Webstock – New Zealand’s Web Conference – Day 2

I had a great day at Webstock on Day 1 and below is my synopsis of Day 2, which was also excellent. If you get the opportunity to attend Webstock, do it. Some of the key themes from the conference can be summarised as:

• Build with love, use common reusable blocks, deploy quickly, deploy often, listen to explicit and implicit feedback from real users, understand what their needs are (not what they say they are), consider social and ethical impacts. Repeat.

The Lean Startup

Eric Ries (@ericries)

• http://startuplessonslearned.com #leanstartup

• Slides can be found at http://www.slideshare.net/startuplessonslearned/2010-02-19-the-lean-startup-webstock-2010

• Why build a startup?

  • Change the world

  • Build an organisation of lasting value

  • Make customers' lives better

• Pivot: change directions but stay grounded in what you've launched

• Speed Wins: If we can reduce time between iterations we increase chance of success.

• In start-ups, it's all about the team

• Shadow Beliefs:

  • We know what customers want

  • We can accurately predict the future

  • Advancing the plan is progress

• If you're building a product that fundamentally no one wants, what does it matter that it's on schedule?

• New plan (e.g. of a startup that worked)

• Shipped in six months – a horribly buggy beta

• Charged from Day 1

• Shipped multiple times a day (by 2008, on average 50 times a day)

• No PR, no launch

• Results: 2009: profitable, revenue > $20MM

• What activities are value-creating and which are waste?

• The unit of progress with a waterfall process is getting to the next stage.

• With Agile the Unit of Progress is a line of working code

• Product development at lean startup unit of progress : Validated learning about customers ($$$)

• Minimise TOTAL time through the loop: Ideas-->Build-->Code-->Measure-->Data-->Learn-->Ideas

• Continuous Deployment

• Deploy new software quickly

• Tell a good change from a bad change (quickly)

• Revert a bad change quickly

• Work in small batches

• Break large projects down into small batches

• Cluster Immune System

• Run test locally (Simple Test, Selenium)

• Continuous Integration Server (BuildBot)

• Incremental Deploy

• Alerting and predictive monitoring (Nagios)

• When customers see a failure

• Fix, Improve your defences

• Five whys route cause analysis: Ask “why” five times when something unexpected happens.

Iterative Design Strategies

Daniel Burka (@dburka)

• “You're going to have an ugly baby” While you love your project, it could always be better, embrace feedback.

• If you review your first site version and don't feel embarrassment you spent too much time on it

• Take chances and release. Build with expectation of change. Listen and iterate.

• “How buildngs learn: What happens after they're built” is a good book

• With modular architecture you can build a castle in no time. This is key to rapid design.

• Don't waste time trying to predict everything

• Good Designers Redesign, Great Designers Realign e.g. don't dramatically change how people use a site.

• Subtraction is iteration too. Try to remove as much as you add – don't be afraid to prune

• Really listen to your users: Both explicit and implicit feedback are crucial

• Digg success:

  • Get it out there

  • Add sophistication

  • Keep revising

  • Establish goals, Measure success, Set new goals

  • Customer focus group, Task analysis, Makes changes, Launch

  • Stay fit: adapt to survive and thrive - If iterative design isn't instinctual be convincing

  • People adapt well to seeing people struggling with a system. It can be used as a way to convince people a change is required.

Double Click to Edit

Amy Hoy (@amyhoy)

• Sweaty Design: because genius is 99% perspiration.

• Is consistency necessarily good? This can be a barrier to evolution. e.g. all blog readers look the same, n to-do list iPhone apps.

• “To affect the quality of the day, that is the highest of arts”: Think about the effect your site/app has on others each day

• Invisible design is a crock of shit. e.g. Who wants an invisible Lamborghini?

  • [sg] Some things should be invisible. This is really just semantics.

• We don't need permission to do the best we can do

• Greatness multiplies greatness

• You're creating change: whether or not you want to

• Understand how people use tools and what their needs are vs just copying what others are doing

• If you can't feel the pain you can't fix the pain.

When Your Idea Doesn't Suck: How to stop working for clients and launch a startup

Mike Davidson (@mikeindustries)

• Mike did not want anything to be tweeted, blogged or recorded so I respect that. Whilst his speech did have some good content I didn't think it warranted being so protected.

Designing for Participation

Bek Hodgson - Blurb, Etsy

• Etsy – marketplace for homemade stuff

• When designing layouts consider how the page will be displayed if little or lots of content (e.g.a user's profile)

• Sets / Collections can be useful to group lots of content. Started with company providing taxonomy then moved to user driven folksonomy.

A Baroque Aria – 10 Tips for Entrepreneurs

Kevin Rose (@kevinrose) - Digg

• 1. Go Build It!

• Running ideas past others isn't always the best idea

• You don't need anyone's approval

• If you believe in something, work nights and weekends, it won't feel like work

• 2. Build & Release

  • List out your features for six months, sort them by importance, and build 3-4 of them

  • Stop thinking you understand your users – learn from what they're actually doing on your site, not what you think they do

  • Build, release, iterate, and repeat.

  • Feature not working, how core is it? Kill it?

• 3. High your Boss

  • “Get people on board that understand your vision and what you are trying to create”

  • Senior positions: Only hire people you'd personally work for – make sure they can help build your vision

  • Junior positions: Hire hustlers, people that will run circles around you

  • Demand excellence: Top 1%

• 4. Raising money

  • Beg, borrow and steal for as long as possible

  • Don't just take the cash, can they add value?

  • Super angels (50-200k) no board seats

• 5. Go Cheap

  • Shared servers to test the waters

  • S3, cheap, fault-tolerant

  • No office (8 months no office)

  • Shared cell phone plans

  • Shared apartments (3 digg employees)

  • Part-time A players (@dburke)

• 6. Connect with your community

  • Start a podcast (it's OK not everyone listens)

  • Throw a launch party, then yearly/quarterly events – invite the press/influencers personally – don't tell the bar (tell them it's your birthday; cheaper)

  • Engage with the community

• 7.Hack the Press

• Invite only system (Pownce, Digg v3)

• Talk to the junior bloggers

• Attend parties for events you can't afford – network with influence, bring a demo

• 8. Advisors

  • What technical problems are you going to have?

  • Advisors can be helpful in a whole slew of areas (marketing/hiring/bizdev)

  • Stock compensation, typically not a board seat, solid advisors help during fund raising

• 9. Leverage your user base to spread the word

  • Facebook notifications is a great example of this (e.g. Farmville and helping out farm).

  • “Tweet to complete” - used by wefollow. This asks user if they want to tweet.

• 10. Analyze your traffic

  • Install Google Analytics

  • Entrance sources (search?)

  • Paths through site

  • Surveys

  • Top exit pages

Elements of a Networked Urbanism

Adam Greenfield (@agpublic) – http://doprojects.org

• What do people want from technology to improve their daily life?

• Consider the social and ethical consequences of Ubiquitous computing

• “By the end of 2012, networked sensors will account for 20% of non-video Internet traffic” (Gartner)

• When you walk past a camera, assume it's on, you have no privacy any more in public spaces

• Information persists

• Real-time sensors are becoming more prevalent. e.g. Visualisation of night-time activity in San Francisco (http://citysense.com)

• The likes of Foursquare can drive usage patterns e.g. people will return to places just to stay the mayor or get a badge

How the Web Works

Jeff Veen (@veen)

• Disruptive innovation: From ice to refrigirators, From typography to CSS to ...

• We need to be “native to the web”

• Shipping code is the way to get consensus / converge

• Get your site out there, then iterate. iterate, iterate

• Speed of iteration beats quality of iteration

• The Robustness Principle: Be liberal with what you accept... and conservative in what you produce.

• Separate interfaces from engines

• Prototype before polishing. Get it working before you optimise

• Design for the future, it will be here sooner than you think

  • [sg] Need to be careful to not over engineer though

Dense and Thick

Mark Pesce (@mpesce)

• Sharing makes something more valuable

• The web is “All things to all people”

• The genius of Apple's iPad design is its simplicity (it is not trying to be a computer). This will impact what the Web is about to become.

• “Augmented Reality is about to make a transition. We will soon wonder how we lived without it.”

  • e.g. Pointing a phone at a book and getting information about it

  • The device becomes the focal point

  • The book is about to be subsumed by the network

  • e.g. Beef Mince

  • Traceability back to the animal, inc carbon footprint, feed…

  • These questions will need to be on-demand to minimise our footprint on the planet

  • e.g. Medicine

    • Get info about medicine, such as if it is a poison and how it complements other medicine you're using

    • An interface to ourselves is where we are moving to