This week - OkCupid stealth-mode, Homophones, and what really happened with Healthcare.gov
OkCupid’s President Christian Rudder made headlines in the past week for his callous dismissal of concerns over not only Facebook’s admission to questionable testing methods, but that of OkCupid as well. In a blog post meant to defend Facebook’s tactics during a test that experimented with the emotional reactions of account holders based on changes to the provided content, Rudder explained that OkCupid also ran a test where match percentages were reversed to provide a sect of account holders with 90% matches where normally the algorithm would have rated them a 30% match.
This has cause a lot of uproar, obviously, because people don’t like hearing that they were duped. Rudder defended his experiment stating that he informed the experiment participants, who unknowingly received the mis-matches, after the experiment had run its course.
This is the one time where I think it is safe to say that pharmaceutical companies did something better than someone else. Rudder used pharmaceutical companies in a brief defense of his methods, but when testing medications participants are informed prior to participation even in the case of double-blind trials. While I understand the need for continual improvement, OkCupid should be allowing users to opt in or out of experimentation where their emotional well-being and the results of a service are in question. Revealing the methods of that experimentation do not need to be made apparent, but bringing that trial to the forefront of the user’s mind will help alleviate the outrage that stems from lowly, lonely, dateless peons finding out they are simply playthings for Silicon Valley execs to do with what they please.
For those of us that heavily partake in the copywriting and copyediting processes of our businesses and spend a good portion of the day making sure your colleagues do not send out emails accidentally swapping the use of “our” for “are”, this story may simultaneously amuse and enrage you.
A Provo, UT man was fired from an English language learning center for writing blog posts expounding on common homophones, not to be confused with homophobes. Notice the subtle difference in the N vs. B there? Homophone is the proper name assigned to words that are commonly mistaken for one another because they are pronounced similarly such as two, to, and too. Tim Torkildson, the social media specialist in question, writes content for the Nomen Global Language Centre as a way to continue teaching English to non-native speakers.
Nomen owner Clarke Woodger defended the decision to let Mr. Torkildson go by implying that the difference between “homophone” and “homophobe” can be confusing to someone just learning English as a second language, thus making the point for the necessity of homophone-centric content on the website to begin with.
Given the comicality of the termination reasoning and the popularity of this story, I anticipate that Mr. Torkildson will not find it difficult to gain more rewarding employment elsewhere. The lesson to be learned here? Stay in school, kids.
What Really Happened With Healthcare.gov
For those of us that make websites from start to finish, the thought of taking on a massive government website build backed by months of heavy press sounds like the absolute last thing you would want to do. A recent investigation by a non-partisan congressional agency known as the Government Accountability Office (GOA) reveals that the creation and roll-out of the Healthcare.gov website was exactly as mangled and mismanaged as you would assume.
If watching politics for the last, oh I don’t know… 60 years or so has taught us anything, its that politicians make lousy project managers. A lack of proper planning and oversight prior to the start of the project and with too many stakeholders making and changing directives during the process caused the massive down-times and site errors that plagued Healthcare.gov during the early days of its release.
That lack of planning and expectation management is what causes projects to go grossly over time and budget, as the healthcare exchange website did - though on a much larger scale than we typically see for most website launches. To address these issues the White House brought in a management consultant (aka project manager) to clean up the mess and eventually was instrumental in the successful enrollment of over 8 million citizens to the healthcare exchange. To read more in-depth on how the healthcare.gov website came to fruition go here.