Snake Oil 2.0 why big data is not necessarily good for your health

Maybe small data is where we should be looking

Yummy but not good for you

Why more data is bad

Remember the old joke regarding college degrees? BS = Bull Shit, MS = More Shit and PhD == Piled Higher and Deeper and HBS == Half Baked Shit.

In Western society, we are schooled to believe that more and faster is better — even though we can see that big data analysis is paying off in a very small number of use cases (everyone is quoting personalized genomics and drugs) and that large scale data breaches are the direct result of hackers going after the big juicy customer data sets.

Your marketing, technology, logistics and business development staff are all information junkies, not getting enough and wanting more.

Is lots of data really good for healthcare?

Our clinical trials customers often feel they are not getting enough information — even though the site and the project management staff feel that they (the staff) provide them (the biotech sponsor) with lots of information via interactions online, by phone, email, at face to face meetings and in formal product presentations.

Your CRM statistics may tell a story of high impact private networks for sale, the number of online seminars and Web site visits and engagement but customers often feel that they are getting no useful information at all from their vendor and account managers.

When customers and decision makers finally do have a private, face-to-face meeting with a salesman and technology expert in the privacy of their office, they almost always feel that they have been given valuable information, even if they are unhappy with the answers or want to seek a competitive offer.

Why does this happen?

Utility is reference-based and not additive

As prospect theory predicts, utility (the value of a product or service) is reference-based and not additive.

In other words, more data from technical, sales and marketing staff and customer support groups is less valuable than data received when the frame of reference is a private consultation with a senior product manager regarding a technology solution — for example, data loss prevention technology to prevent data leakage of patient records in a large hospital organization.

Framing favors customers overrating a face-to-face visit with an expert sales engineer and underrating digital communications — even if the technical content is identical.

Framing effects in the customer relationship may also be related to cultural and societal factors.

In countries where managers function within a hierarchy, decision makers will tend to value personal visits from senior sales engineers over email, social media, Dr. Google and online technology forums.

Framing effects create mismatched perceptions and expectations in an asymmetric relationship — where technical decision makers (at the bottom of the totem pole) get information but do not value it and sales and engineering staff (the experts at the top of the totem pole) provide information and expect the customer to value the information and then become frustrated when their prospective customer downgrades the value of their messages.

Closing the gap between vendor messages and customer assessment of quality is critical to customer satisfaction, improving the customer relationship and achieving higher sales and product satisfaction.

From a data security perspective — storing less data is more secure than storing more data.

From a sales and marketing — getting a small number of right messages out to the customer is good marketing and effective sales.

I’d love to hear what you think — clap or comment and tell me where I’m wrong.

Danny is founder of Flaskdata.io. He is a solid-state physicist by training, serious amateur musician and tech entrepreneur. He is involved in a cutting-edge project that harnesses AI to help us achieve transparency for our clinical data.

Proud to be working with some very smart people, all smarter than him.

I am a physicist by training, serious amateur musician and everyday biker. Working in cybersecurity and AI-driven monitoring of clinical trials.