Could Big Data Predict Funeral Work?

    One of the most common annoyances of the funeral industry is the lack of predictability for the work.  However, recently I was in a class on quantitative research where we discussed 'big data' and I realised how it could possibly apply to the funeral industry.  That this new form of qualitative research might be able to accurately predict specific details about funeral work.

Which would be quite a benefit to many companies and individuals!

    Qualitative research is basically statistics, it breaks down the data into numbers and tends to look at larger concepts.  Political surveys, which look at who will vote for which party, are a good example of qualitative research.  Another example is the way the Australian Bureau of Statistics looks at Australia's national population.

    Traditionally this type of research has been done with 'random sampling' where a random segment of the population is surveyed.  The idea behind this is that it would be impossible to survey everyone, so a random sampling is most likely to be accurate of the larger population.  While not perfect this method has been the standard for a while, and worked quite effectively.

    However, due to recent social changes random sampling is becoming harder and at the same time access to larger segments of people is more possible.  First, random sampling is harder to achieve in a  few ways, take phone surveys for example.  Now only certain segments of the population have a landline phone.  So phone surveys will tend to get older people and exclude younger people.  On the other hand internet based surveys are just as bias for the same reason, in that they tend to be dominated by younger people and exclude the elderly.

    This is just one basic example but essentially random sampling is doable, yet becoming tricky and less sure that the sample is truly random.  At the same time we are seeing a rise in 'big data', a new qualitative method which is surprisingly accurate when done right.

A good video on 'Big Data':

    Big data is not a smaller random sample of the population.  It is a larger sample, of the whole or majority of a population, or alternatively a collection of many, many, random sampling style studies.  A good example is of the recent prediction of the American election.  Nate Silver, a statition accurately predicted the election, even down to swing states which won by as little as 2%.

    He did this not through random sampling of populations, rather he looked at voting history in America over the last few decades.  In other words he performed a HUGE study, looking at the relevant statistics on a massive scale.  As a result he was able to very accurately predict the election results on a specific and localised scale.

    Another example of big data is with facebook and social media trends.  While this information only represents a certain group and excludes others it has been found very reliable.  The data is so large, there is so much of is, that it becomes reliable in itself anyway.  If done right the volume of information makes up for the exclusions and bias and gives a very reliable result.

    Here is where the funeral industry could be interested, and a perfect place to practice or refine big data.  A large issue with the funeral industry is the unpredictable nature of the work, some days are busier than others, but generally one cannot predict when work will spike or what type of work it will be.  As a result funeral companies need to be very flexible, but this is a bit of an inefficient (while effective) system.

    I remember a coffin manufacturer saying how if he could predict what coffins would be needed when that he could then streamline his business and save money.  As it stands he has to have a decent surplus of coffins sitting about, until sold they are essentially lost money.  Or with funerals homes, having to quickly call in casuals when it gets busy.

    Big data could potentially solve, or at least mitigate, this lack of predictability.  Imagine if you could reasonably know how busy the company will be in two months.  And more importantly, one would be able to predict the type of work in two months.  If it will be cremation, what coffin they will want, where the transfer will be, and so on.

    We will never know if big data could answer this unpredictability, but I think it would be worth a try.    However, the issue with this study is not the idea but the logistics.  A big data study would require the whole industry across Sydney, and possibly even Australia, compiling their records into one big study. This compliance would be difficult at best as most funeral companies are in direct competition with each other.

    If everyone came together, if all companies put their data into the study, then the funeral industry might become predictable.  There is a lot to gain for the companies, so hopefully one day it will be attempted.  Maybe it will fail, maybe the data will find nothing.  But we will never know if we never try.  It is something that should be considered and more companies should ask if Big Data could remove or reduce the unpredictability.

Further reading on big data:



On a "new" profession -- data science: 



Predicting elections and big data: 






No comments:

Post a Comment

Never hesitate to ask a question or comment on something, this is an open minded and free space.

If you want to contact me privately do so at: theothersideoffunerals@gmail.com