Chaff - When Measurement becomes the Problem

Wednesday, March 25, 2015

Does increasing contact center diagnostics insure better customer experience?

At one point in my career I worked in the field of medical diagnostics, selling diagnostic testing to physicians. 

If you've ever looked at your lab results, you noticed that they report where your test results lay within a reference range. The reference range is based on variables like your age, location, and the test results from a healthy population.

Clinical laboratories generate reports that highlight outliers, alerting your physician to potential issues. However, physicians know that outliers can represent 'false positives'; incorrect results, or results that are correct, but don't represent actual clinical or pre-clinical condition. These false positives are 'Chaff', if followed they result in unproductive or counter-productive treatments.

Reference Ranges - 'Normal Ranges'

Because reference ranges (except for some lipid studies) are typically defined as the range of values of the median 95% of the healthy population, it is unlikely that a given specimen, even from a healthy patient, will show "normal" values for all the tests in a lengthy profile. Therefore, caution should be exercised to prevent overreaction to miscellaneous, mild abnormalities without clinical correlate.

From: Dr ED Uthman MD

False Positives

Don’t blame the testing laboratories. There is inevitably a trade-off between a test’s sensitivity (its ability to spot illness) and specificity (its ability to give a clean bill of health). Says Dr. Jon Nakamoto, an endocrinologist and a lab director for 

Quest Diagnostics : “Society and the medical community have to make a choice: How often do we get these false positives and how often do we make a difference in the outcome”–that is, save a life.

Lyme Disease - False Positive Rate 91%

Lyme incidence is 31 per 100,000 people in infested states. The Elisa (enzyme-linked immunosorbent assay) test for telltale antibodies turns positive for about 6% of people who do not have the disease. So false positives would outnumber true positives 200 to 1 in a random sample of testees.

Source: Forbes 6.23..2010 

Experienced physicians look at, talk to, and touch their patients to 'reality check' lab results.

Call center's are experiencing a rapid growth in diagnostic technologies. Technologies that are supposed to spot negative trends. Big data is collected from call distribution and call handling systems, post call surveys, transactional content (CRM data and Speech Analytics), etc. 

As these technologies are introduced, and the resulting data integrated into the diagnostic picture of customer contacts, it becomes increasing important to understand what is being measured, and confirm it's relevance to the health of the customer's experience. 

In other words, we need to introduce processes to 'reality check' Big Data diagnostics, to be sure we're not chasing after chaff.

Call Intelligence - Better by Design

Call Intelligence Inc. is a Transactional Intelligence and Design Consultancy with over 25 years helping our clients to better understand and improve the outcomes of interactions with their prospects and customers.

Our analyses allow us to map, correlate and prioritize thousands of possible relationships between marketing programs, call handling systems, calls, callers, call components, concerns, agents, competition, etc...and outcomes.