Facebook, first— the rest are on the way.
A couple of years ago, bots were all the rage. In the spring of 2016 at F8, Mark Zuckerberg went into some detail of the promise of the strategy of developing Messenger into a platform— and the promise of bots.
Today, we at SBCPH know that the remarks were understated. Bots are stealthily revolutionizing the way that things are done at our offices that are engaged across an incongruent set of industries and markets: from human capital management, to healthcare, to financial services.
But as those of us in international markets know, Facebook has failed to penetrate and replicate success that they’ve realized in Western markets, in countries such as Korea, China, Japan, and elsewhere— where Messenger and the Facebook platforms have failed to come close to posing a threat to the top social networking, social media, and messenger platforms in those respective countries / markets.
Besides, Facebook and the markets that it domiciles in and toes in (social networking, social media, video, messenger, etc.) isn’t the only game in town; Twitter is finding its way— maturing as one inadvertently does under the pressures of operating as a publicly traded company (shareholder pressure; public opinion; regulatory scrutiny); Snapchat is another paradigm and approach; LinkedIn is beginning to flex in markets and spaces where there used to be a “gentleman’s code” to keep a distance from Facebook.
So Facebook is first to the bot game, and it’s where our offices are seeing more traction of the platforms available thus far, but we’re prepared for other platforms to follow suit— and soon (perhaps a platform that isn’t only more accommodating of bots, but built to suit the bot paradigm).
What are bots?
Which leads us to the topic of what a bot actually is? Is it AI?
It is not AI. Not today, anyway. A bot as it exists today on platforms such as Messenger [btw, know the distinction between the Facebook platform and Facebook’s Messenger platform—it’s important] is simply a series of programmed content-responses. If you program your content-responses well enough, the UX comes off as AI-esque, but it isn’t AI. There isn’t intelligence in the capacity of decision making or evaluating that goes on with bots— yet.
That bots aren’t AI per se should notin any way discredit or take away from the utilitarian value of mainstream bots. As we’ve mentioned, we’ve deployed several bots in healthcare and medical scenarios, HR and training scenarios— all scenarios that go well beyond typical superficial marketing, brand building, and other “noisy endeavors.”
In 2013 at SBCPH, as a wide and horizontal organization, we subscribed to and encouraged a culture of technical content creation from the top, down. For a long time, our attorneys and managers subsisted on a short staff of technical writers, content creators, as well as clerical and administrative staff to produce and maintain their “paper infrastructure.” In 2013, we encouraged a weaning off of the dependence of these teams, placing the minds at the heart of all of this paper infrastructure at the helm.
There was some cultural resistance, but overall the growing pains that we incurred and the investments put forth proved worthwhile and very, very profitable.
Attorneys, managers, and c-level executives took up CMS courses, learning how to administer their own intranet sites built atop SharePoint and WordPress. These higher-level authors of all of this paper infrastructure took up power user courses in Google Apps and Microsoft Office 365. There was a strong correlation between offices and teams that adopted and adapted the “new tech” (new to them) to productivity and efficiency quotients as well as business generation and overall business results.
What we saw then, was that the tech of these CMS and doc management systems, such as G Suite (formerly Google Apps) were so mature and user friendly that no engineering or technician experience was required to become a power user of these systems; we came upon a mind meld — an equilibrium point in which tech has matured enough and mainstream users had matured enough; in this particular case, CMS and doc management systems had become just easy enough to use that folks such as attorneys, physicians, and other authors of our paper infrastructure of a non-technical, non-IT background could become power users in, in a matter of weeks.
Mind meld— that’s what’s happening with bots. Again, we’re finding that authors and contributors to our paper infrastructure (authors of HR forms for example; authors of policy memos— decision makers and authorities), are taking to bot creation platforms just as easily as they had to Gmail and SharePoint not too long ago.
Last year, our first bots were simply reincarnating and iterating on old HR manuals. HR teams and human capital managers set up permissions based Facebook pages and manually added select staff and team members to those pages; picking staff and team members who in turn were managers— allowing for these first bot creators to manage whole swaths of teams, departments, and offices by managing bots. A policy memo would be drafted, referencing a page in an HR manual, and copies of these pages would be uploaded for the bots to reference in natural language exchanges.
What our bots are up to.
Today, our bots are “talking” to patients in healthcare. View one of those bots here: https://m.me/pediahealthmedicalgroup. Some may view these activities as simply enhanced email or automated chat, but really, in its current form, it’s more than that. And the nature of the bot, versus the nature of an email or a manual response chat interaction, imposes the difference upon you as you engage with the bot.
With an email, the interaction with it can fall flat for so many reasons.
For one, inboxes are inundated. We all know the ailments: junk mail; too many chain mail messages from baby boomers; too many automated notification emails from services like Facebook, our banks, etc.
Two: with emails, you’re given one big slab of meat. The first few bites might be enjoyable, but by the time you get to the end of it, the thing’s a cold carcass and you find yourself drowning the stuff in sauce— trying to make it edible / quaffable.
With bots, you’re given morsels at a time, and at each break in a paragraph or break in train of thought (stop to catch a breath), you’re given options on what path to take in the content (whether you want to skip ahead, go off on a tangent, exit entirely, or view relevant resources). And at each serving, the morsel is hot, freshly prepared, and so, all that more engaging.
It’s freshly prepared because bot creators maintain content for bots in modules. They don’t have to rewrite whole tree diagrams of conversation that branch out to hundreds of generations every time they want to tweak something. All they have to do is hunt down the module that they want to update for relevance —for example, any mention of Snapchat— and change that one part —change mentions of the name Snapchat as a company to Snap. [Toss on some links to vids of Snap IPOing and you’re fresh again.]
There are a number of approaches that you can take to maintaining your bots. Right now, as early as it is and as quickly and rapidly as things are evolving in this space, we’re of the mind that we don’t flesh out overly complex content. We want to keep content close to the surface, and we want to retire content that ages quickly— again, simply because we’re at such an early phase wherein things are evolving much too rapidly to commit to and invest in a deep approach to content that is exclusive to bots and messenger platforms.