PyCon Ireland 2013 - Python Adventures in Startup Land

(via http://python.ie/pycon/2013/talks/python_adventures_in_startup_land/)
Speaker: Rachel Willmer (http://twitter.com/rwillmer)

5 years ago, I started a side-project to help me find ebooks I wanted to buy.

This is now Luzme.com, and I'lm turning it into a real business, writing large quantities of Python/Django along the way.

This talk will share some of my technical experiences; what I've tried, what's worked, what hasn't; and how I'm designing to meet the current challenges of making it faster and making it mobile.

Transcript

1. PyCon IE, Oct 2013 Python Adventures 
 in Startup Land Rachel Willmer rachel@luzme.com ! Slides: http://slideshare.net/rwillmer Website: http://luzme.com
2. PyCon IE, Oct 2013 What is Startup Land? • You aim for “Minimum Viable Product” • You don’t have a long-term development plan • You write code you know you’ll throw away • You know that you don’t know what you’re doing
3. PyCon IE, Oct 2013 Why Go There? • It’s fun! • It’s creative! • It doesn’t have to be Facebook/Twitter • It could be your own business
4. PyCon IE, Oct 2013 What is a Startup, anyway? “A Project in search of a Business Plan that works”
5. PyCon IE, Oct 2013 Which Means Continuous Change • Design Changes • Test Changes • Deployment Changes • EVERYTHING changes
6. PyCon IE, Oct 2013 “Practice Trumps Theory” Ash Maurya, Author of “Running Lean” http://practicetrumpstheory.com/
7. PyCon IE, Oct 2013 The Agile Manifesto, 2001 • http://agilemanifesto.org/ • KEY POINTS: – Our highest priority is to satisfy the customer through early and continuous delivery of valuable software – Welcome changing requirements, even late in development... – Deliver working software frequently, from a couple of weeks to a couple of months ... – Simplicity, the art of maximizing the amount of work not done, is essential – And others….
8. PyCon IE, Oct 2013 What This Means in Practice • Have a long-term Big Idea • Plan only short-term work plans, a “sprint” • Continuously engage with customers • Be very willing to change everything as you learn • Change != Failure • Change at an early stage = cheap
9. PyCon IE, Oct 2013 It’s a marathon, not a sprint © http://www.flickr.com/photos/wwarby/4782201357
10. PyCon IE, Oct 2013 The Lean Startup, 2011 • Term coined by Eric Ries, based on earlier work done by Steve Blank • KEY POINTS: – Product/Market Fit – Minimum viable product – Continuous deployment – Split testing – Actionable metrics – Pivot
11. PyCon IE, Oct 2013 Lean Links • http://theleanstartup.com/ • http://runninglean.co/ • http://leananalyticsbook.com/
12. PyCon IE, Oct 2013 Fail Early, Fail Often, (Fail Cheap) • “Fail” = “Try,Learn,Change” not “Failure” ! “Good judgement comes from experience, But experience comes from bad judgement” Mark Suster, http://www.bothsidesofthetable.com/
13. PyCon IE, Oct 2013 How to Fail Early ! • Don’t Wait until it’s ready until you talk to your customers • Customer Discovery interviews • Minimum Viable Products • Verifiable Hypotheses • Landing Pages • Ask for money early!
14. PyCon IE, Oct 2013 And Now… The Python Bit…
15. PyCon IE, Oct 2013 Luzme • Luzme is the ebook search site • 1000 registered users worldwide • USA, Canada, UK, Australia, India • 25 VMs in 4 countries • 1 dev/ops/support/marketing/sales = ME ☺ • Written in Django/Python • Started in 2008
16. PyCon IE, Oct 2013 • Switching to OmniGraffle for architecture walkthrough…
17. PyCon IE, Oct 2013 And now in words… • Original: webserver + database + cron job • #1: Extra worker nodes: scraping, API lookups • #2: Datafeed node • #3: Overseas workers (geographical location matters!) • #4: Queueing (celery/rabbitmq) • #5: Caching (redis) • #6: Search (elasticsearch) • #7: Host Monitoring (zenoss) • #8: Business metrics (logstash, graphite, statsd)
18. PyCon IE, Oct 2013 QUESTIONS? • Slides are available on slideshare • http://www.slideshare.net/rwillmer