From Zero to a Million Users - Dropbox and Xobni lessons learned
- 1. Zero to a Million Users@drewhouston, @asmith
- 4. Summer 2006: Humble beginnings
- 5. Xobni, Dropbox – VC-backed startups in SFBoth reached 2 million users in 2 years2 ,000,0000
- 7. Biggest risk: making something nobody wants
- 8. Weak product-market fit cannot be fixed by good marketingGood marketing will not fix a bad product
- 9. But good product will be wind at your back
- 10. Xobni Analytics: Solved problem no one had
- 11. Goal: product-market fit ASAP
- 12. learn early, learn often
- 13. Cost of change at each stageIdeaPrototypeLaunch
- 14. Learning on a $0 budget: Talk to people!
- 15. Learn without launchingAdWords tests Hacks: Fake landing pages, screenshots, etc.
- 16. Dropbox’s minimum viable product: 3 min screencast on Hacker News (Apr 07)
- 17. Simple landing page: capture interest/email address
- 19. Go where your early adopters hang out
- 20. Private beta launch video 12,000 diggs; beta waiting list jumps from 5,000 to 75,000 in one day (Mar 2008)
- 21. Avoid ghost towns: bootstrapping to critical mass
- 22. Niche first, world later
- 23. Fake it till you make it: Aardvark & simulating features that weren’t ready
- 25. Tie yourself to a bigger trend
- 26. AdamMeet journalists in person. All the time.
- 27. Journalists are really busy – come up with the angle for them.
- 28. Media resources page on web site
- 29. Word of mouth rule #1: have a great product
- 30. Generate word of mouth with scarcity
- 31. Help users generate tell their friends
- 33. Auto-ranked
- 34. Outlook detection via looking at mail headers = FUN!
- 35. Sex it up! Have fun!
- 36. Inbound marketing / “social media”
- 37. how do you get people to use your product?
- 38. Focus: do a few things really well instead of a lot of things poorly
- 39. Every 10% easier 50% larger audience
- 40. Don’t make me think: No decisions nothing to screw up
- 41. Don’t make me read, either:Designing landing pages & signup flowsearly betaconcise beats comprehensive
- 42. call out the next step
- 43. simple converts betterlate beta
- 44. Hook the user first, educate over time(tours, tip emails, etc.)
- 45. You do not see the same product your users do! (craigslist, usertesting.com)
- 46. Make feedback painless – then iterate!
- 47. how do you make a product that spreads itself?
- 48. The best products turn users into evangelists
- 49. Encouraging word-of-mouth: Referral program increased signups by 60%
- 50. Tricks of the trade: FB & Twitter feed, emails, contact list importers Trailing 30 days: 3 million invites sent
- 51. metrics: know thy funnel
- 53. Startup Metrics for Pirates. Live it.
- 54. Output I: know where to invest marketing $$
- 55. Xobni finding: AdWords users don’t convert to paid BUT they refer others who do!
- 56. Output II: allocate engineering resources between usability, engagement, & virality
- 57. Example: user education around people search
- 58. scaling without virality:“dollar in, dollar out” marketing
- 59. Some markets don’t elicit virality or buzz
- 61. Netflix example(numbers not exact)($16 rev - $10 svc cost) per month $6/mo x 22 mo lifetime= $132$40 advertising, affiliates, etc.$92 lifetime gross margin/user
- 62. a quick taste…
- 63. example user acquisition costs
- 64. example monthly rev per user
- 66. example freemium conv. rates
- 67. Thank you!Questions?Feel free to reach out!@drewhouston - drew@dropbox.com@asmith - adam.smith@xobni.com
Editor's Notes
- #14: Xobni analytics -> prototype stage.Segway example -> millions spent on launch; best investors in the worldCould have realized much earlier that the problem wasn’t the technology, it’s that you look like a douche while ridingGoal is not to never fail, it’s to fail as early as possible
- #17: So shipping code was out of the question – bad experience tell all friendsYC app – ship in 8 wks vs 18 mosPrototype worked; video could show product in best light; get much of the same feedback as if we shipped working code
- #18: Video created fans; capture that interestEarly adopters more forgiving
- #20: Building product for ourselves; where would we look for something like Dropbox?
- #22: Google Wave
- #24: AardvarkexmapleReddit users faking comments to avoid ghost town feeling
- #37: Always lose feature comparisons
- #38: One invariant of our installer
- #39: One invariant of our installer
- #51: “I don’t mean to imply that you want to gamble your VC dollars..”
- #53: Engagement: email people who haven’t used certain features, etc. Active user versus registered user.
- #56: What kind of markets tend to be dominated by DIDO marketing?
- #57: Virality can e.g. double TLV.
- #58: Virality can e.g. double TLV.