Facebook and the Power of Algorithms
According to an article written by Professor Karrie Karahalios, from the University Illinois, and published in the MIT Technology Review October 2014 Issue, the power of algorithms is becoming a, mostly unknown, breeding ground for “corporate psychology research” on consumer behavior. In the article, Professor Karahalios identifies the Facebook study on “emotion contagion”, which documents and identifies the contagious effects of newsfeeds, pitting negative vs. positive feeds posted on unknowing Facebook users pages, to identify their reaction.
The result? More negative feeds, more negative posts. More positive feeds, more positive posts. Professor Karahalios then conducted a small study to determine what people knew about how algorithms affect their opinions online – only this time, the participants knew they were being monitored and manipulated on their feeds. The result of this study? One person “quit” Facebook all together after the project. The other 39 participants modeled their own feeds, with more control, and a better result.
Algorithms, interactions online and across mobile devices are now basically ubiquitous across most societies. The surprise is not that the algorithms affect the way people interact, nor even that Facebook would use this method to “study” users actions. After all, Google, Yahoo, eBay, Amazon and iTunes have all done this for years. There are even algorithms we really love, like Amazons “other recommended books for you” based on our behaviour of reading interests, or Netflix’s suggestions, or iTunes “Genius” playlist creation that saves us the time of creating a playlist, based on the mood of the music we are listening to. But then this kind of data mining can have dangers too, like when Facebook got a ton of bad press after their “Year in Review” algorithm ended up highlighting the worst events of some people lives, and replaying them alongside a snappy tune.
Of course, it can be argued that algorithms, that are more stealth, surreptitiously invading our most inner psychology of surfing, searching and posting, are all “bad” and “evil” plots, but, like anything it is more a matter of the grey area. We must ask the questions like “what if” and “how” and “why” with this evolving part of our social infrastructure that tracks, collects, recommends and influences us, most of us, unknowingly. Like anything in life, there’s both a good and a bad side to the level of detailed information algorithms make possible.
How to use Algorithms Positively
What if, in enterprise, we make our businesses smarter with algorithms? What if, those practices that prove to be the best practices are tracked, in the daily work we do, whether on a production line or, like most businesses in the western world today, and increasingly globally, with intelligent service tasks?
What if, algorithms were linked, through enterprise Apps to not only the information we deal with, but the processes we follow after having information, better information, more clear information, about how the things we do daily are done best? Call it big data analytics, meets process dynamics. Call it micro data of our HITS “Human Intelligence Tasks”, meeting the process we followed to accomplish the best result in the least amount of time, and/or, with the best outcome.
The argument can be made that “emotion contagion” which Facebook seem interested in, when coupled with actual functions of workers in their roles, can capture a better set of “best practices”. Perhaps this will help save staff’s time, or create better engagement with customers on a broad scale. Perhaps making our businesses better, more emotionally positive places to work, with increased productivity, could result in a win-win scenario of teams of people and the organizations they work for, getting an improved careers, bottom lines, lifestyles, while customers could have improved interactions with products and services.
The Genius Process List
Call it the “Genius Process List”. How do we create Genius Process Lists, smart “to-do” lists? With what we call smart tags, process flow chains, where we start with a study of what our business does, at each role level, and then extend that to mobile, to desk tops, to the devices people already use daily to get their work done. It doesn’t have to be “hidden” either. Like Professor Karahalios research shows, it’s better if it is not. When people are more aware about what they are doing, they improve the results, improve their “genius”. This phenomenon, also known as the Hawthorne Effect, was demonstrated way before the internet age, in a study done at the Hawthorne Electricity plant in 1924-1932. People work harder at becoming smarter and thinking more about how they do things, when they know they are being observed.
The challenge isn’t with algorithms, or how to use them. The challenge is with the old style of management, the things that businesses do in the “hierarchical mindset” of management who love “control”. It is well known, now, that the more progressive organizations, who let go of “control” and let people become more natural and passionate about what they do, are the most successful organizations in the world. Daniel Pink has spoken on TED talks about this reality, what he calls the “Puzzle of Motivation“.
Organizations, Enterprises, can effectively engage process flows, algorithmic “smart lists” tagged To-Dos to help capture, direct and engage participants from their workers, if they begin to engage Enterprise level tools and let their staffs participate at the level where the HITS “Human Intelligence Tasks”, the work “actually” gets done.
The problem is with algorithms. The challenge lies with the way that management traditionally “thinks” about BI (business intelligence) analytics, and so they (management) can use BI to figure it out.
Shocker for managers, and executives. Your teams and people are smart, and getting smarter. They do the work, so let them participate. When Harley Davidson did this a few decades ago, talked to the mechanics and the workers on the “floor” – and learned how they would manage the failing bike manufactures business, it re-invented the company into a brilliant global resurgence and brand that dominated the new “bike” culture – offering some of the most expensive and sought after motorcycles in the world. And this was learned well before algorithms could be accessed to help any business get smarter.
Getting smart in our businesses is in and of itself, a process. Business is a process. A decade ago when I wrote about Business in Process Enterprise Design (BIPED) and developed the process flow algorithms that drive our Patents-Pending which drive the 1to1Real platform, I was thinking how business managers could learn more from their staffs than any Accenture type system design of the day. Today, a decade later, we are seeing the reality of algorithmic intelligence that is not going to go away. The only questions now are, how and who will become the Harley Davidson businesses, the smart ones, to grow with intelligence from the Way things actually get done the best.
Who knows, maybe even management and their techniques, will become “smarter” along with more productive teams, and better actual intelligence about what those people who make your business tick do everyday. All it takes is the right tool, available on the right devices, and a start in implementing process flows that tell about the “real” way things are getting done, and how they get done best.