture of the entrepreneur’s viewpoint of the world.
I enjoyed these exercises though at times it was mentally draining. Every so often we fell into the rabbit hole of entrepreneurs defending their visions instead of trying to help us understand what the world needs.
This was something I encountered at times with the founders at JFDI. I believed that founders should have a stronghold on their visions. For many, this was the reason they got out of bed each morning. The difficulty was, how to let go of the details of that overarching vision, and to allow early adopters’ needs define the foundation of the products.
Too often we got into a heated debate, referring to qualitative and quantitative evidences. In the end, I finally submitted that these customer development data were there to help make informed, educated guesses. Guesses, at best.
Hugh told me that working with early stage startups was akin to raising children. Sometimes you saw a cliff coming headlong. You could warn them. You couldn’t make them take another route.
Great entrepreneurs will listen, but will follow their instinct.
“In hindsight, I think David was right and I was wrong. I wanted him to build something that felt more like Wordpress or Typepad (where I blog). He had something different in mind. And to David’s credit, he had the courage of his convictions to follow his own instincts.” — Fred Wilson
I watched Cloud Atlas for a third time yesterday.
It has been a privilege to have access to media files so easily. The last movie I saw three times was Avatar.
These were both movies with a lot of details and many moving pieces. For Cloud Atlas, I watched the film again and again because of its complex story line (hadn’t read the book). For Avatar, I saw it in 3D and in 2D to relive Pandora, and really enjoyed the big cinema screen for that
Three years have elapsed between the two films. Although AppleTV, Netflix, Amazon, YouTube all existed in 2010, I was under the impression that films were getting released sooner and sooner after its debut in cinema. A quick look online for data didn’t amount to much.
When asked where the Internet was headed, Jason Kottke responded with a link to the movie Network. Was that too dark of an outlook for a medium or were future of all media predictable from past performance?
The way it was going, the content distributors would undoubtedly become the future generation of “networks.” In fact, Hulu was one audacious leap the US TV networks undertook. This may be a glimpse of how corporations were always fighting for controls. A defiant theme in many sci-fi stories.
For both of these films I mentioned, the Seaboard Corporation / Nea So Copros and the Resource Development Administration were the oppressors that upheld orders in their respective worlds. Ultimately, these corporations became the roots of apocalypse.
While I don’t 100% subscribe to pure hackerism anarchy, I do hope that the media companies will still allow us the flexibility today’s Internet offers. Don’t be evil.
Preparation is tough. I saw all the JFDI.Asia teams prepared for the mock pitches today. Most of them came a long way, and much of it had to do with the fact that they spent everyday thinking about their pitches. Everyday to prepare for 6 minutes.
Contrast it with my own preparation for talks that I gave in the past. A lot of times I had to prepare materials for 10-20 minutes talks. On a couple of occasions, I did 3 hours long workshops. Most times I failed to build up my preparation and ended up scrambling for the last minute.
Very often, I was still putting together slides, rehearsing for the very first time, and didn’t know my content by heart when I stepped on stage. That was bad. I saw how founders fumbled their content and that was how I must have sounded: underprepared and overwhelmed.
Steve Jobs was perhaps every founder’s idol when it came to presentation. He came off very casual, but his delivery was always superb. He must have prepared so thoroughly for each of those performances that no one could tell it was practiced. He was a great orator, and he was able to deliver his thought on the spot even under attack.
It was probably from years of delivery and used to being in front of a crowd that allowed him to stay calm. I was most impressed with him taking his time and paused to gather his thoughts before responding to the ambush.
We all want to get to that level of coherence on stage. It will take years of practice for stage presence, and diligent preparation for the content. But we must work on the content, first.
A couple of my friends from New York are now in, or are getting into accelerators this year.
About six weeks ago, TechStars NYC announced its 2013 class of 11 startups. There were 1,700 companies from around the world that applied. With this type of scope, per chance I know two teams well. I also found out today that another friend from NYC got accepted into the upcoming YC class. It is not by accident.
A lot of my friends have worked extremely hard since the beginning of NYC tech boom. That was when New York Tech Meetup was less than 350 people in attendance, and Eric Ries and Lean Startup were still relatively unknown.
Many people regard getting into an accelerator as the mecca. They say, wow, these guys are amazing. In the midst of the hype, it is easy to overlook the fact that many of the New York entrepreneurs were new to this startup thing in 2009.
Thinking back on my friends’ journeys to date, I also took a minute to reflect on mine and how I got to know them. You see, I took the full plunge in 2011.
Everyone I knew then were grinding. They were working day-jobs, and building things at night. We bonded through this. I made a lot of wonderful and likeminded friends. They were the main reason that kept me going.
There were also sacrifices. I moved out of my comfort zone to save money. I lost a lot old friends that didn’t quite get why I was exhausting myself. I gave up on a lot of partying and relationships. I got involved with learning Rails. I volunteered at Sparkseed, co-hosted Ruby Nuby, and joined Lean Startup Machine. I did StartupBus. I stretched myself really thin. Eventually, I had enough savings to quit my corporate job and go full time. The next thing I knew, I was spending time in Boulder, and then in Asia.
By no means was it easy. I slept 3-4 hours a night for most of 2011 & 2012. It was the same for many people I knew, and they put themselves out there.
I am very happy with the fact that I now work with amazing people and startups in our portfolio. Equally excited to see old friends making their dreams happen with guidance from some of the best programs in the world.
I am really lucky. I knew it was going to take a few years to get something going. Often we won’t know when the next “break” is going to be.
This journey, of pursuing dreams, is never going to end. I’m glad to have taken the plunge.
Reed Hastings is one of the more controversial executives.
I recall the day that I dropped off Netflix. It wasn’t because I didn’t love their service or because of Reed Hastings. I dropped off because I moved out of the US in 2012.
There was a huge uproar regarding Qwikster just shortly before my departure. Up to that point, I hadn’t given too much thought. In fact, it didn’t bother me much because I didn’t care for the DVD rentals. I merely remembered thinking the name Qwikster has to be the worst in the world. The tech world seemed to think the same.
But many people felt otherwise. According to the NYTimes article that featured the company and its comeback since the Qwikster debacle, Netflix had lost 800,000 subscribers and the stock price dropped 80%. It wasn’t the first time Netflix had experimented, but definitely one experiment to be remembered.
Image Source: New York Times
People cared enough to abandon the service. It was a case of utter surprise for me, and it must have been the same for the Netflix management. I doubted that they hadn’t done their home works on its user demographics. Perhaps the same way that Facebook changes its privacy settings, they were hoping that they could crush all other competitors by moving subscribers and studios up and away from old behaviors and beliefs.
In hindsight, this must have been a reality check for Reed Hastings and his team. They learned that they couldn’t rush it. They had succeeded because people cared enough and were telling them so.
It is an extremely competitive arena that Netflix is in. Given Amazon, Apple, Youtube are all competing on the same eyeballs, it is incredible what Reed Hastings and co. have achieved. When asked about the course of the events, he reflected,
“Don’t get distracted by the shiny object,” he said. And if a crisis comes, “execute on the fundamentals.”
“The future has been shrinking by one year per year for my entire life” — Danny Hillis, 1993
Autocatalysis, a word I learned today from listening to Danny Hillis’s TED talk from 1994 in Monterey, California.
It was a word he used to describe the current technological leaps that we experienced through the advent of mediums: of Internet, and an entire lineage of past medium shifts, such as CD-ROMs, Telephony, to language itself.
I particularly enjoyed the perspective he offered in terms of the evolution of information. Perhaps McLuhan prescribed a similar explanation?
Hillis described evolution in terms of the history of information storage mediums. At first there were droplets of chemicals that stored information, then they became cells. Cells furthered their chances of survival by grouping and exchanging information/DNA with other cells. Through evolution these cell groups morphed into communities.
The communities allowed cells to specialize in functions; eventually they became interdependent, and they formed what we called, organisms. The evolution of information took on the next stage in an interesting turn through the development of neural systems within these organisms.
Evolution of information at that stage no longer depended on genetics or the guidance of survival the fittest. In humanity’s case, it became a matter of learning and storing information, and of exchanging information. We developed languages to communicate those ideas with each other. This was when the mediums started to accelerate.
The rest was history where we went from language to print, to morse code, radio, telephony, internet.
All of this, was autocatalytic. The product of the evolution only accelerated the next product in line.
His talk put the image of Conway’s Game of Life in my mind. Especially the way he described programming, and the application of this type of evolution in creation of software. This was the future in 1994, and still is, in our distant future.
As autocatalysis goes, we will need to learn and give up controls of our medium.
Here is the TED talk by Danny Hillis: Back to the Future (of 1994):
I recall when I used to try and grow a big online Twitter presence, we were using TweetAdder. I hated the service because it was spamming by following anyone, everyone, and it caused new Twitter accounts to get banned.
Recently I noticed a startup favorites tweets by many other people I don’t know online, related to a blog post I wrote. I thought to myself, they must be targeting keywords in those tweets.
It seems smart, and here’s my logic:
After checking it out online, I learned:
Here’s what I tried, and it works:
VERSION BUILD=6011206 RECORDER=CR
URL GOTO=https://twitter.com/search?q=growth%20hacking&src=typd
TAG POS=1 TYPE=B ATTR=TXT:Favorite
TAG POS=2 TYPE=B ATTR=TXT:Favorite
TAG POS=3 TYPE=B ATTR=TXT:Favorite
TAG POS=4 TYPE=B ATTR=TXT:Favorite
TAG POS=5 TYPE=B ATTR=TXT:Favorite
You can probably read that block of code easily. It basically tells the macro to go to the URL and run these queries. Note that this job searches the term “growth hacking” and favorites the newest five tweets. Oh, you also need to be logged into your Twitter account.
If you can run this job periodically (can probably automate this further with a script), you can get a list of prospects on Twitter and potentially a new following. Another way to hack this is to build it with Twitter API and GET the latest “search tweets” and see if you can do a POST for favoriting them. Might be too much engineering.
Hope this will generate some leads & a following for your Twitter account. Be careful out there, don’t get banned!
Marketing and growth hacking have become startup household terms in the last few years.
Recently at JFDI.Asia, we’ve been doing a lot of courses on hacking the funnel. Yesterday I came across an awesome slide deck by Sean Johnson @intentionally.
Ruby has a great ecosystem, and there are many wonderful contributors that created content for people to learn from. I found that learning the ropes of growth hacking a little more challenging. There are less forums (this one on Quora is great) and overall fewer resources in terms of books or interactive tutorial or codeacademy type of courses.
I signed up for a weekly email subscription from Montana Flynn @montanaflynn, and I’m excited to receive tips for the next 9 weeks.
But I can’t help and wonder if there is an opportunity here for someone to create an interactive course on Growth Hacking to teach all the different aspects of this vast topic.
Maybe even a scaffolding that lets you plug in GA and set up goals, do SEO, SEM, A/B testing. It may even have a sandbox to simulate traffic being driven to the site and gamify the whole thing.
Perhaps it has been done/being worked on already? Anyone knows?
I met a set of founders recently that were solving a real problem for a known market. So far so...
After learning my flight was detained 4 hours,
I heard the announcement:
If anyone in the vicinity of gate 4-A understands any Arabic,
Please...
Frequently seed stage founders ask me my opinion of whether or not they should...
In Israel, they’re experimenting with a novel business model: subscriptions to unlimited cups of coffee at...
‘You can’t connect the dots looking forward; you can only connect them looking backwards. So you have to trust that...
Feynman diagrams explained. For a more in-depth look, see A Guide to Feynman Diagrams in the Many-Body Problem.
And don’t miss Edward Tufte’s...
New York
(This post introduces howtoprogram.tv.)
In eighth grade, at the peak of the tech bubble, I won a family fantasy...
