In the field of e-commerce, the Amazons and Wayfairs of the world have mastered the art of leveraging data to show the right product to the right customer at the right time.
In the face of plummeting workforce numbers due to the COVID-19 pandemic, some healthcare organizations are looking to e-commerce strategies to improve their own staffing and scheduling processes – showing the right shift to the right per diem worker at the right time.
Whether shopping online for a couch or combing through available shifts, people will tolerate scrolling on their phones for only so long, so data science must be offering users options they are likely to pursue.
The stakes are, of course, much higher in healthcare. If a consumer doesn’t buy a couch, another may; but if a shift goes unfilled, a healthcare organization may not meet mandated staffing ratios – and quality of care can suffer.
Ike Nnah, cofounder and chief technology officer at IntelyCare, a vendor of healthcare staffing and scheduling technology, discusses with Healthcare IT News the draw of e-commerce strategies, how healthcare can use these strategies, the role artificial intelligence plays and some examples of all this in action.
Q. Why would healthcare look to e-commerce for help with data science? What is the draw?
A. E-commerce already has undergone this digital transformation that healthcare staffing is in the middle of. The e-commerce space has a wealth of talent, experience and insights that the healthcare staffing industry can harness.
There are similarities between the world of e-commerce and healthcare staffing, namely how a consumer purchases a product and how a healthcare provider chooses to pick up a shift. But there are plenty of differences that must be addressed, too.
For instance, geography is critical in healthcare staffing, while e-commerce typically prices on the national level. And healthcare staffing’s product (available shifts) has a short shelf life, compared to the products listed on sites like Wayfair or Amazon.
But, healthcare organizations can be successful in adopting data science practices if they can harness the learnings of e-commerce models, apply them in a thoughtful way that considers their similarities, but tailors their operations to fit the needs of providers and facilities.
This does not apply to brick-and-mortar staffing agencies. It only matters for businesses that let providers choose which shifts they want. When it’s more of a shopping-for-shifts experience, e-commerce is a natural place to look for ideas.
As healthcare staffing becomes more digital and increasingly grants providers more freedom to choose the shifts they want to work, the staffing experience starts to look a lot like a shopping experience.
Only instead of buying books, providers are choosing what shifts to work. But the key differences are these: Shifts are more ephemeral than books or mattresses, and providers are only choosing shifts from nearby facilities rather than a global catalog.
Q. What are a couple of data science strategies and practices from e-commerce that healthcare staffing can leverage?
A. Personalization is one example. We know that consumers live on their phones, and providers are no exception to that rule. They like the ability to build their schedules and pick up shifts all from their phones. Doing so is easy and natural for them.
However, there are sometimes thousands of shifts to choose from, and they can’t possibly view all of them on their mobile device. So, healthcare staffing needs to be selective about how they display shifts. If the best shifts aren’t on the first page, providers are not going to see them.
Making shift recommendations based on a provider’s settings and behavior can optimize the shift-filling process. This makes it much easier for facilities to fill shifts while providing a better experience for providers to hunt for their perfect schedule.
Dynamic pricing provides the ability to change prices frequently, constantly leveraging all available data from purchase history to get the most successful prices in the future.
While this happens all the time on e-commerce sites, it’s not used to its full potential in healthcare staffing. Dynamic pricing allows healthcare staffing vendors to raise the price of the hardest shifts to fill and lower the prices of the shifts easiest to fill, effectively improving the chances of filling both shifts.
Ultimately, this strategy benefits the provider, the facility and the patient. Providers optimize their earning potential, facilities optimize their schedule, and in turn, floors are well-staffed and patients get the care they deserve.
Q. Where does artificial intelligence come into the mix here? What role does it play?
A. As the algorithms for pricing and personalization become more sophisticated, they start to move into the realm of true artificial intelligence. These algorithms can respond to patterns businesses may not be aware of.
If engineered correctly, these algorithms are basically another team member. They make business decisions like any employee – just limited to whatever decisions the business is comfortable allowing it to make.
So, as long as the business creates appropriate bounds for these algorithms, they can work for the company around the clock, they can be scalable, and they can generate positive outcomes for a business and for the future of healthcare staffing.
Q. What is an example or two of principles of behavioral economics that can develop algorithms to more effectively “sell” finite, time-restrictive resources like shifts?
A. The first is loss aversion. In some cases, instilling the fear of losing out on something (like a product or a shift rate) may be more motivating than simply presenting good products, savings or bonuses.
How you frame the price of something to a consumer also is tremendously important. For example, you can advertise a shift as $16 per hour or promote it as $108 for the shift. While they’re the same payout, in the end, consumers respond differently to their framing – so considering how you communicate a reward is essential.
Finally, the timing of a promotion matters a great deal. Data science can create practical exercises in determining what is more motivating to a consumer or, in this case, a provider – the promise of larger, future rewards versus smaller, instantaneous ones.