2. How I ended up here
• Developer for 10 years
• Graduate of Computing Systems
• Running 2 startups, some success
• Bootstrapped, no outside investment
• Actively involved with other startups
• Organiser of StartupMill events
6. People who defined
my thinking
• Eric Ries
• Steve Blank
• Sean Ellis
• Dave McClure
• Look these guys up!
7. Eric Ries
The Lean Startup
• Use of platforms enabled by open
source and free software
• Application of agile development
methodologies
• Customer-centric rapid iteration
8. Minimum Viable Product
• Version of a new product which
allows you to collect the maximum
amount of validated learning about
customers with the least effort
• Avoid building products nobody
wants
• Probably much more minimum than
you think
9. Eric Ries
The Lean Startup
• The “pivot”
• Looking back at the beginnings of
successful companies shows they
often started as a different idea
• Many examples: PayPal, Flickr, ...
17. Steve Blank
Customer Development
• http://bit.ly/FourSteps
• http://custdev.com
18. Sean Ellis
Product/Market Fit
• Comparing results across over 100 startups,
those that struggle have not reached
“Product/Market Fit”
• Sean’s metric: ask existing users how they
would feel if they could no longer use the
product. 40% very disappointed = P/M Fit
• Engage existing and target users to learn
how to make your product a “must have”
23. Dave McClure
Pirate Metrics
• More from Dave McClure:
• http://500hats.typepad.com/
24. Learning
• Embrace change, avoid assumptions
• “Pivot” & iterate through feedback loop (Eric
Ries)
• “Get out of the building” (Steve Blank)
• Strive for Product/Market Fit first (Sean
Ellis)
• Use metrics, not opinions (Dave McClure)
29. Split tests
• Test a change against an existing feature
• For example, a new landing page versus the
old one
• Show 50% of people the old one and 50% of
people the new one, and test the metrics to
decide which is better
30. Continuous
deployment
• The time taken for code to be committed to
the repository to it being live on the
production server is less than 20 minutes
• Code is passed through tests in order to
determine whether it should go live
31. How I see lean, and
ways to bootstrap
• Stop wasting time
• Activities aren’t always wrong, just
often timed wrong
• Just do it, there are too many
reasons not to start
32. Lean & bootstrapping:
my experiences
• 2 things which worked for me
• 2 things I will now always question
• 3 examples of keeping things lean
33. Worked for me:
Open source
• Use a framework, careful with a CMS
• Use a familiar language
• Save learning a new one for a side
project which isn’t a business
• Use other people’s code
34. Worked for me:
3rd party services
• It makes no sense to do non-core activities
• Email - sendgrid, mailchimp, ...
• Deployment - beanstalk, github, ...
• Metrics - mixpanel, KISSmetrics, ...
• Feedback - uservoice, GetSatisfaction, ...
• Hosting - AWS, Rackspace, Heroku, ...
35. I will always question:
delaying a launch
• I didn’t think I was ready to launch OnePage
or Buffer when I did
• OnePage - 4 months
• Buffer - 7 weeks
• “be notified when we’re ready” works well
• Users are very forgiving
• Question every feature
36. I’ll always question:
“closed beta”
• “closed beta” for far too long with OnePage
• No closed beta for Buffer
• Use it only when it makes sense
• You have people desperate to see what
you built next
• You have a very clever social hook
37. Keeping things lean:
Launching incrementally
• Start small, measure throughout
• Example: Buffer
• Started with “coming soon” page
• Added full landing page, no other pages
• Then added pricing page, still no app
• Finally added the app
38. Keeping things lean:
No fancy signup process
• Buffer has no signup process yet
• Launching it without was scary
• Pleasantly surprised with outcome
• Better to launch, measure conversions and
then build a better signup process
39. Keeping things lean:
Integrating PayPal
• PayPal has a feature to allow you to
automatically switch a user’s account level
in your app when they make a payment
• Maybe considered necessary, I launched
without implementing it
• Better to spend the time getting the first
customer
40. Act tomorrow:
Starting from scratch
• Keep your MVP minimal
• Don’t get carried away with code
• Measure before you build more
• Focus on qualitative feedback
41. Act tomorrow:
Already launched?
• Add in actionable metrics (Dave McClure)
• Lots of traffic? Start split testing
• Check out Optimizely.com