First 100k users are always the hardest
Download as PPT, PDF10 likes2,648 views
The document provides tips for building and scaling a startup based on Rashmi Sinha's experience with SlideShare. Some key points discussed include: focusing on solving problems well and making the app social; the importance of speed in development, design, and releases; listening to behavioral metrics over just the loudest users; outsourcing complexity; and understanding potential market size and revenue models from the beginning. Execution is emphasized as far more important than ideas.
First 100k users are always the hardest
- 1. The first 10k users are always the hardest Rashmi Sinha, SlideShare
- 2. Who am I?
- 3. SlideShare... best place in the world to share presentations
- 4. Taking over the world (one PowerPoint at a time!) Start May 06 > Beta > Launch Oct 06 > Grow > Grow more > Angel funding Dec 07 > Series A Apr 08 > Grow more
- 5. 1. Make your app useful
- 6. Solve one problem really well
- 7. Make app social SlideShare as non social app PowerPoint plugin? No virality
- 8. 2. Speed is imperative
- 9. Why startups need speed?
- 10. Small & fast
- 11. Set the tone for speed In hiring, development, design, releases, biz dev
- 12. Be agile (worry about scaling later)
- 13. 3. What to build
- 14. Why we built products we did MindCanvas: Could sell into market SlideShare: Understood market
- 15. 37signals: Build less
- 16. Build for right reasons Start with less Add complexity (when necessary) Remove features Get people used to change
- 17. To build a feature or not to Listen to loudest? Watch behavioral metrics? Read blog posts?
- 18. Loudest users OR silent majority slidecasting example
- 19. How to double traffic ?
- 20. Anatomy of a feature launch Groups: V1, V2 & V3
- 21. 4. How to Launch
- 22. The Alpha Feedback to the embed No publicity Browser compatibility issues
- 23. How developed should a Beta be? SlideShare beta Enough to get point across Little of social layer Not buggy
- 24. SlideShare launch Invite only: Before or after launch? Alpha Stealth mode Open to all Launch Invite only
- 25. Initial Momentum
- 26. 5. Use metrics
- 27. What metrics? Performance: page load times... Errors: 404, 500s... Traffic: Web logs Database tracking: No. registrations, login, uploads, contacts Customer feedback emails
- 28. Get yourself a shadow app Metrics we care about Simple console with daily, hourly Simpler than web analytics Whole team uses it Measure project success, iterate Proxy AB testing
- 30. 6. Scaling Community is hard
- 31. Community is a great way to grow No marketing costs, users evangelize Constant stream of feedback
- 32. Scaling community with millions of users Users help each other?
- 33. 7. Outsource what you can
- 34. Outsource complexity Amazon S3, EC2 Dedicated servers instead of colocation Google Ads Google Search Different widgets (Gigya, AddThis)
- 35. 8. Working with remote teams
- 36. Challenges Getting to know each other Hard to develop informal rapport Need to look at same screen. Body clock Benefits: 24 hour cycle
- 37. Synchronous is must Skype GoToMeeting
- 38. SCRUM End of day reports
- 39. 9. Ideas are dime a dozen
- 40. NO EXECUTION = $1 WEAK EXECUTION = $1000 SO-SO- EXECUTION = $10,000 GOOD EXECUTION = $100,000 GREAT EXECUTION = $1,000,000 BRILLIANT EXECUTION = $10,000,000 AWFUL IDEA = -1 WEAK IDEA = 1 SO-SO IDEA = 5 GOOD IDEA = 10 GREAT IDEA = 15 BRILLIANT IDEA = 20 Ideas are a multiplier. Execution is all that matters X http://www.oreillynet.com/onlamp/blog/2005/08/ideas_are_just_a_multiplier_of.html
- 41. Care about execution Scrutinize ideas: “can we execute on this?” How fast? When can we start work? What will first version take? Which team member will do it? What metrics to track progress? What will we push back to work on this?
- 42. 10. Understand market size (& revenue model)
- 44. Potential market size No of PowerPoint presentations Blogs & Video
- 45. Potential revenue streams AdSense Affiliate Networks Sponsorships Fremium Licensing White label version Marketplace Job board Professional services
- 46. Good early legs Adsense Consulting revenue
- 47. Revenue stream impacts application Advertising Pro accounts & Fremium White label solutions -Revenue stream should not kill virality -Don’t loose focus on majority
- 48. Problems with fremium
- 49. VC money? Retain control till you can Take money (if you absolutely need it) From right partners
- 50. Finally slides at http://www.slideshare.net/event/webvisions-2008/slideshows find me at www.slideshare.net/rashmi We are hiring!
Editor's Notes
- #4: Why choose Describe three popularity methods
- #13: What type of hosting Developers love it Its good for fast, agile teams We hit scalability issues, but Twitter hitting them first! So far, its been solvable Not enough CPU. Doubled spend on machines. Split Web servers from App servers As system gets load, some parts of system get used more, code needs to be optimized Database scaling issues currently Solving problems as they arise, but architecting for scaling Dedicated hosting provider is compromise Having tech resources outside organization you can call on Reduces initial expense Good sys admins harder to find than good devs
- #18: Customer requests? At any given time List of 15 to-do features Many bugs to fix Customer requests
- #45: This workshop is more about building large-scale web products, rather than web products for niche markets, consulting scenarios