Covid-19 and what we can learn about Data

If you are inside the marketing or management world, I′m pretty sure you said and heard several times that “data is the new oil”. It is also very likely that some friend of yours from a totally different background called you “data guy” and laughed at you for being so paranoid about that.

Well, he is not laughing anymore.

From the moment that Covid-19 hit the world with impact and became a pandemic, everyone suddenly became a data addict, searching daily for all kinds of information, such as the number of cases worldwide and in the home country, as well as hospitalized and mortality rates.

Obviously not the ones we wanted, but this times have proven how data is important to understand the present and to plan and predict the future.

At the same time, Covid-19 also incentivized us to be critical about data. Data is only valuable if we understand the context of it. We have to comprehend how it was collected and what does it really mean, otherwise data is just a bunch of numbers.

In the case of Covid-19, this has become pretty clear in many aspects:

• The population size is different between countries. It is predictable that countries with more people will therefore have more infected and deaths. It is not enough to look at the total numbers in order to tell a country is doing better than other. We need to look at that in terms of proportions.

• The same thing applies to tests. Only the people who take the test can be counted as infected and the capacity to make tests is still low in many countries. For example, comparing at this point the two countries with more positive cases confirmed, USA has 51.110 tests done p/ 1M habitants and Brazil only 4.378. Even though far from perfect in both, the number in the first is much likely closer to the reality.

• Without more tests, we cannot be sure of the real mortality of the virus. Although this is terribly bad, a positive thing is that we are only seeing the mortality rate amongst the ones that were detected as positive, so it is predictable that this rate is considerably lower.

• Some people are tested more than once. Given this, the number of tests made do not match the actual number of people tested so far.

• The type of tests made varies from country to country. Also, no tests are 100% trustworthy, with some research suggesting that the rate of false negatives might exceed 30%. This is another clear evidence that the number of infected is higher.

• Covid-19 data is not reported and published immediately as it needs to pass through complex operational processes. Therefore, countries publishing numbers once a day are not necessarily showing data exclusively related to the last 24 hours.

• The data we see regarding Covid-19 is out of sync. Some people die weeks after being tested positive. Some people only have their test result days after being hospitalized. And it usually takes more time to recover than to die. Therefore, the infected, hospitalized, dead and recovered curves need to be seen separately.

• We need to be aware of fake data and data manipulation. Taking China as an example, it is hard to not be suspicious when analyzing their numbers.

• In some countries, people that die before having the test are not tested afterwards. Although this may be done with a good purpose (in order to save tests to the ones that need them) it obviously needs to be taken into account.

And the list could go on.

For instance, we can′t just look at Covid-19 data and just forget everything else. Countries that have less deaths of Covid-19 but are experiencing more deaths in non-covid related causes cannot be considered success cases.

But does this mean that data is useless? Not at all!

We shouldn′t stop looking at data just because it is not perfect. Excluding the manipulated one, the data that we have in a moment is the closest we are from understanding the reality. It is crucial that politicians base their decisions on the available data during this pandemic, and it is fundamental that they continue do so on every decision they make, after this is over. The same thing, obviously applies for marketeers and entrepreneurs.

To conclude, the point is: It is not just about having the data. It is about how you use it.

Stay safe!