Archive | February, 2020

The Iowa Caucuses. A study in how NOT to throw technology at a process.

4 Feb

This Iowa Caucus mess will make one hell of a business case of what not to do when you are faced with applying tech to a mission critical challenge. While what I’m about to say is constructed out of reading a bunch of news reports and picking up nuances and details in the first hours after the disaster, I’d say this is a pretty complete laundry list of the failures:

1) Start with a completely byzantine process that was engineered in the early 1970s and has never been actually rethought, just incrementally changed to become even more complex

2) Decide to throw an app at the process, without considering that your userbase will consist of largely techno-phobic post-retirement volunteers

3) Start late and go long on your development process

4) Be so afraid of hacking that you literally do not allow your users to test your product prior to using it

5) Be so afraid of hacking that you also forgo full-scale stress testing of your infrastructure to make sure you have enough server-side strength to support the usage level you’re going to get

6) Deploy the app widely and simultaneously across a 1700 person user base and expect them to install the app on their personal smart phones, not knowing a) what version they might be on; b) how much memory and bandwidth resources they may have; c) if they actually even know how to download and install apps (note userbase demographics above). Do not conduct training, and document the app minimally.

7) Do not deploy the app for installation until the moment it needs to be installed and used

8) Deploy the app across a rural environment with thin cell coverage except in areas of major metros and major highways, again with mission-critical timing

9) Assume the 1700 locations that your 1700 elderly/technophobic users will be visiting all have solid, available wifi connections, OR in lieu of good wifi, have good cellular bandwidth (note point 8 above). Don’t bother visiting and testing the locations to verify connectivity.

10) Create a telephone support system but design it and staff it only for supporting the application, not replacing it

11) Don’t create a disaster plan that assumes the app is completely unusable

12) Do all of this while

a) you’re supporting an event that has global importance

b) the global media is watching

c) the national news is broadcasting live

And they wonder why this went pear shaped?

#digitaltransformation #applications #apptech #apps #Iowacaucuses

%d bloggers like this: