“Often data can have greater than just monetary value,” said the head of Grant Thornton Baltic’s digital services, Vladimir Rüntü, on Äripäev daily’s radio programme “Kasvukursil”. He gave the example of vaccinations and how Pfizer decided to partner with Israel mainly for sharing data. “The vaccine producer saw that the value of the data was greater than just a sum of money; the data could be analysed and used to grow the market.”
Making data profitable begins with secure use and this, according to Rüntü, is founded on three premises. “Data should be integral, available and accessible only by the right people,” he said. Rüntü emphasised that there is nothing to be gained from fear-driven disproportionate measures to protect data from leaking out. “Often, people think that by encrypting all data and squirreling them away, all will be well, but in fact, to be able to reap the benefits for a company, data must be analysed.”
To keep data accessible only to the right people, internal rules and control mechanisms should be established at the company. “Data security must be well-thought-out and the rules approved – someone has to be tasked with responsibility for verifying the data,” said Rüntü.
Data should be stored where they will be used
Rüntü said there was no special secure place for storing data, and the guiding principle should be to store data where they are needed. “It can’t be said that the cloud is more or less secure than anywhere else – the data just have to be available and visible only to those who have to see the data.”
Data should definitely be categorised – that makes them easier to use later on. Rüntü gave an example from private life. “If you file your photos into one photo album, it’ll be harder to retrieve the right memory later on – so better to categorise everything by date, location or some other classifier. It’s the same way with the data companies generate: if the metadata are well-organised, they will help in finding the right information later on.”
To keep the process of organising the data from becoming too time-intensive later on, it has to be an ongoing process. Organising data is an everyday task, new customers are added every day and errors may be introduced in the data entry process, which is something that could be corrected continuously,” said Svea-Elen Peters, head of business development for Luisa Translation Agency.
Avoid manual entry
In addition, experts recommend that all incoming data should move as smoothly as possible to where it needs to go. “Often an online retailer’s orders don’t feed straight into the system; rather, they are accepted by email and then entered individually into the accounting program,” said Rüntü. “Instead, solutions for integrating the e-store with the accounting software should be explored – this raises efficiency.”
As the first order of business, Rüntü advises every entrepreneur to write down what data they have. “Are the data related to accounting or customer management?” “Once that is done, you should go into more detail and organise the data better, add classifiers,” he said, adding that all of this did not have to be done at the same time; rather, the data development should be seen as a constant evolution.
At that point, the data can be harnessed. “If conditions are favourable for analysing data, better business decisions can be made, for example, to assess whether a customer is getting away and make them a better offer,” said Rüntü.
Invest into people
When data are in good order and can be analysed easily, something you should think about is whether your staff have the capability to use them. “It’s not enough to acquire a program – your people have to have the analytical frame of mind and acumen to read the data, and the company’s executives should think about how to develop that ability.”
Peters noted that data analytics training shouldn’t be compulsory. “An environment where the employees themselves are excited to be working in this field and have a real curiosity about what lies behind the numbers in the datasets should be created.”
He added that definitions have to be put in place for the people performing the analytics. “For example, who is considered a ‘lost customer’ – one who hasn’t made a purchase in the last two months or in the last six months? Once issues like that are in place, the analysis will be more accurate.”