Announcement: We’ve partnered up with The Healthcare Rap Podcast’s Jared Johnson & Virtua Health’s CRM & Digital lead on a webcast designed to help you learn effective precision marketing. Learn more below:
“You need to aim better, Dad!” Fortnite and Precision Marketing
“You need to aim better, Dad!” This soundtrack has been on repeat over the past few months at my home in Lafayette, Louisiana. My 11 year old, Merrit, routinely showcases his prowess in the immensely popular/cultural phenomenon video game Fortnite. He consistently provides feedback on my skills (although it feels eerily like a little league coach’s rebuke). Fortnite, for those not familiar, offers players the ability to compete in a battle royale survival game. Properly taking aim in high stakes scenarios is a premium skill (and one I seem to be lacking).
While I understand I need to “aim” better, putting my son’s coaching into action is a wholly different challenge. What’s especially difficult are the constant distractions and moving targets that take my attention off one priority target and onto the next. In my efforts to improve my aim, I’ve deployed a bevy of tactics. These include on-line research, self-help videos on Youtube, and analyzing my game statistics. All these suggest that this will be harder to master than Super Mario Brothers.
“In healthcare marketing… competing and changing priorities cause confusion as to who and what to target”
It occurred to me that while I understand what I need to do, I’m not exactly sure how to do it. In healthcare marketing, we are challenged with a similar plight. Competing and changing priorities cause confusion as to what and who to target. In Fortnite, I’m informed of the success or failure of my targeting at the end of each game. That said, I don’t get the data I need to know how to consistently improve my precision.
In healthcare, however, we have an immense amount of data at our disposal. And just like my videogame plight, the challenge isn’t understanding the necessity of data, but rather the guidance in how to use it.
Where data analysis goes wrong in healthcare marketing
Data-driven marketing isn’t new to the healthcare industry. With the rise of Electronic Medical Records (EMR) and Health Information Exchanges (HIE) has come a flood of patient data. Healthcare marketers are tasked to use this data to determine strategy, campaigns and patient/consumer outreach — all while staying HIPAA compliant.
However, most marketers aren’t trained data scientists and don’t necessarily know how to work with all these disparate data sources. That’s not their fault — data may be widespread in healthcare, but analytical training for marketers lags behind. As a result, healthcare leaders often make decisions requiring high data prowess without awareness of cognitive biases and data fallacies. This can cause ineffective campaigns and untrustworthy reporting, or worse, it can hurt trust among patients and consumers.
Data fallacies dilute efforts to drive patient acquisition and loyalty
Cognitive biases & data fallacies are human. So human, in fact, there are dozens of them with fun names and psychological studies behind them. Human though they may be, they still hammer patient engagement and outreach efforts. Let’s discuss three data fallacies that we commonly see healthcare marketing: Cherry Picking, Selection Bias, and The McNamara Fallacy.
Cherry Picking – Choosing stats that you like
Everyone wants to believe that their marketing efforts are working (why wouldn’t you, your job depends on it!). That said, it makes us prone to cherry picking, where we selectively pay attention to data that confirms our position while ignoring data that doesn’t. Cherry picking may lead to over emphasis on certain channels or campaigns at the expense of others. It may also lead to inaccurately assessing how well some initiatives work.
Cherry picking feels great during the presentation, but it’s awful on the health system balance sheet
Example: Quarterly meetings with a marketing vendor are upbeat and positive, but the bottom line never seems to change. Each quarter, the vendor shows growth in various service lines when compared with previous month, year over year, or quarterly performance. Without consistent reporting, cherry picking runs rampant. It feels great in presentations, but it’s awful on the health system balance sheet.
Selection Bias – Data not properly collected
Intuitively, most people would think that the best way to understand who is likely to have an illness is to study traits and characteristics of other people who have had the illness. This is the basis of most propensity models, but where can this go wrong in a campaign?
The “selection” in this example is skewed toward observed people who have come in for care and received a diagnosis. This ignores unobserved populations whose needs are just as urgent, but who may not have been able to afford care. We saw this play heavily with AI ethics issues in 2019 that could have been avoided.
Example: A data model incorporates many attributes on a patient to identify who is a good fit for a campaign, but the marketing group fails to ask what these attributes are (or the vendor refuses to disclose). The algorithm is then optimized for non-clinical factors such as the insurer, income levels, and psychographic profiles.
There are two possible outcomes: People in vulnerable populations with clinical need are left out of important communications, or these same people are targeted by competitors who are using a better data strategy!
McNamara Fallacy – Too much faith in available metrics
Metrics aren’t everything. No, that isn’t a typo. Also known as quantitative fallacy, the McNamara fallacy refers to measuring only what can easily be measured and disregarding what can’t. Falling into this trap leaves organizations prone to using ineffective proxies and vanity metrics to measure more conceptual or qualitative goals.
Example: A hospital marketing team wants to report on how much their new email content resonates with recipients. Within two weeks, they report a 26% open rate. Great, right? Not necessarily. Though the vanity metrics available looked great, this group of patients booked appointments at a lower rate than others, and they unsubscribed at 17%. The high open rate could have been an urgency to unsubscribe!
Marketers need more data skills, but that doesn’t mean they need to become full-fledged data scientists. Rather, healthcare organizations should adopt a culture of Data Enthusiasm all throughout.
Data Enthusiasm: why it matters and how to embody it
Data is crucial for healthcare marketers. It tells you who you are talking to. It lets you have conversations with people without even speaking to them. Through understanding heuristics and creating algorithms and models, you get more explainability than human decisions can offer. With machine learning in particular, you can make more accurate decisions that are less prone to bias.
Executives are briefed on data reports but may lack the data science knowledge to look at these reports critically…
And yet, our organizational approaches oftentimes do not enable this use of data. Marketers aren’t trained enough in data analytics to use it effectively on their own. Data scientists are highly trained, but work in such a technical field that marketers and executives alike may struggle to understand what they’re talking about. Executives are briefed on data reports but may lack the data science knowledge to look at those reports critically and truly engage with the health of their business.
With Data Enthusiasm, data becomes a mentality. It’s an organizational approach wherein we see the value of individuals knowing how to fight bias and make data-driven decisions, even if they never learn how to implement machine learning in Python. Marketers can speak data science fluently enough to infuse it into their strategies, data scientists speak marketing enough to give marketers the deeper insights they need in plain English, and executives speak both so they can more deeply engage with the insights presented to them.
Data Enthusiasm correlates with patient trust
Healthcare has a higher calling than other sectors. You may ignore constant sale emails from big box stores, but still shop there when you need something. If your doctor spams you with disingenuous emails, however, you may lose trust in them and take your healthcare elsewhere. It’s not about getting patients to buy more things, it’s about helping them invest in their health.
This means that the same accuracy metrics that are so critical to medicine — precision, recall, false negatives, etc. — are also critical to healthcare marketing. In order to achieve these metrics at scale, it requires Data Enthusiasm across the organization. The right technology can help, too. AI-powered CRMs make it easier for marketers to practice Data Enthusiasm and make better decisions.
At SymphonyRM, we call those decisions Next Best Actions. The example in the link shows the business intelligence and prioritization needed to prioritize vaccinations, annual wellness visits, and patient portal enrollment. In addition to services like these, Next Best Actions use data science and algorithms to prioritize most interactions and outreach you’d have with patients and providers.
Data Enthusiasm in a post COVID-19 world
As the pandemic continues to unfold, most everyone in healthcare is impacted in one way or another. In a post COVID-19 world, organizations will need Data Enthusiasm to effectively reschedule and schedule lost revenue, engage in strategic service lines, and build and affirm trust with both providers and patients. In doing so, they can structure outreach based on service line capacity, patient need and risk, and CMS prioritization factors. Our Post COVID-19 Road to Recovery can help you implement this in your organization.
I, like many of you reading this, looked for the silver lining that this COVID-19 disruption has caused and the coming “new normal.” Distance provides perspective and insight, and so does data. While it may take many more months for me to gain my son’s approval regarding my ability to aim and succeed in targeting, we as an industry need not wait. We have the ability to embrace the data that informs our decisions, the data that directs our focus, and the data that confirms our success and impact.
Our communities need us to use every tool at our disposal in the new normal to engage them in proactive, personalized, and precise communication. To that end, the ability to make sense of the data deluge is the most powerful tool we have! Ok, I have to play a quick Fortnite match before my son wakes up and starts his rebuke…
*Banner photo credit: frankv @ unsplash.