How to create the perfect prototype

Prototyping is hard. As a digital agency, Mint Digital has created a number of prototype applications for clients, however these have been some of our least satisfying projects. Whilst we have been inspired by the project and put in the extra mile to delight, things sometimes just don't quite click.

It's taken us a while, and we've been through quite an experience to get to this point, but at Mint, we think we have solved the issue in a way that works for us, for our clients and for the end product. And all it takes is 4 days!

We'll look at the problems that prototypes can face; slippage, lack of focus and bloat. We'll see how our solution materialised out of Mint's company culture in a way that truly surprised us. We'll recap those initial issues and how we solve them and finally we'll look over how Rails helps us as developers and how we can help ourselves when creating prototype applications. By the end, 4 days is all you will need to create anything!

Transcript

1. How to create the perfect prototype
2. Phil Nash @philnash MINTDIGITAL
3. What is a prototype?
4. A prototype is not: •  Polished •  Fully featured •  Production ready
5. A prototype is: •  Single minded •  Well thought out •  Ready to put in front of users
6. Prototyping is hard
7. Agenda •  The problems with prototypes •  How we came to a solution •  How to code a perfect prototype
8. Problems with prototypes •  Slip •  Bloat •  Lack of focus
9. Slip •  Lack of deadline •  Hard decisions get put off
10. Deadlines Are GREAT!
11. Bloat •  Trying to build out the whole project •  Extra, unnecessary features
12. Lack of focus •  No full time attention •  Other priorities
13. Finding a solution
14. How a solution found us
16. WebApp Weekender
17. Hacking
18. Going to the pub
19. Dressing gowns
20. Hacking
21. Hack week
22. Hacks over the years •  Snppr •  What the Tweet? •  Rafiki •  Quizables
24. Then we got serious
26. Hack days are the answer?
27. The story continues...
28. Hack weeks as a training scheme •  Tightly focused prototypes •  All round team agreement •  A clear proposition
29. Hack weeks with a client
30. Hack weeks with a client •  Deadline •  Tight set of features •  Focus
32. Problems with prototypes •  Slip •  Bloat •  Lack of focus
33. The perfect prototype •  Small, dedicated team •  Fast decision making •  Clear proposition •  A testable application
34. Coding the perfect prototype
35. Confession I lied, you cannot code a perfect prototype. Especially in 4 days. No software is perfect.
36. Coding the best prototype you can in a short space of time
37. Frameworks
38. Frameworks Use everything available to you to get a headstart. •  Rails •  Devise •  Active Admin/Rails Admin •  JQuery •  Bootstrap (if you must)
39. No time for new things
40. Templates
41. Templates Use everything that your chosen framework gives you to jump even further ahead before you even start. There are certain things that you will always need.
42. Templates $ rails new app_name
43. Templates $ rails new app_name -m http://example.com/template.rb
44. Templates Have a preprepared project that is ready to jump into.
45. Templates Concentrate on the domain problem
46. Testing
47. TAFT
48. Testing •  Tests are your insurance policy •  They allow you to be confident that, when you put even hastily written code in front of users, it will work •  You can write code quicker without breaking things
49. Deploying
50. Deploying We want to be able to deploy as often as possible to show our progress to the rest of the team.
52. Shipping •  Start shipping from day 1 •  Make it easy to ship •  Be aware of the limitations of your server
53. It's a marathon, not a sprint
54. It's 4 days, not 24 hours
55. It's a marathon, not a sprint •  Get some sleep •  Write good code, not more code
56. Throw it all away
57. Throw it all away Not before you test with users!
58. Throw it all away •  This is not your best code •  You will have made bad decisions •  Launching a full service will cause you pain in the future •  Iterate on the ideas in the prototype, don't launch the prototype
59. Coding the best prototype you can in a short space of time •  Frameworks •  Templates •  Testing •  Shipping •  Sleeping •  Iteration
60. How to create the perfect prototype •  Focus, over a short time scale •  Have a deadline •  Have the decision makers in the room •  Give yourself as many headstarts in the app as possible •  Ship it
61. Thank you
62. Any questions? Phil Nash @philnash http://philna.sh Mint Digital http://mintdigital.com