Month: March 2017

AI and Machine Learning in Health IT Marketing, PR, and Social Media

“Many tasks commonly performed by marketers are being augmented by machine learning, deep learning and cognitive computing” (Introducing the Marketing Artificial Intelligence Institute)

There’s a great fit between this week’s #HITSM (AI and Machine Learning in Healthcare, hosted by and ) and next week’s #HITMC (Make the Most of the Healthcare IT Marketing and PR Conference) tweet chats! If I know and , there will be lots of discussion about this intersection at the HITMC conference itself (May 5-7). Therefore I combined topics and adapted HITSM topics to apply to marketing health IT.

One of my degrees is a masters in Intelligence Systems (half AI and half cognitive science). I’m old enough to have lived through a previous AI hype cycle and then AI winter due to disappointment with results compared to expectations (but also see AI Winter Isn’t Coming). Since I spend so much time marketing health workflow IT I’ve naturally thought about how AI, ML, HIT marketing, and workflow all fit together.

Of the top of my head (helped by a quick Google search, and especially 15 Examples of Artificial Intelligence in Marketing) here’s a non-exhaustive list of ways AI and ML can help (HIT) marketing.

  • Recommendations and content curation (Netflix already does this, marketing automation systems are starting to do so, imagine using similar to chose different health IT marketing news letters to different subscribers)
  • Intelligent search engines (Just think what search engines that actually understand what they read on the web will do for both SEO and health IT conference web content. For example, I’d like to search all and only HIMSS18 exhibitors for workflow related content, and to have that search engine understand web site content regardless of whether the string “workflow” actually appears on a webpage)
  • Smart social sentiment analysis (understand not just what folks are saying about your brand, but also how they feel…, this is getting awfully close to journey mapping…)
  • Product pricing (Price is one of the classic Ps of the hoary old, what if you could truly optimize your prices relative to profit … and in realtime! )
  • Predictive customer service (imagine if you could predict complaints on social media and provision necessary resources to instantly solve a customer’s problem, and just think of their reactions on social media)
  • Ad targeting (“machine learning helps to increase the likelihood a user will click”, no brainer!)
  • Chatbots and conversational UIs (guess what, your brand is going to become a personality that can actually talk to customers and potential customers, kind of like a virtually intelligent Flo from the Progressive insurance commercials! )

    • “Have a question about a particular insurance package? Flo can answer your question. Want to get price comparisons between Progressive and Geico? Again, Flo can help, with the same sassy personality that the consumer has come to expect when thinking about the Progressive brand. This will extend throughout the entire customer journey, starting from the top of the funnel with marketing campaigns (TV commercials, digital advertising), continuing to the sale of product (ecommerce checkout, reducing shopping cart abandonment), and maintaining the customer relationship through support (automating call center inquiries), which ultimately comes full circle and leads into the next customer journey cycle.” (The 200 billion dollar chatbot disruption: part two)

  • Customer segmentation (obvious relevance to patient risk segmentation, here artificial intelligence, machine learning, and marketing combine to aid population health management!)
  • Content generation/marketing (yes, computers now write news articles and blog posts… therefore computers will, if not already do, write health IT news and blog post marketing content).

Here are six observations about AI and the future of marketing:

  • “2017 will be the year of the bot” (bots won’t replace websites, they’ll power them)
  • “The “marketing conversation” will become a human-machine conversation” (see above comments about conversational UIs)
  • “AI will accelerate marketing and sales” (“autonomous, self-driving, marketing automation”, BTW IMO workflow tech will play big-big role here!)
  • “Marketers will not be replaced by AI and will be able to skip the boring stuff” (Whew! That’s a relief! ““Anything that seems rote or mechanical … it’s all going to go to AI”)
  • “Algorithm development will become a commodity and data will become the key differentiator” (you can buy ML algorithms “off the shelf”, the difference will be the data fed to them)
  • “The Link Graph is going to be replaced by the Engagement Graph” (“quality of content is determined by the number of people listening, interacting, getting engaged”)

What do all of the above have in common? You knew I’d get to it… Workflow! Or, more specifically, process-aware workflow technology, such as data and machine learning pipelines, modeling conversation as workflow, and orchestrating of cloud-based services. Please see my three-part HIMSS17 series on what I call the New Workflow Technologies.

OK! That’s a lot of think about. Let’s the early morning of an AI/ML savvy and empowered HIT marketing and PR professional….

6:00 AM I glance at hundreds of email subject headers that are just the tip of the iceberg of thousands of possibly relevant news, brand, and social media alerts. I say…

“OK Hitmike🤖 (get it? HIT-M(i)C(ke)!) Please summarize this morning’s incoming health IT and workflow alerts”

GOOD MORNING CHUCK. ONE MOMENT PLEASE. YOU HAVE 786 UNREAD ALERTS. 77 SPECIFICALLY MENTION “WORKFLOW” IN HEALTHCARE. 50 MORE MENTION WORDS TYPICALLY USED IN DISCUSSIONS OF WORKFLOW, SUCH AS PROCESS, ORCHESTRATION, OR TASK (BOTH SINGULAR AND PLURAL).

(Hitmike🤖’s holographic UI flickers into existence above my bed, and two hands gesture the scare quotes around “workflow”)

Thank you Hitmike🤖. Do any alerts involve past or current clients?

YES, 7 PAST AND 3 CURRENT CLIENTS.

Have I had any Twitter interactions with any of the others?

93 PERCENT ARE ON TWITTER. 52 PERCENT FOLLOW YOU. YOU FOLLOW 75 PERCENT. YOU HAVE HAD DM CONVERSATIONS WITH 20 PERCENT.

Hitmike🤖, what is the total number of alerts involving past or current clients, plus Twitter follows and DM conversations.

32 ALERTS

Hitmike🤖, please convert all 32 alerts to tweets and upload to Hootsuite and assign them draft status. Make sure you include Twitter handles. Schedule for times that optimize impressions and replies.

DONE

Please remind me to review and schedule after breakfast but before 9AM.

OK

Hitmike🤖, today’s #HITSM tweet chat scheduled today during my dentist appointment. The subject is e-prescribing. Find all tweets about that and related subjects on my account.

DONE

Combine those tweets with #HITSM topics and create a one-time, one-use chatbot.

DONE

Configure to attend the #HITSM and use the chatbot.

DONE.

Immediately after the first #HITMC tweet asking who is participating, tweet the following on the #HITMC hashtag. UNFORTUNATELY CANNOT PARTICIPATE IN THE #HITSM TODAY. CHUCK SENDS HIS REGARDS. LUCKILY I AM AVAILABLE.

THIS WILL BE DONE

Oh, have favorite and retweet every tweet mentioning “workflow” or “workflows” during the tweetchat.

THIS WILL BE DONE

Hitmike🤖, any interesting health IT marketing automation news today?

YES. ONE PRESS RELEASE ONE BLOG POST.

Summarize press release.

THE RECENTLY HELD HEALTH IT MARKETING AND PR CONFERENCE ANNOUNCED RECORD INTEREST IN MACHINE LEARNING AND ARTIFICIAL INTELLIGENCE DRIVEN MARKETING AUTOMATION PLATFORMS.

Summarize blog post.

A BLOG POST HIGHLIGHTING THE PRESS RELEASE APPEARED YESTERDAY ON THE HEALTH IT MARKETING AND PR COMMUNITY WEBSITE.

Have any tweets from any accounts I follow mention either the press release or the blog post?

YES,

Please auto schedule a comment retweet contain the word “Cool” exclamation mark, an applause emoji, and relevant hashtags.

DONE

Thank you, Hitmike🤖! I am working on a project all day today researching the latest trends regarding use of workflow technology in marketing automation platforms. Please turn off all notifications except those potentially relevant to this project.

UNDERSTOOD

Here are the HITSM topics adapted to apply to healthcare IT marketing:

T1: What are the most promising applications of #machinelearning based #AI in health IT marketing today? #HITsm

T2: What challenges do health IT marketing organizations face in implementing #machinelearning technologies? #HITsm

T3: What political, cultural, or other factors drive adoption of #machinelearning or other #AI technologies in health IT marketing? #HITsm

T4: For orgs that have implemented, what’s the impact been (pos or neg) on health IT marketing lead generation and sales? #HITsm

T5: With so many options, what are some do’s & don’ts for health IT marketing orgs looking for a #machinelearning #AI partner? #HITsm

Bonus: Some say #CognitiveComputing is a buzz phrase & synonymous with #machinelearning or #MachineIntelligence Thoughts? #HITsm

I will see you there!


On Periscope!

Cognitive Technology for Goal-Driven Healthy Habits: An Intelligent Systems Approach

I am truly delighted to do my research in preparation for today’s #HITsm tweetchat with . I have such fond memories of her marathon Blabs (3, 4, 5 hours?). The topic of today’s chat came up frequently: How Technology Helps and Hurts Healthy Behavior Change.

I usually introduce myself an industrial engineer who went to medical school, hence my interest in healthcare workflow and workflow technology. I don’t often mention I when to medical school and then studied Cognitive Science, which was one half of my MS in Intelligent Systems (the other half being Artificial Intelligence). The CogSci portion included psychology, linguistics, philosophy, and neuroscience. As a graduate student, playing my way, I worked on computer models of aphasia, dementia, and depression. I even spent a week studying with the man who founded of cognitive therapy, Aaron Beck, in Philadelphia.

I took a look at using tech to change patient behavior from the point of view of cognitive scientist. In doing so I hit on this paper (full text freely available!), Habits, Action sequences, and Reinforcement Learning. It summarizes and synthesizes a number of topics I studied decades ago. Topics I feel are relevant to using tech to move human behavior away from unconstructive toward constructive.

Believe it or not, (and I suspect you’ll believe it, given my workflow-centric reputation) there is a workflow angle. A workflow is a sequence of actions, consuming resources, achieving goals. Humans evolved from more basic animals. These animals exhibit, what may be thought of as, instinctive workflows. A fixed action pattern is species-specific characteristic sequence of behaviors (actions), which, once triggered, runs to completion. For example, if an egg is displaced from a nest, certain geese will roll the egg back into the nest, even if the egg somehow magically disappears. They continue to maneuver the imaginary egg back into the nest. The animal kingdom is rife with FAPs. We even know a lot about the neural networks that generate FAP behavior.

What do FAPs have to do with human behavior? Well, FAPs are a lot like habits, a sequence of behaviors, automatically executed, in the presence of some “releaser”. They happen automatically, seemingly without purposeful or mindful control. Of course, unlike FAPs, human habits are not instinctive. Through a variety of techniques, we can break old habits and create new ones. However, doing so is difficult! This is where technology comes it.

However, before we get to how technology might be useful in this respect, it’s useful to have a model of what is going on inside our heads. The degree program I mentioned, Intelligent Systems, viewed robots, software artificial intelligences, humans, and even some animals, as “intelligent systems” that, to vary degrees, shared certain properties and characteristics, including perception, memory, action, reasoning, and learning. Further, intelligent systems research combined techniques from cognitive science (psychology, linguistics, neuroscience, philosophy) with artificial intelligence and machine learning, to actually create computer simulations of these intelligent agents, to better understand them. We’d create software simulating them, and then we’d conduct experiments, comparing their behavior in response to manipulated environment stimuli, and to intelligent agents in the real world. Sometimes we’d even “break” the intelligent agents, to try to simulate mental and neurological disease. As I mentioned previously, I worked on a variety of such projects, from aphasia (language difficulties), dementia (memory, reasoning, personality), and depression (where I actually published a number of papers!).

All of that, and it is a lot of personal history, is backdrop for what I will do next, which is describe our human mind as if it is computer simulatable intelligent systems, with an eye toward thinking about changing bad habits into good habits.

The Habits, Action Sequences, and Reinforcement Learning paper describes an intelligent system in which there are two complementary but also competing information processing modules. One module is “closed-loop” meaning it has a model of the world and in that model of the world it behaves (acts on its world) to move the world toward a preferred goal state. The perceiving-reasoning-acting loop is closed in the sense that the difference between the current world state and the preferred goal state is continually fed back to the intelligent system so it can continually chose actions that will eventually achieve it’s goals.

Contrast above with the second behavior module. This module is similar to a Fixed Action Pattern. It has a set of “hardcoded” workflows, sequences of behavior, which, once triggered, execute from beginning to end, without reference to whether they move the world from a bad (less preferred) state to a good (more preferred) state. The great thing about these automated personal workflows is they are fast, consistent, and require no thought. The bad thing about these automated personal workflows is that they are fast, consistent, and require no thought. If you change the environment, “good” habits can become “bad” habits.

The two systems can profitably work together. Once one’s environment changes, fall back on the closed loop thoughtful goal-oriented behavior. Over time find new personal workflows that work, then turn them into open loop fast, consistent, and “thoughtless” workflows. This frees up the closed loop goal orient system to focus on other, higher level, more strategic issues. Also, you can think of an intelligent agent has having different bundles of related workflows for different environments. As it move through these environments, different clumps of workflow potential become active. Let’s suppose an intelligent agent has about a dozen different environments it frequently or occasionally needs to navigate. Eight or nine may be stable and the current open loop personal workflows are perfectly appropriate. However, several environments may be problematic. So our closed loop problem solving systems focus there. Over time, as all of our different occasionally frequented environment change, each is dealt with in turn, converted from open loop to closed loop and back to open loop personal workflows. But imagine if all your environments change at once! That is indeed stressful and even your wonderful dual system, open and closed system partnership, can be overwhelmed!

On a moment-by-moment basis, current thinking is that these two, open loop and closed loop, modules compete with each other. Consider the following quote:

“some have suggested that these processes may compete for access to the motor system…. in which the goal-directed and the habitual systems work in parallel at the same level, and utilize a third mechanism, called an arbitration mechanism, to decide whether the next action will be controlled by the goal-directed or the habit process”

So, now let’s think about how technology might be used to help these two, open loop and closed loop, systems work together.

Let’s consider the open loop personal workflow system. How might we extinguish is highly automated responses, in preparation for instituting new, healthier responses?

  1. Prevent the workflows from being triggered in the first place.
  2. Detect when the workflows are executing and disrupt them.
  3. Emphasize the negative consequences of these workflows running to completion.

This last device is interesting because it is essentially attempting to convert open loop behavior into closed loop behavior.

I can imagine technology being used in all three ways.

  1. Don’t go there! (You know what always happens if you do…)
  2. Look! Squirrel!
  3. Ouch. Be honest with yourself. That hurt! (But also be constructive, give yourself a brief scold, and lay plans to avoid triggering similar future behaviors, or at least figure how to stop one if it get started.)

At the same time we are trying to hobble destructive open-loop personal “workflows,” we need to enable constructive closed-loop personal workflows.

  1. Make the future preferred world goal state particularly vivid.
  2. Figuring our how to solve new problems, or old problems in new ways, is hard. Provide help.
  3. Once you find a tentative solution, capture it! Institutionalize it in some way, to make it more the more likely to execute open loop behavior than the old destructive open look behavior.

Regarding the arbitration mechanism, both the open loop and closed loop personal workflow system spring into operation, race along in parallel, and then demand that they be given control. In this last regard, a basic insight is this. One way to become more “meta-cognitive” is to have some sort of model of yourself. This model can be used to explain and understand, and to guide what to do. I think this model of you and intelligent system in eminently teachable and learnable. In fact, cognitive theory works a bit like this. One of its goals is get you to think like a “personal scientist”. Scientific thinking involves weighing evidence and conducting experiments. Simply viewing yourself as a “scientist” is itself esteem elevating. I think something similar might be true of viewing yourself as an intelligent system.

Anyway, back to what technologies could be useful.

The stimuli that trigger personal workflow are often spatially and temporarily circumscribed and specific. Here wearables and the Internet of Things can be the eyes and ears of a system to detect you may be heading into a bad workflow stimulus rich environment. If a bad workflow can’t be avoided, and starts to execute, workflow execution itself can be detected. (This is currently an active area of artificial intelligence and machine learning research, recognizing which goals, plans, and workflow of an intelligent system are currently active.) Once the bad workflow is detected, mid-execution, send notifications, call someone to call you, ring the fire alarm, whatever it takes (no, don’t ring the fire alarm unless there is a fire, but you know what I mean!)

And if, heaven forbid, that bad-bad-bad personal workflow can’t be prevented… document it. And do so in such a way that the next time it can be held up and waved in front of the intelligent agent… NO, YOU REALLY DON’T WANT *THAT* TO HAPPEN AGAIN, DO YOU?

Relative to closed-loop problem solving and workflow creation, preferred workflow goal states might be vividly representing using virtual or augmented reality. (THIS is what you’ll look like in that bathing suit, this is what will feel like when you walk across that graduation stage!)

Relative to helping to find new workflows that work, that’s what many workflow and task management systems do. They help manage potentially useful tasks, to string them into candidate workflows, and then, when executed, keep track of state (success, in progress, timed out, failed, escalated, etc.)

Finally, one you find new workflows that work, you need to move that insight and actionability down into the system that senses whether you are in danger of executing one of the bad-bad-bad workflows, and offers a different, more constructive workflow instead. Increasingly, every single digital device we interact with is aware of each other and work together. They will talk to your fridge and your minibar. They will, if necessary, act on your behalf, perhaps even stepping to literally prevent you from doing what you are about to do.

Yeah, scary. But also, possibly, fascinating, in positive and constructive sense.

A lot of the technologies I just listed already exist in bits and pieces. Some are already being woven together, to act in purposeful and useful manner, at our behest, to help break and make personal workflow habits. In a sense, there will be (at least) two intelligent systems: you and the system you create around you.


On Periscope!

Health IT Marketing and Boomer, Gen X, and Millennial Workflows

I gotta brag. I’m not an anthropologist, but I love reading about anthropology, and lots of my friends in medical school were medical anthropologists. I use ideas from anthropology to inform my systems engineering thinking about health IT. With some trepidation, I finally waded into the anthropology of health IT, and wrote a blog post. Eventually I sent it to the former HHS CTO, who happens to have an anthropology degree. And she tweeted this!

That’s my brag! Made my week, it did!

Since that Eat Your Beans post, I’ve been thinking a lot about digital and computational anthropology … wait! Don’t run away! This really is connected to health IT marketing and generational divides. First, just a little background. Anthropology is the study of human societies and cultures and their development. The theory of generations (generations X, Y, Z, did you know that “Boomers” are Gen I?) was proposed in 1923 by Karl Mannheim, a sociologist. Sociologists and anthropologists both study societies and how humans behave. Sociologists focus on the big picture, such as social groups and society. Anthropologists drill down into the nitty-gritty of individual behavior in groups. If you put the word “computational” in front of something, it usually means the nitty-gritty is so nitty-gritty that there is enough detail to actually simulate behaviors. Closely related to computational anthropology is digital anthropology, in which anthropologists study interactions between humans and digital technologies.

Aha! Digital anthropology is clearly relevant to how “generational perspectives are influencing healthcare technology, and additionally, how can we (as health IT leaders) can strive to incorporate and include diverse generational needs into the industry roadmap” (from this week’s #HITsm chat). But computational anthropology is also relevant to “including every generation in our health information technology thinking”.

The rise of computational anthropology is fueled by so-called “big data.” Digital technology is so woven into our professional and persona lives, that its “data exhaust” (love that term!) can be used to track us, understand and empower us, but also raises enormous ethical and privacy issues. The fascinating thing in this latter regard is that the field of anthropology has adopted a sophisticated system of ethics regarding dealing with data about human behavior. In some ways, it surpasses current health IT principles for handing sensitive personally identifiable health data. For example, consider the following from the Wikipedia entry about digital anthropology…

“Online fieldwork offers new ethical challenges. According to the AAA’s ethics guidelines, anthropologists researching a community must make sure that all members of that community know they are being studied and have access to data the anthropologist produces. However, many online communities’ interactions are publicly available for anyone to read, and may be preserved online for years. Digital anthropologists debate the extent to which “lurking” in online communities and sifting through public archives is ethical.”

Lurking during a tweetchat potentially being unethical? Wow!

But let’s assume, for the moment, that anthropological ethics and Internet-Of-Things cybersecurity issues can be adequately managed.

How might ideas from digital and computational anthropology potentially guide a health IT marketer?

The first thing to realize is that digital anthropology is applied anthropology, from which marketing research increasingly incorporates methods. In fact, there is a Journal of Business Anthropology (a sub-discipline within applied anthropology) and the sub-discipline of marketing anthropology. Anthropology is an increasingly popular minor among marketing students. Degrees in digital marketing anthropology are surely just around the corner.

What about workflow? Digital anthropology can be used to collect and interpret consumer and patient life-flows (essentially “workflows”, but more general than mere work settings, including family and other personal activities). Computational anthropology provides representations and models into which these data and interpretations can flow and inform. At the top of this list are agent-based simulations. Agent-based simulations are really cool. So cool, I recently attended the Anylogic user conference in Nashville to learn more about agent-based simulation. Anylogic develops and markets the most sophisticated agent-based simulation software on the market. Anylogic can also simulate more traditional discrete event simulations (popular among industrial engineers for simulating patient flows, where I got my start in healthcare workflow) and dynamic systems. Agent-based simulations simulate “agents”, which are basically simplified representations of humans, though I am sure they could simulate other kinds of agents, such as cattle behavior at the level of individual cows, and so forth.

Here are some of the various workflow notations compatible with AnyLogic.

With so much compute power available today, look at the scale of current agent-based simulation research! Surely human behavior after a nuclear attack is an important public health topic!

Here are a couple animations driven by agent-based simulation.

The following is a simulation of conference attendees interacting with the lunch queue. This not as impressive as either of the previous agent-based simulations, but there is the thing. It was created, from scratch, in just a couple hours in front of an audience. Each of the “attendees” (in the simulation, not the audience in which I sat) is essentially a tiny, virtual workflow system. Each attendee is modelled as a state machine, which is the formal terminology for a model of workflow being executed by a workflow engine while interacting with environment inputs.

Watch the above animation and just think of the possibilities for modeling different generations and their interactions with digital technologies! Increasingly we have the data. We have means to model workflow behaviors and execute workflow models. We can study personal and professional workflows executing within interactive environment. And we can do so within and between demographic generations within families, among friends, and between patients and healthcare systems.

Sound like science fiction? Workflow research really is finally moving out of the healthcare organizational setting and into patient’s lives. Check out this diagram of workflow interactions and information flows between a patient outside of a healthcare organization and the healthcare organization itself (from the recent Healthcare Systems Process Improvement Conference in Orlando).

Also see out my previous post, Actuarial Science, Accountable Care Organizations, and Workflow.

“Workflow⁰ is a series¹ of steps², consuming resources³, achieving goals⁴.”

⁰ process
¹ thru graph connecting process states (not necessarily deterministic)
² steps/tasks/activities/experiences/events/etc
³ costs
⁴ benefits

If one modifies my definition of workflow, though within my subscripted limits, to …

“Process is a series of events, consuming expected resources, achieving expected benefits.”

… you’ll arrive at a stochastic process closely resembling actuarial science’s generalized individual model (page 35 in Fundamental Concepts of Actuarial Science, a great review or introduction by the way!).

During my student days, we spent a lot of time estimating parameters and distributions, and then predicting behaviors of these stochastic processes. Sometimes we did so analytically with complicated equations (Markov Models). Sometimes we fell back on computer simulation (Monte Carlo).

A quick review of actuarial science literature indicates many of these same techniques are used today.

Back to the subject at hand…

Patient journeys are workflows. If they are workflows, then we can model them, inform and test those models with data from digital medical anthropology research, and then simulate those patients interacting with digital healthcare technology using ideas from computational anthropology.

If you Google generational differences, you’ll find hundreds of tables that look like this.

These generational difference tables compare and contrast live experiences, goals and values, resources and constraints, and typical behaviors of Baby Boomers, Gen Xers, Millennials and other generations. Adapt these insights to adopting and consuming digital health technology and information. Collect increasingly available data (subject to ethical constraints). Use data to inform and drive simulations of personal and professional life-flows and workflows. Compare simulations to what we observe in the real world. And then systematically improve these simulations.

In doing so we will gain greater insight into the differences and similarities between different generations regarding adopting and consuming digital health technology and information.

Consider this scenario, one I believe will be possible within five short years.

Consider a population health system covering five million members. Imagine 20,000 medical and administrative staff (by the way, I just pulled that number out of a hat). Further, imagine various pieces of the IT systems being proactive, that is, agent-like. Roughing in the models would start with a combination of generational differences and risk stratification. Patient states include well, acutely ill, chronically ill (and if so, which chronic conditions). Staff states include off-duty, on-duty, ideal, and busy (and if so, which patient-directed activities are they qualified for). Now imagine you are a health IT marketer. Instead of working for a health IT vendor or health IT oriented marketing and PR agency, you’ve made the transition to working for a health system. You’re job is to understand and facilitate the diffusion health IT technologies into the homes and hands of covered population health system members. Here are some additional states: unadopted, adopted-but-not-optimized, optimized. Now, based on a variety of data, from qualitative and quantitative applied digital anthropology research, estimate the probabilities of transitions between states. (Possible role for machine learning here!) Workflows are series of these state transitions, which can be simulated, to fit various other data sets and generate predictions. For example, which kinds of health IT technologies (apps, calls, chatbots…) introduced to who (Boomers, Gen X, Millennials…) influence transition probabilities between which states (well, acutely ill, chronically ill…), and probabilistic models of impact on population health system resources (number of personnel required, kinds of personnel), under different assumptions about which technology initiatives are undertaken (which kinds of patients are supplied with which kinds of health IT technologies). If you think this kind of simulation requires astounding amounts of data, it does. But we now live in the Big Data era. The data is there or potentially there. The real problems with this simulation are managing its complexity and data ethics issues. However, if researchers can undertake an agent-based simulation involving between 10 million and 20 million individuals in the aftermath of multiple Manhattan nuclear blasts, then agent-based simulations of health IT diffusion and effects on clinical outcomes and costs are surely at least almost already possible!

If I have stimulated your imagination and interest, check out my Health Standards article, Marketing Workflow Is An Incredible Opportunity To Differentiate Health IT Products, And You!, which ends this way:

“Workflow: It’s not just for industrial engineers anymore!”

I’ll see you at the Including Every Generation in our Health Information Technology Thinking #HITsm tweetchat! Noon EST today.

Further reading:


On Periscope!

What Is Business Process Management, or BPM? My Own Short Informal Description

A colleague asked me for a three or four sentence description of business process management. I managed to whittle something down to five sentences!

Business process management (BPM) is a “process optimization process” approach and technology. BPM includes discovering, modeling, and executing models of processes and workflows. During model execution, tasks are tracked to completion or escalated, resulting in fast, consistent, error-free task management. BPM software is exceptionally agile, because executable models of processes and workflows are so easily changed and optimized. Increasingly, many applications not traditionally categorized as BPM software, such as customer relationship management (CRM) and interface/integration engines, are embedding BPM-like functionality and behavior.

Digital Transformation of Healthcare with Business Process Management: Two Books To Consider!

[This post was written in preparation for today’s Digital Transformation in Healthcare #HITsm Tweetchat!]

I’m always looking for ways to get health IT and workflow technology folks together, in real life and on Twitter. For example “digital transformation” is a popular phrase and concept. It describes the changes due to digital technology in all aspects of human society. More specifically, it’s about transforming business activities and processes. Processes? Workflow! So, workflow technology, also called business process management, is in the digital transformation conceptual mix. In fact, there’s a wonderful book about BPM and digital transformation coming out the 20th of this month.

Digital Transformation with Business Process Management: BPM Transformation and Real-World Execution

I obtained an advanced peek. I highly recommend it!

From the foreword, Nathaniel Palmer:

“Today’s BPM platforms deliver the ability to manage work while dynamically adapting the steps of a process according to an awareness and understanding of content, data, and business events that unfold. This is the basis of intelligent automation, enabling data-driven processes adapting dynamically to the context of the work, delivering the efficiency of automation while leveraging rules and policies to steer the pathway towards the optimal outcome. For these reasons, BPM is the ideal platform for digital transformation.”

The introduction and case study abstracts are available online.

If you are intrigued with the idea of using workflow technology to transform healthcare organizations, I hope you’ll also consider Business Process Management in Healthcare, Second Edition.

I wrote the Foreword and contributed a chapter, the full text of which are available here (Foreword) and here (Marketing Intelligent BPM to Healthcare Intelligently).

And, since this is your digital transformation lucky day, here are some recent articles about BPM and digital transformation.

Transform your ideas about transforming healthcare with business process management!


On Periscope!

Why is Health IT behind in workflow-friendly technology and process awareness? How do we fix?

Republished from Health Standards.

Why is Health IT behind in workflow-friendly technology and process awareness? How do we fix?

MAY 11, 2015 BY CHARLES WEBSTER, MD 0 COMMENTS

Thank you to Health Standards for allowing me to press my case for using workflow technology to counter healthcare’s Workflow Problem. I’ll argue that we should adopt a new metaphor. Instead of “data silos,” let’s speak of “workflow silos.” Instead of waiting until we understand healthcare workflows to automate them, let’s use workflow technology to create and leverage understanding. Let’s promote a Workflow Triple Aim in service of healthcare’s Triple Aim. Let’s educate, highlight, and recruit the best workflow minds to improve care, outcomes, and costs.

From Data Silos To Workflow Silos

Metaphors are not just flowery language used by poets. The metaphors and analogies we use, user-friendly (treating computers as people), data silo (farming, nuclear war), and data liquidity (flowing water), powerfully influence how we think. That is the point of Metaphors We Live By, an influential book in cognitive science.

I propose we stop talking about ‘data silos’; start talking about ‘workflow silos”. Data and workflow are related concepts, but very different ways of looking at healthcare. In fact, almost everywhere you see ‘data’ (especially in a headline), just replace it with ‘workflow.’ You’ll be pleasantly surprised by the innovative ideas that just seem to begin to, well, flow!

The occasion of my plea to “Think different.” is my response to a special live #HITsm tweetchat that took place Tuesday, 14, at the HIMSS15 conference in Chicago. I wasn’t present, sadly so, since a question I submitted was asked of four health IT experts before an audience of HIMSS15 attendees. I do, however, have the livetweeted #HITsm transcript. I did respond, two days late (oh, I wish I’d been there in person!). Those tweets and my dozen tweeted responses are embedded in a blog post on my Healthcare Business Process Management Blog. As usual, I’d love your comments, here, my blog, or even to the embedded tweets themselves.

Let me focus on three of those tweets.

“T5 Why does the HIT industry lag behind in terms of supporting workflow-friendly technology & process awareness? How do we fix?”
“Missing interoperable workflows, but first need the data to bust down silos, establish what those are”
“We still need interoperable workflows between providers but we don’t yet know what those workflows should be”
Let’s talk about data silos first.

I grew up on a corn and soybean farm, so I happen to know lots (well, probably more than average) about grain silos and elevators.

Let’s consider a definition of “data silo.”

“A data silo is a repository of fixed data that an organization does not regularly use in its day-to-day operation. So-called siloed data cannot exchange content with other systems in the organization. The expressions ‘data silo’ and ‘siloed data’ arise from the inherent isolation of the information. The data in a silo remains sealed off from the rest of the organization, like grain in a farm silo is closed off from the outside elements.”

If one accepts this description of a silo, then one’s thinking tends to head down some paths, but not others. Of course, “busting” silos is more of a World War III metaphor than a farming metaphor. Keep in mind the result of busting a nuclear silo is not the freeing of its content, but its destruction. So back to the farming analogy, which seems more constructive for our purposes. What if we think about silos from the perspective of how they are *actually* used in agriculture?

After the recent HIMSS15 conference in Chicago, I continued to my family farm in NW Illinois. I tweeted pictures of farm houses, antique tractors, and grain silos. I asked a well-read farmer about what he thought of the data silo analogy. He said something profound (which I tweeted at the time), “Seeds are alive & in the spring they know it’s time 2 germinate, they swell & begin 2 sprout”. This dynamic process can actually start in the silo! Further discussion confirmed he did not think of seeds or grain as inert, at all. In fact, they are alive, dynamic, tightly coiled bundles of potential energy (a “spring” metaphor, if there ever was one, and borderline pun, to boot). By the way, a really great book about seeds is The Triumph of Seeds. Seeds are metaphorical supercomputing self-assembling, micro-robots with sensors and actuators. They put today’s Internet of Things to shame.

What’s inside seeds is much more like workflow than data. Data is static, inert, and tactical. Workflow is dynamic and strategic. Workflow acts on data: transporting and transforming.

What about a pile of seeds? Take it from me; big piles of seeds are not static or inert. They are shifting and treacherous. 31 people died in grain-bin entrapments last year. Two young men died in my home county in 2010. (My mom knows the families.) I never worked in a silo. I did sneak into a corn bin once, until I was chased out and given the lecture of lifetime. After that, the couple times I peeked into a grain silo, I saw riptides and treacherous currents, not a static pile of “data.”

What about grain elevators, those collections of interconnected silos one sees throughout our Midwest?

“Grain elevators play a key role in U.S. agriculture, and fulfill three main functions: post-harvest handling and storing of cereal grains and oilseeds, conditioning and preserving of grain, and facilitating the delivery of grain to domestic feeding and processing, as well as overseas, end-use destinations. These facilities have evolved from mere storage sites to large, high-throughput, highly automated, processing plants…. grain elevators represent a key intersection in our food production chain” (Design Considerations for the Construction and Operation of Grain Elevator Facilities. Part II: Process Engineering Considerations)

Consider that phrase,”evolved from mere storage sites to large, high-throughput, highly automated, processing plants”… again, sounds more like workflow, than data, to me.

Seeds, piles of seed, silos, and collections of silos, are in constant motion, channeled by agricultural organizations and technology. There’s a giant grain-to-food conveyor belt workflow from thresher to table.

I’d rest my case, but I have two more tweets go!

From Workflow Oppressed To Workflow Owners

3. “We still need interoperable workflows between providers but we don’t yet know what those workflows should be”

To which I tweeted back:

“IMO we should NOT wait 2 understand workflows B4 using workflow tech, cuz WF can be adjusted!”
“that’s POINT of workflow tech, since workflow is liquid, can implement WRONG WF but fix later”
If you think my ‘workflow silo’ analogy is a bit, well, ouut-theere, just wait until you see my next one! Stick with me!

Throughout the history and evolution of democracy and democratic traditions, autocratic regimes sometimes agree that democracy is good, but citizens need to be taught about democracy first. Only after citizens have matured, can they be trusted to actually vote. This is how revolutions happen. The citizenry see through the ruse, won’t wait, and deposes the despot. The rejoinder to these despots is that one learns by doing. Perhaps badly or imperfectly at first, but this is the only practical route to democratic civil society.

Something similar can be said of healthcare workflow. In fact, I’ve sometimes used the Twitter hashtag #OccupyHealthcareWorkflow. Healthcare, but more specifically, health IT, has a “workflow problem”: usability, interoperability, safety, patient experience, and more… What do all of these have in common? Workflow-oblivious technology (see my five-part series on healthcare workflow tech).

Users of health IT, especially clinical staff and patients, need to own their workflow. What could I possibly mean by “own workflow”? How can we make it possible that patients, physicians, the intended beneficiaries and users of these IT systems, should own their own workflows?

The key to solving the workflow problem, and repatriating healthcare workflows to their most important stakeholders, is workflow technology. In other industries, when you have a problem X, X technology arises to help solve or manage. Think pollution/pollution technology. Think, healthcare workflow/healthcare workflow technology.

Saying we can’t automate workflow because we don’t yet understand workflow makes sense if you hardcode workflow. That is, if you use third-generation languages such as Java, C-sharp, and Mumps, to automate the series of tasks that make up a workflow. Since you can’t easily or inexpensively change workflow after the code is written, then, By God, you better get the workflow right in the first place.

The problem with this stance is that there is no single correct workflow. Workflow changes all the time. Government regulations change. Patients change. Society changes. Science changes. Medicine changes. Everything is changing, all the time.

What is the alternative, then? Low-code software development.

Use workflow technology. Draw workflows in workflow editors. The results are both executable by workflow engines and understandable by non-programmers. Some of these systems look like traditional workflow diagrams, such as produced by Visio. If you don’t think non-programming users of these systems can understand workflow diagrams (many can, in my experience), then there are systems that present simplified, but still usefully editable, workflow.

What if patients and physicians don’t want to click or touch anything during design? BPM (Business Process Management) systems can be changed, even implemented, a magnitude faster than traditional health IT system. Analysts (business and clinical) can quickly iterate through a series of workflow designs, until converging on workflow satisfactory to patients, physicians, and staff. In either case, super-users creating super workflows, or healthcare organization analysts doing the same in close coordination with users — break the workflow monopoly that has been imposed on us by workflow-oblivious legacy health IT.

In fact, I’m seeing a convergence, between patient experience and user experience. The tech that will make this practical and scalable will be workflow technology. Some of the most sophisticated uses of workflow tech have been in customer- and consumer-facing “Systems Of Engagement”, those systems at the edge of the enterprise (to contrast with “Systems of Transaction”, mission critical processes deep within the enterprise).

At a recent BPM conference I attended, “empathic” workflow was a hot topic of conversation, which was aligning the customer journey of a daughter and her mother regarding a home emergency medical bracelet. Backend and customer-facing workflows and touch points must be redesigned, to respond to their customers’ journey, through experiences of worry and feeling overwhelmed, relieved, anxious and frustrated… It is the ability to change workflows and processes quickly, to quickly improve to workflows serving this mother and daughter, which make this critical alignment possible. Workflow technology will be an essential tool and platform for co-designing healthcare workflows.

Don’t wait to implement healthcare workflow technology until we completely understand healthcare workflow. In contrast to traditional health IT workflow-oblivious tech, it is the implementing of workflow tech that creates understanding of workflow. Instead of blue ribbon commissions and dusty academic research telling citizenry “correct” healthcare workflows, use workflow tech to quickly get to satisfying, shared, co-owned patient and provider workflow. Pave the cowpaths. Then straighten and widen into eight-lane super highways. Health IT that is easily molded to patient and physician workflows, which can then be systematically improved while respecting normal human tolerance for change, is the key to health IT adoption.

Of course, healthcare needs also needs interoperability to achieve this vision. However, focusing exclusively on message transport and translation has had untoward consequences. The lowest level of interoperability, syntactic interoperability, gets messages from machine to machine. The next level up, semantic interoperability, makes sure the content of these messages means the same on both machines.

But workflow interoperability (also task or pragmatic interoperability) isn’t just the end game. It is the begin game too. Do exchanged messages accomplish the goals they are intended to accomplish? Each level helps the other levels. In other words, workflow tech can greatly facilitate lower level message exchange. Modern workflow platforms support a plethora of adaptors and connectors for gluing disparate technologies together. In fact, healthcare interface and integration engines are an important area of diffusion of workflow engines, editors, and analytics into healthcare. A C-level health IT executive just told me he’s getting a BPM systems because it will work so will his regional HIE’s BPM-based infrastructure.

Communication among EHRs and other health IT systems must become more “conversational,” if they are to become more resistant to errorful interpretation. And workflow tech is the best and most natural means to enable these conversations.

Workflow Triple Aim In Service Of Healthcare’s Triple Aim

Finally, the “How do we fix?” tweet:

1. “T5 Why does the HIT industry lag behind in terms of supporting workflow-friendly technology & process awareness? How do we fix?”

There are many different stakeholders and many different skill portfolios, when it comes to the healthcare workflow problem.

You may have heard of the Triple Aim, to improve care, health, and cost. My means to contribute is a Healthcare Workflow Triple Aim, to educate, highlight, and recruit the best workflow minds.

Hence my tweeted response:

Need education
Hilite success stories
Recruit workflow tech
Educate users and buyers of EHRs and health IT about workflow and workflow tech. Find and highlight success stories to promulgate best practices. Bring into healthcare the modern social, mobile, analytics, cloud-based workflow technology called BPM, for Business Process Management. Just as health IT is a large and varied continent (hey, metaphor sighting!), BPM is too. There are different kinds of BPM. So this also means sorting through and adapting workflow technology to healthcare’s unique needs and purposes.

Every year I search every HIMSS conference exhibitor website for content about workflow, workflow technology, workflow editors, customizable workflow, workflow analytics, and business process management and related case management software. I do this to find informational and encouraging healthcare workflow tech stories to share via social media during each HIMSS conference. Five years ago, at HIMSS11, I didn’t even get to the two percent threshold. Every year since, the percentage has doubled — 2%, 4%, 8%, 16% — until this year’s HIMSS15 25 percent plus. The percentage may have actually doubled, but there was so much good and relevant workflow material I literally ran out of time (somewhere in the Qs). I’m seeing a surge of new and embedded workflow tech across almost every category of product and service. Five percent of HIMSS15 exhibitor websites actually mention “workflow engine” (the engine that executes workflow definitions).

As I put it in my HIMSS15 Social Media Ambassador blog post about thirty submitted posts about the Future of Connected Health (in which I highlighted, wait for it, healthcare workflow!)…

“The future is so bright, I gotta wear shades!”

Get ready for the bright sunshine of a healthcare workflow technology spring, and the sprouting of a million workflow blossoms. How’s that for flowery metaphor?

The following two tabs change content below.

My Twitter profile
Charles Webster, MD
Charles Webster, MD, MSIE, MSIS, evangelist for great health IT workflow, is the self-anointed King Of All Workflow In Healthcare (AKA The “Workflow Bear” per !)

Marketing Workflow Is An Incredible Opportunity To Differentiate Health IT Products, And You!

Republished from Health Standards.

Marketing Workflow Is An Incredible Opportunity To Differentiate Health IT Products, And You!

JANUARY 17, 2015 BY CHARLES WEBSTER, MD 0 COMMENTS

Editor’s Note: The following article is a Guest Column from Charles Webster, MD, a health IT workflow expert and advocate for process-aware technologies in healthcare, including workflow management systems, Business Process Management, and dynamic and adaptive case management. You can contact Dr. Webster on Twitter or through his blog at ChuckWebster.com. To learn more about submitting a Guest Column, Click Here.

Anyone who has wrestled with how to sell a health IT product has wrestled with features and functions versus benefits: what a product is and does versus what important problem it solves and how that will make someone feel. I’ll argue that workflow is the bridge from features to benefits. This bridge is missing in much health IT marketing today. With a modicum of self-study, any health IT marketing professional can use workflow to find clients, understand their products, and tell a vivid and credible story about how they will help health IT consumers prosper.

Back when I took my three-credit undergraduate marketing course, I learned about the original Four Ps of the marketing mix: Product, Price, Promotion, and Place. The Four Ps are now over a half-century old. Since then we’ve had the seven Ps (then eight), the Four Cs (consumer, cost, communication, and convenience), seven Cs, and, finally, four new Ps! (People, Processes, Performance, and Profit.) I could go on about interesting connections among these marketing frameworks (Processes!) and workflow, but I won’t (in this piece!).

Regarding workflow, I took courses about it during an MS in Industrial Engineering. I’ve looked a hundreds of definitions since. This is the short definition I’ve settled on: Workflow is a series of tasks, consuming resources, achieving goals. In marketing terms, you can think of goals as benefits, resources as prices or costs, and the series of tasks as what the product does. All purposeful human activity involves workflow.

So, how is workflow a bridge from features to benefits? Achieving a consumer goal is a benefit. Using a product requires a series of user-product interactions (steps, tasks, activities). Resources consumed? They start being consumed the moment a consumer realizes they have a problem to solve. They continue to be consumed after a product is acquired. And they only stop when a product is finally retired or discarded.

Workflows exist within workflows within workflows, all the way up to, and including, the workflows of life itself. Workflow extends all the way down to the micro-workflow of a series of button clicks.

Let’s imagine that a product has three salient features: A, B, and C. For example: HIPAA-compliant user authorization, ability to look up patient info, and direct staff to do something. A, B, and C are steps in a workflow. They accomplish a goal, the goal of the workflow, goal D. But goal D can be a step in a higher-level workflow, such as Help My Patient. And that workflow is embedded in an even higher-level workflow, such as, What I Do Every Day At Work. And that workflow is part of a life flow, How I Live My Life. Think I’m being silly? I’m not. Understanding how a product (the first P in the original four Ps) fits into lives of users is perhaps the single most important strategic insight a health IT marketing professional can impart.

How can you, a health IT marketing professional, use workflow to find, understand, and help health IT vendors and customers?

“Workflow” is becoming a bigger and bigger meme within the health IT industry. What makes the workflow meme so interesting and so strategic, is that workflow, in a sense, glues together all the other memes. Take SMAC, for social, mobile, analytics and cloud, for example. If you are going to create the next great health IT SMAC-based product, in what order does the user do what, at what cost to achieve what benefit? Workflow!

Furthermore, those tens of thousands of health IT products out there? No one product does everything, so products need to be combined into usable (wait for it) workflow. The biggest pain points within products (usability) and between products (interoperability) all critically involve workflow.

Before every HIMSS conference, I search over a thousand conference exhibitor websites for “workflow.” I tweet links to the most interesting content on the HIMSS conference hashtag. Last year over eight hundred of my tweets on the #HIMSS14 hashtag contained the words “workflow” or “workflows”. By the way, I’m delighted to be a HIMSS Social Media Ambassador again!

Early on, honestly, I had trouble finding much of interest about workflow on exhibitor websites. However, starting at HIMSS12, it really started to take off: four percent of websites, eight percent, and last year, sixteen percent. I’m only part way though the websites of exhibitors for this year’s HIMSS15, but I can already see this trend continuing (though I don’t know if it can actually double yet again).

So, “workflow” is in the air, in hallway conversations, in tweets, marketing, technical documentation, user forums, etc. Search in Google and Twitter for “workflow” and X, where X is a subject you already know. If you are already an ICD-10 expert, become an ICD-10 workflow expert. If you’re already a patient experience and engagement expert, become a patient experience and engagement workflow expert. You can pivot from workflow to any health information management area, and you can pivot from any health information management area to workflow. Doing so deepens your understanding and adds tools to your portfolio.

Network about your workflow interest, through contact pages, emails, listservs, blog comments, LinkedIn, and Twitter. Once you’ve started a conversation, ask for details. Ask about workflows. What happens first? And then what happens? And then what happens. What if something slips between the cracks? Do you have any workflow diagrams? Videos? Do you mind if I draw a workflow diagram and run it past you to make sure I understand how you do what you do? Use Visio, PowerPoint, draw on a napkin and snap a photo. You’ll be surprised by the degree you’re forced to prove to yourself that you really understand a product. And your interlocutor will be impressed (or pestered, you still gotta sell your value relative to the cost of their time).

Now that you understand a health IT product workflow, you have a detailed roadmap between low-level product features and higher-level user goals. If you need more context, bump up a level and understand how that workflow relates to even higher-level workflow and goals. If you need more nitty-gritty, drill down to screenshot-by-screenshot micro-workflow.

If above sounds like a lot of work, it is. It’s worth it. First of all, you’ll prove that you really understand the nuts and bolts of a product and how it fits into a user’s world. Second, you’ve got some great content. Workflow diagrams, simplified, annotated, and made aesthetically attractive are great for blog posts, white papers, presentations, and so on. The health IT market is a collection of complicated and intricate micro markets. The biggest, most costly, and (to the point) beneficial differences between health IT products are differences in workflow. Any means to more deeply understand, represent and communicate these differences is all the good: the consumer’s, the vendor’s, and yours.

Everyone is an expert on their own workflow. If you can vividly and credibly show me, an expert on my workflow, that your product fits perfectly into my workflow, I’m impressed. This is what I mean by the title of this column. Marketing Workflow Is An Incredible Opportunity To Differentiate Health IT Products, And You!

Workflow: It’s not just for industrial engineers anymore!

The following two tabs change content below.

My Twitter profile
Charles Webster, MD
Charles Webster, MD, MSIE, MSIS, evangelist for great health IT workflow, is the self-anointed King Of All Workflow In Healthcare (AKA The “Workflow Bear” per !)

Healthcare Systems Process Improvement Conference #SHS2017 Tweetchat Off to Great Start!

I started thinking about a Healthcare Systems Process Improvement Conference-themed tweetchat back in 2015, when I saw Jim’s tweet.

This year a bunch of stuff fell into place. I suggested that SHS2017 ought to have what are called Social Media Ambassadors (just like HIMSS17!). I ended up one of 10 SMAs in this year’s inaugural group. I love tweet chats, where folks tweet about related topics at a predetermined time using a common Twitter hashtag. I participate in about a half dozen tweet chats a week (#HITsm, #HCLDR, #KareoChat, #AskAvaility, #MEQAPI, and often another couple of one-off special-purpose chats). So I decided to host a tweet chat during SHS2017.

Then I realized that the #MEQAPI (Measurement, Evaluation, Quality Assurance, and Process Improvement… you can’t get more HSPIish than that!) tweetchat occurs during the afternoon of the first full day of HSPI! Why not have a joint #SHS2017/#MEQAPI tweet chat?

Before we get to the tweetchat tweets, I’d like to highlight a kind DM from a participant.

“I was really pleased by the ethos and approach to the topics today. It’s refreshing as a patient and a clinician.”

To which I replied:

“Thank you for kind words. You happened upon an unusual community. The tweet chat was associated with the annual conference of healthcare process improvement professionals. They (we) look at the bigger picture, but in a practical way. The attendees would appreciate your sentiments. Thank you participating in the tweet chat. I hope we do it again sometime! — Chuck “

On to the #SHS2017/#MEQAPI tweets! (Just a small subset, by the way!)

There were mostly crickets in direct response to this question (should we be concerned?). However there was also lots of good-natured palavering and debate about other stuff. (Not a problem, that’s just like real life conversations!) For example, there was an interesting conversation about consultants between and

That’s enough examples of tweets from the tweetchat. Tweet chats are way more fun to participate in than read after the fact! However, here are another couple tweets. I include them due to their positive sentiments toward this tweetchat and having another tweetchat next year.


On Periscope!