To Bavasso the phrase ‘data has a better idea’ means when data is used properly it can be converted into insights which, in turn, can be converted into an automated process. The automated process will then challenge the way in which others in the company perceive problems or formulate solutions.
“We need to let the data speak.”
Bavasso, who is the director of big data – digital, said: “We need to let the data speak. We need to have that mindset in which we embed it into our strategy. In theory, a company shouldn’t have a data strategy. Data should be in the strategy of every business function in the company.”
Having data in the strategy of every business function is the ideal destination but he admitted that the journey towards it may not be a short or an easy one. To get there, a company would need to change its systems and structures, culture and way of working, and team operating model.
He set out the order in which a company should go about it. Defining the strategy is the first step, in which there are two components; defence and offence.
The defence is enabled by a single source of the truth, and is aimed at control. The key objectives are to ensure data security, privacy, integrity, quality, regulatory compliance and governance, and the core objectives are to optimise data extraction, standardisation, storage and access.
In contrast, offence is focused on flexibility and is enabled by multiple versions of the truth. The key objectives are to improve the competitive position and profitability, and is characterised by optimising data analytics, modelling, visualisation, transformation, and enrichment.
In Bavasso’s view, it is imperative that offence and defence take place at the same time. He said: “If you do them in a sequential way that will never be perfect because it is continuous evolution and you will constantly block the innovation. It is a dual strategy and it has to be done in parallel.”
After setting out the defensive and offensive blueprint, the business decision makers should then seek proof of value, improve the data governance, scale up and from that point they will start to see the returns.
He pointed out that this advice is particularly pertinent to companies that create data products. There are three types of such companies and Bavasso gave examples of each. The first, like Flightradar24, monetises data as a service. The second builds enhanced data products, like Netflix. The third uses data to get valuable insights, as Javelin Group does.
One person, the data product owner, should have a product vision, evolve the product and oversee its delivery. That person also needs to have a strong data background, understand the data needs and know where to source as well as what is needed to develop the product. For this person, it is imperative to have access to and oversight of data.
“At the time of product vision, know what data you need.”
Bavasso said: “If you don’t know where the data is and you don’t have it, you don’t develop anything. And access to good data is difficult so it is important that at the time of product vision or development process, you know what you need and where to get it.” If not, they could get stuck in the concept phase.
He also said that the data product owner also should maintain alignment among all the products being developed, deliver exemplary user experiences with data and develop the input of other people they are working with.
With all of these changes and transformations taking place, the company should never forget the importance of consent and its obligations to its users. Bavasso said: “It has to be very, very regulated by a very good consent ecosystem. You have to have the right consent to do anything and all the anonymisation and GDPR compliance has to be kept in the whole ecosystem.”
“We need to let the data speak,” Bavasso said in conclusion. “We need to have that mindset which we embed in our strategy.”
Lorenzo Bavasso was speaking at DataIQ Summit in June 2019.
