TOP 10 REAL-WORLD APPLICATIONS OF ANALYTICS - PART 1
Too often, discussion of analytics and big data focuses either on high-level abstractions about its power to boost competitiveness or customer loyalty (spruiked by academics and some of the large consultancies) or descends into the specifics of IT investments and software (spruiked by software vendors).
Neither of these captures the real, hard-hitting benefits of what analytics can accomplish. After the conceptual discussion at 30,000 feet, and the specifics of (very expensive!) IT investments the question remains: Why are we doing all this?
This article answers that question in concrete terms: it sets out my top 10 real-world analytics applications. All are actual, on-the-ground applications which are documented: there are no theoretical examples or mere hypotheticals. Contact me if you would like the reference for any of them.
Let’s dive in, in the first of a two part series, to see how analytics is being used to benefit businesses, infrastructure departments and health services …
1. Amazon’s ‘Customers who bought this also bought …’
This is possibly the use of analytics with the widest exposure: everyone is familiar with this application even if they don’t know that analytics underpins it. Amazon’s recommender system runs while you are browsing or buying; it matches your preferences with items or the preferences of people with similar browsing or buying histories.
While the actual algorithm used by Amazon is proprietary, the principles and techniques (correlation analysis, collaborative filtering) are well canvassed and can be adapted and used by businesses to enhance customer experience and maximise cross-sell and up-sell opportunities.
2. Improving heart attack response times
Of people who suffer heart attacks that occur when coronary arteries suddenly become blocked (so-called ST-elevation myocardial infarctions, or STEMI) 15 percent die within 30 days. Rapid intervention to unblock the artery is key to achieving better outcomes: the US national standard for the period of time from when the patient enters the hospital to relief of the blockage is 90 minutes.
US health provider Intermountain Healthcare is a pioneer and leader in the application of analytically-driven services. Starting in 2011, its cardiology team set internal time standards and measured results for treatment of STEMI patients, with data on each heart attack patient circulated to the whole team. This process improvement and feedback loop helped reduce Intermountain’s median response time to 57 minutes.
In the last three years, every Intermountain STEMI patient has been treated in less than 90 minutes, and its rate of patient survival beyond 30 days is 96 percent. Put another way, Intermountain’s mortality rate is only four percent, compared to the US average of 15 percent.
It’s hard to think of a more high-impact application of analytics, and it was data-driven.
3. Data-based health care
Another Intermountain health care application: an analytics tool runs at 9.15 every morning in the cardiovascular unit in each of Intermountain’s 22 hospitals and takes readings from patients’ vital signs. An email alert informs clinicians of those patients at risk of heart failure and their likelihood of readmission or death. This helps Intermountain adapt its patient care pathways, perhaps assigning patients to palliative care, or providing home care.
Impressive. And analysis-based.
4. Improving asset condition … and budget funding
The New Brunswick Department of Transport (NBDoT) in Canada has implemented an elaborate analytical model to estimate and optimise its spending on road maintenance and asset management.
With a portfolio that includes 18,000 kilometres of roads and 2,900 bridges, the NBDoT moved from a ‘worst first’ approach, to an approach that programs the amount and timing of expenditure to make the greatest improvement on asset condition.
As a result of this, its managers can now make long-term decisions with confidence, and they can quickly model the effect of changes to its plans. The consequences of deviating from the optimised plan can be easily quantified and communicated, thereby largely removing the politics from the decision-making process.
Because of this NBDoT is realising $72 million in annual savings ($1.4 billion over 20 years) and they have secured substantial additional funding for pavement rehabilitation from the historical level of $50 million to a new amount of $120 million.
All of which makes for a massive return on their investment in analytics.
5. Controlling employee calories
Analytics doesn’t only involve crunching numbers and performing logistic regressions on huge datasets: it also entails conducting experiments to isolate specific impacts and using the results to improve performance.
Google does this. And it applies its analytical rigour not only to AdWords campaigns and the like, but also to its people management.
Google provides free food to its staff (co-founder Sergey Brin stipulated that every Google employee be no more than 200 feet away from free food). However the tech giant was concerned that its employees were consuming too many free chocolates, with negative effects on workers’ health and happiness. So it conducted an experiment and hid the chocolates in opaque containers, while prominently displaying healthier options such as dried figs and pistachios.
The result was that in the New York office alone, employees consumed 3.1 million less calories over seven weeks.
In a similar test, Google experimented with plate sizes in an effort to get people to eat smaller portions. Around one-third of staff chose the smaller plates without going back for additional helpings, and the use of small plates increased a further 50 percent after Google posted the experiment result on cafeteria signs.
* * *
I hope this gives you a sense of what analytically-based management can accomplish. The second instalment on this topic will deal with analytics applications as they’ve been used with casinos, banks … and pop-tarts.
Until then, if you would like to discuss how analytics can be used in your organisation to accomplish benefits such as these, please contact me by return email, or on 0414 383 374.
Director I Michael Carman Consulting
PO Box 686, Petersham NSW 2049 I M: 0414 383 374
© Michael Carman 2015