Tuesday, March 17, 2020

Coronavirus: Just the stats...

ItalyUSA
  # Cases  NetNew  New%   # Cases  NetNew  New% 
2/23             155  3/5             159  
2/24             229                74 32%3/6             233                74 32%
2/25             322                93 29%3/7             338              105 31%
2/26             453              131 29%3/8             436                98 22%
2/27             655              202 31%3/9             603              167 28%
2/28             888              233 26%3/10             808              205 25%
2/29          1,128              240 21%3/11          1,135              327 29%
3/1          1,694              566 33%3/12          1,762              627 36%
3/2          2,036              342 17%3/13          2,247              485 22%
3/3          2,502              466 19%3/14          2,836              589 21%
3/4          3,089              587 19%3/15          3,680              844 23%
3/5          3,858              769 20%3/16          4,663              983 21%
3/6          4,636              778 17%3/17 
3/7          5,883           1,247 21%3/18 
3/8          7,375           1,492 20%3/19 
3/9          9,172           1,797 20%3/20 
3/10        10,149              977 10%3/21 
3/11        12,462           2,313 19%3/22 
3/12        15,113           2,651 18%3/23 
3/13        17,660           2,547 14%3/24 
3/14        21,157           3,497 17%3/25 
3/15        24,747           3,590 15%3/26 
3/16        27,980           3,233 12%3/27   

Sunday, March 15, 2020

Coronavirus: A Defining Moment for Great Companies

Somewhere between the coverage of Fox News (@FakeNews) and CNN (@ZombieApocalypse) is the truth around how COVID-19 will impact the USA – and I actually think the data is scary:

ItalyUSACalifornia
 # Cases  # Cases  # Cases 
2/23             155 3/5             159 3/12             198 
2/24             229 3/6             233 3/13             247 
2/25             322 3/7             338 3/14             369 
2/26             453 3/8             436 3/15
2/27             655 3/9             603 3/16
2/28             888 3/10             808 3/17
2/29          1,128 3/11          1,135 3/18
3/1          1,694 3/12          1,762 3/19
3/2          2,036 3/13          2,247 3/20
3/3          2,502 3/14          2,836 3/21
3/4          3,089 3/153/22
3/5          3,858 3/163/23
3/6          4,636 3/173/24
3/7          5,883 3/183/25
3/8          7,375 3/193/26
3/9          9,172 3/203/27
3/10        10,149 3/213/38
3/11        12,462 3/223/29
3/12        15,113 3/233/30
3/13        17,660 3/243/31
3/14        21,157 3/254/1


Keep in mind that the Italy numbers after around March 5 are reflective of that country’s draconian measures that may not be possible in the USA (for example, a virtual 24 hour curfew imposed by a national police force).  It is reasonable to think that the numbers in the USA may be much higher by the end of March.

We at AnalyticsIQ began mandatory ‘work from home’ this past Friday for our 20 Atlanta-based employees (the rest of our staff already work in home offices).    This is a small step – but it is very likely that some sort of centralized government action will be required.  While schools have been closed and events have been cancelled, it seems that restaurants will be next (which has just started in Belgium and France).

If the USA turns out to be an analog to Italy, it will be result in a serious economic shock.  So, this is where great companies will separate themselves from the pack.  Great companies will treat their employees and vendors as partners during this potential crisis.  Great companies will make it clear to investors that a short-term drop in profit is not reflective of long-term value.

Tuesday, March 10, 2020

Pandemic Math

It always interesting when the domain of analytics enters the broad market.  The ongoing outbreak of COVID-19 (“coronavirus”) has naturally led to much public speculation around infection and mortality rates.

Of course, the press has always struggled with interpretation of statistics (and I can only assume that highly analytical people are not drawn to become journalists).  A classic case is confusing correlated and causal events.  For example, it is true that the majority of tall people where big shoes, but it is unlikely that wearing big shoes makes you tall.

In viewing various stories related to the pandemic (and press conferences held by the president, the VP and various CDC executives) it is obvious that there is no common understanding of what the infection and mortality statistics actually mean.  For example, President Trump and his CDC appointee threw out wildly different mortality rates in the same press conference.  Not comforting.

At the time I write this, the latest statistics are as follows:

·      762 reported cases of infection
·      28 reported deaths

Given that 28/762 equals 3.7%.   End of story?  No – here is the first problem with using 3.7% as the mortality rate:

·      It would seem likely that infection cases are underreported.  There are reports that extremely healthy people heal from infection without hospitalization
·      On the other hand, a death is very unlikely to be under (or mis-) reported.


This implies that the mortality rate is something less than 3.7%, right?  No.  The infection cases are at all stages of illness, with a weighting towards recent infection (due to the very rapid spread of the virus).    Without analyzing the time-based cohorts of the infected, it is impossible to estimate the actual mortality rate – but it is very likely to be > 5% of those who are reported to be infected. 

Tuesday, February 25, 2020

AI and Predictive Analytics


First off, let’s get something straight:  we are many years away from developing technology that can match humans’ ability to do things like answer complex questions or create art.  For example, a machine that can develop a solution for the question “How do we increase our marketing response rates?”.  In some future time, perhaps a computer will solve this question by accessing relevant databases and developing a strategy that optimizes response.  Sounds great?  Of course, if everyone has access to this same technology, no one will have a competitive advantage.

Luckily, in the year 2020 a human would need to be heavily involved in solving complex questions like this one.  Make no mistake, a computer can perform necessary calculations much faster (sometimes millions of times faster) than a human.  But pointing the computer at which items to calculate is still the exclusive domain of humans.

At AnalyticsIQ, we do test every relevant technology.  Some have promise as tools that can create better (defined as more predictive and/or more stable) algorithms – although nothing ‘game changing’.    While there is an understandable fascination with AI and machine learning – they are still analogous to a car:  they will get you somewhere much, much faster than walking, but you still have to tell them where to go.