Social media firms need to pro-actively manage user expectations: Bacuvier

Guillaume Bacuvier, Global CEO, dunnhumby
With the recent developments around Facebook and violations of data breach, what are the checks and balances that social media and analytics companies need to serve users and clients without compromising data security?

Good privacy and security practices begin with processes and understanding. You need to make sure you get the basics right which involves knowing the following: what data you use and the risks it presents, where it is held, what purposes it is used for (and having clear governance over changes to those purposes), who has access to it internally and externally, who your suppliers are and how good their security measures are, and knowing where you want to go as a business in the near future.

The culture of the company is also very important. The right culture is instilled through communicating with employees right up to board level regularly about privacy, security and the associated ethical issues, so they understand the risks and their role in mitigating them. If you have a good grip on the basics and the right culture, you can then build more detailed controls and processes, including information security ones, around the actual demands of your business. You will also have confidence that those controls and processes are proportionate and effective. Getting the balance right is very important. You have to guard against rash or short-sighted commercial decisions and other mistakes, but also have to act in a proportionate manner that still allows business to flourish. For consumer businesses like social media companies, there is also the need to be transparent with your users about how their data is being handled and for what purposes. You need to think through how your actions could be perceived or misconstrued and pro-actively manage their expectations.

What are the major areas and decision making cases for which Indian executives are turning to analytics?

A growth in high-speed internet availability, high smartphone penetration, acceptance to e-commerce, a greater usage of social media and a good local talent pool have led to an increase in the data being generated which is critical for organisations. Data-driven decisions are now finding use-cases within different business functions in an organisation — they span from business planning to optimisation, workforce management to recruiting, customer relationship management solutions to acquiring customers, managing supply-chains efficiently to forecasting inventory and retrospection of previous year’s business drivers to forecasting revenues for the next few years.

dunnhumby specialises in the 4 Vs of big data—velocity, volume, veracity and variety. How are you helping clients solve issues leveraging all the elements?

The 4 Vs can be fully leveraged if we have the right computing power and the data sources for solving the relevant problem. In today’s world where the majority of our clients are active in both online and offline space, it’s imperative for us to source various data types/channels to get a good view of their needs and service them based on those needs. The better you know your customer, the better you can serve them, which reflects in sales.

Real-time personalisation uses the 4 Vs of big data to revolutionise the retail marketplace by reaching customers with personalised messages at the very moment they are making a decision. Media personalisation in the digital space is no longer a “good to have” but a “must have”. Our data scientists apply sophisticated techniques to help brands and clients identify the right set of audiences based on business objectives and the type of offers, which help in either driving additional product purchase or incremental shopping trips. Besides big data, a key enabler for achieving individual customer level real-time personalisation is powerful dunnhumby technology that can quickly process data and instantly apply solid relevancy science.

What are the key innovations that dunnhumby’s Centres of Excellence (CoE) in India currently working on to solve customer problems through analytics?

Our CoEs based out of Gurugram is the only dunnhumby office that serves all our clients around the world. It is at the forefront of developing best of science innovation using machine learning and artificial intelligence to predict which customers will buy what and when, through which channel and at what price-point. 

Through an understanding of deeper customer behaviour and use of advanced analytical techniques, the data scientists help retailers from across the world take data-driven decisions and use their marketing budgets efficiently across multiple marketing levers. The CoEs have been the leaders in leveraging new technologies to support dunnhumby’s strategy of cloud and open source first environment.

What are the distinct features of dunnhumby’s Technology Renewal programme that takes a cloud-first approach given technology giants like Adobe, Accenture and others have specialised cloud offerings?

Technology Renewal is a multi-year, multi-discipline programme adopting a cloud first, open source approach to our foundational technologies, bringing greater freedom, flexibility and speed to dunnhumby and its clients. The Technology Renewal program is a necessary investment which ensures that we continue being at the cutting edge and able to quickly react to these global trends. It will provide flexibility, increased agility and reduced pace to serve our clients with the most advanced analytical methods.

India teams are key to this programme and have the advantage of having a global perspective. This makes it is easier to share and learn about challenges across markets and build scalable and common solutions, for multiple clients. India also provides much-needed talent for execution as well as for training others.  With a large concentrated pool of great talent, India teams continue to innovate and challenge each other through activities such as hackathons to ensure that the final solutions are brilliant and best-in-class.