It is the phenomenon called data obesity, which today represents one of the most common traps that those who deal with data analytics fall into.
To understand how to set up an effective data collection and analysis system, therefore, we need to start from the basics: we need to ask ourselves what is the very nature of the information we want to collect.
Only in this way will we be able to focus on their actual usefulness and, from there, identify the systems for collecting and interpreting the numbers obtained.
The first step: let's identify the really useful KPIs
It is vital for every company to have exactly the data that can be useful for its activity, life stage, size. Generally, this type of data is indicated with the acronym KPI, or Key Performance Index.
The KPIs that are interesting for a company that is launching rather than mature, or that deals with intangible services to companies rather than marketing food products, will obviously be different. But what are these KPIs, ultimately?
We like to point out that they always, in fact, concern transactions .
Regardless of the nature of the company or the type of activity, companies must ultimately focus on measuring transactions in the broad sense, that is, all those movements with which the consumer approaches the brand.
They are transactions, in this sense: the moment a customer enters the site (or the store), signs up for the newsletter, requests information... and so on, up to the actual purchase. All these actions have a value, for the company, precisely because they represent measurable steps in this path that leads, finally, to the purchase - the "transaction" par excellence, the one with the highest value.
Of course, we do not claim that these are the only KPIs to monitor: other values must be carefully monitored, for example because they represent alarm bells. This includes the number of returns, or reports of problems in general. Then there is everything that monitors the efficiency of the company organization.
However, too often we end up monitoring too much, with the only real result being to distract ourselves from the real objectives, which are the transactions.
And this, ultimately, is what KPIs need to tell us: how effective and efficient we are in bringing the customer closer to us, all the way up to the actual financial transaction.
More sophisticated analytics… worse analysis?
The ease with which we can collect data today is something extraordinary, especially on the web. Even tools that are now very common, even free, have potential that a little over a decade ago would have been unthinkable.
Google Analytics alone provides over 600 measurable objects, including metrics and dimensions. And this, paradoxically, is partly the cause of the data obesity problem.
The analyst tends to show the widest possible picture, since the tools allow him to do so with little effort; at a superficial glance, this approach seems to be the correct one for those who want to do a good job. By doing so, instead, pages and pages of essentially useless reports are filled, if not potentially harmful (because they distract from the truly central data).
In how many cases, for example, can it be really useful to know which operating system users are using? And what interest could the sales manager have in knowing what the loading time of each page of the site is?
