- Machine Learning for Plant Properties — startup building database of plant genomics, properties, research, etc. for mining. The more familiar you are with your data and its meaning, the better your machine learning will be at suggesting fruitful lines of query … and the more valuable your startup will be.
- Dissecting Message Queues — throughput, latency, and qualitative comparison of different message queues. MQs are to modern distributed architectures what function calls were to historic unibox architectures.
- 1915 Data Visualization Rules — a reminder that data visualization is not new, but research into effectiveness of alternative presentation styles is.
- The Broken Promise of the Mobile Web — it’s not just about the UI – it’s also about integration with the mobile device.
"startups" entries
Four short links: 20 August 2014
Plant Properties, MQ Comparisons, 1915 Vis, and Mobile Web Weaknesses
Four short links: 11 August 2014
Startup Anthropology, Ends to Means, Permission to Test, and Distributed Systems Research
- Anthropology of Mid-Sized Startups — old but good post about the structures, norms, and dimensions of startup culture. Like a religion, a startup will care for its collective interest by defining certain things as sacred. A classic example is the company’s logo. This symbol is, quite literally, “set apart and forbidden” by brand guidelines, which often specify exactly how the logo must be presented and how far it should sit from the other elements on a page (thus separating the sacred from the profane).
- What Leads To — I love the elegant mechanic of decomposing an end back to a means you can do right now. Lots more sophistication obviously possible, but the fact that it’s not just about “thumbs up this end!” or about actions divorced from intention, makes it a step ahead for social software.
- Researching Link Rot (Pinboard) — graceful notification of a test, and with the simple ability to opt-out.
- The Space Between Theory and Practice in Distributed Systems (Marc Brooker) — I went through everything I’ve read on distributed systems and arranged them on a spectrum from theory to practice the two ends would be really well populated, but the middle would be disturbingly empty. Worse, changing to a graph of citation links would show a low density from theory to practice.
Four short links: 7 July 2014
GV Library, Blockchain Equity, Organisation Anti-Patterns, and Cognitive Biases in Software Engineering
- Google Ventures Library — collection of design, engineering, founder docs.
- SWARM — crypto equity. Stock via the blockchain. (via Jesse Vincent)
- Organisational Anti-Patterns (Leigh Honeywell) — failure modes involving power and labour.
- Cognitive Biases in Software Engineering (Jonathan Klein) — failure modes for estimations, testing, and evaluations explained with psychology. Because brains.
Four short links: 12 June 2014
Our New Robot Overlords, Open Neuro, Anti-Surveillance Software, and LG's TV Made of Evil and Tears
- Norbert Weiner (The Atlantic) — His fears for the future stemmed from two fundamental convictions: We humans can’t resist selfishly misusing the powers our machines give us, to the detriment of our fellow humans and the planet; and there’s a good chance we couldn’t control our machines even if we wanted to, because they already move too fast and because increasingly we’re building them to make decisions on their own. To believe otherwise, Wiener repeatedly warned, represents a dangerous, potentially fatal, lack of humility.
- Open Ephys — open source/open hardware tools for neuro research. (via IEEE)
- Startups Selling Resistance to Surveillance (Inc) — growing breed of tools working on securing their customers’ communications from interception by competitors and states.
- Not-So-Smart TV (TechDirt) — LG’s privacy policy basically says “let us share your viewing habits, browsing, etc. with third parties, or we will turn off the `smart’ features in your smart TV.” The promise of smart devices should be that they get better for customers over time, not better for the vendor at the expense of the customer. See Weiner above.
Four short links: 28 March 2014
Javascript on Glass, Smart Lights, Hardware Startups, MySQL at Scale
- WearScript — open source project putting Javascript on Glass. See story on it. (via Slashdot)
- Mining the World’s Data by Selling Street Lights and Farm Drones (Quartz) — Depending on what kinds of sensors the light’s owners choose to install, Sensity’s fixtures can track everything from how much power the lights themselves are consuming to movement under the post, ambient light, and temperature. More sophisticated sensors can measure pollution levels, radiation, and particulate matter (for air quality levels). The fixtures can also support sound or video recording. Bring these lights onto city streets and you could isolate the precise location of a gunshot within seconds.
- An Investor’s Guide to Hardware Startups — good to know if you’re thinking of joining one, too.
- WebScaleSQL — a MySQL downstream patchset built for “large scale” (aka Google, Facebook type loads).
Four short links: 23 January 2014
MSR Open Access, Light Table Open Source, Virality Unleashed, and Holacracy's Founder
- Microsoft Research Adopts Open Access Policy for Publications — +1
- Light Table is Open Source — this matters because these experiments in semantic interactivity inform technical UIs of the future, and the more ubiquitous this code is then the more effect it can have and the sooner we can have the future.
- The Six Things That Make Stories Go Viral Will Amaze and Astound You (New Yorker) — Berger and Milkman found that two features predictably determined an article’s success: how positive its message was and how much it excited its reader. The obvious part is that we develop immunity to things that catch our attention: our brains are well-developed systems for filtering, and the only constant is that advertisers will need novelty.
- The Story of Holacracy’s Founder (Quartz) — background on the interesting flat organisation culture system that’s gaining traction in startups.
Four short links: 5 August 2013
Startups Class, Container Deployment, Cryptopocalypse, and Program Design
- EP245 Downloads — class materials from the Udacity “How to Build a Startup” course.
- scrz.io — easy container deployment.
- The Factoring Dead: Preparing for the Cryptopocalypse — how RSA and Diffie-Helman crypto might be useless in the next few years.
- How to Design Programs — 2ed text is a work-in-progress.
Not Your User’s Problem
Understanding the Difference Between User Problems, Business Needs, and Solutions
First, let me say that I love all the emphasis on Customer Development, Early User Research, and Product Market Fit that I’ve been seeing these days. What I don’t love is the massive confusion that often comes along with it.
There’s a particular type of confusion I’ve seen on teams at the very beginning of the product development process that I’d like to try to clear up. Or possibly add to. We’ll see.
Some people don’t seem to understand the difference between a Business Need, a User Problem, and a Solution. But you have to understand the difference, because if you don’t, you’ll end up doing the wrong sort of research and designing the wrong product.
A Business Need
At its very simplest, a Business Need is what a product will do for your company. This can often be expressed in the form of a metric that needs to be moved or a hypothesis about how building a new feature or product will make you a billionaire.
Here are some examples of business needs:
- Improve the conversion rate on a landing page so that we get more people trying our product.
- Increase revenue by selling more widgets.
- Get more registered users for free by getting our current users to share our product.
- Increase engagement with our product so that people are more likely to be retained users.
- Build a huge user base so that we can eventually monetize it.
What’s interesting about these Business Needs? Well, in one way or another, all of these things, if executed correctly, should eventually increase our revenue or decrease our spend. We need to do these things to have a viable business. But there are all sorts of ways to do them, some of which are great for users and others that aren’t.
To identify a business need, typically you’re going to want quantitative data. You need to know what your metrics are in order to figure out which ones need to be higher. You don’t determine a business need by talking to users.
Obviously business needs might be caused by user problems. For example, if your onboarding process is hard to use, you could have low conversion rates. But the business need is increasing the conversion rate, which you might do in a number of different ways.
A User Problem
Your users have problems. Some of the problems they’ll pay you to solve for them. Some of the problems you’re probably causing for them with your terrible UX. Some of the problems they don’t even know they have.
Here are a few examples of user problems:
- It’s hard to share documents across different computers.
- The first time experience with a particular product is confusing and complicated.
- The user can’t use an app when it’s not connected to the Internet.
- A person has trouble finding a good hair salon in her area and booking an appointment.
You’ll note that these user problems are all quite different. The first one inspired lots of companies, like DropBox. The second one is common to many products. The third one is mobile specific. The fourth one could be solved by a number of different types of products, some of which are quite low tech. There are roughly an infinite number of other user problems that could exist.
The common factor here is that these are problems experienced by humans. The other common factor is that there is no guarantee that solving a user problem will actually fulfill a business need. Sure, solving problems for people is generally a good thing, but there are some user problems that people will pay you to solve and others that they won’t.
To identify a user problem, your best bet is observational and generative research. Watch people in the wild using your product or other products. Follow people around while they perform various tasks or do their jobs. Understand the things that make life difficult for people and then identify the biggest, most important problems that you could solve for them.
A Solution
A solution, as the name implies, is how you solve a problem. Ideally, your solution will solve a user problem which will fix a business need.
Here are a few examples of solutions: Read more…
Head Games: Ego and Entrepreneurial Failure
“Ever tried. Ever failed. No matter. Try again. Fail again. Fail better.” —Samuel Beckett
Entrepreneurial success hinges in large part on a founder’s mastery of psychology. This requires the ability to manage one’s responses to what Ben Horowitz calls “The Struggle,” that is, the emotional roller coaster of startup life. Paul DeJoe captures the ups and downs of being a startup CEO in a post reprinted in a book that I edited, Managing Startups: Best Blog Posts.
It’s all in a founder’s head: the drive to build something great; the resilience to dust yourself off when you repeatedly get knocked down; the passion powering a Reality Distortion Field that mesmerizes potential teammates, investors, and partners. But inside a founder’s head may also be delusional arrogance; an overly impulsive “ready-fire-aim” bias for action; a preoccupation with control; fear of failure; and self-doubt fueling the impostor syndrome. That’s why VC-turned-founder-coach Jerry Colonna named his blog The Monster in Your Head. In a recent interview with Jason Calacanis, Colonna does a nice job of summarizing some of the psychological challenges confronting entrepreneurs. So does a classic article by the psychoanalyst Manfred Kets de Vries: “The Dark Side of Entrepreneurship.”
Read more…
How to Get Fast, Cheap Feedback on Your Product with Tiny Tests
User research you can do now
There’s a lot of advice about how to do great user research. I have some pretty strong opinions about it myself.
But, as with exercise, the best kind of research is the kind that you actually DO.
So, in the interests of getting some good feedback from your users right now, I have some suggestions for Tiny Tests. These are types of research that you could do right this second with very little preparation on your part.
What is a Tiny Test?
Tiny Tests do not take a lot of time. They don’t take a lot of money. All they take is a commitment to learning something from your users today.
Pick a Tiny Test that applies to your product and get out and run one right now. Oh, ok. You can wait until you finish the post.
Unmoderated tests
Dozens of companies now exist that allow you to run an unmoderated test in a few minutes. I’ve used UserTesting.com many times and gotten some great results really quickly. I’ve also heard good things about Loop11 and several others, so feel free to pick the one that you like best.