Is Data Science at the core of your business?
While it might have been acceptable for organisations to take the ‘big data’ movement lightly a few years ago, more and more evidence is emerging that data science needs to be put at the core of any business.
Here are just six mind-boggling stats on big data from this very comprehensive list.
- We have generated more data in the last two years than in the entire history of the human race
- By the year 2020, we will generate 1.7 megabytes of new data every second
- We perform 1.2 trillion searches a year on Google. This equates to 40,000 search queries every second, on Google alone
- In August 2015, over 1 billion people used Facebook in a single day
- 73% of organisations have already invested or plan to invest in ‘big data’ by 2016
- Only 0.5% of all data is ever analysed
(And if that interests you here’s a different set of similarly impressive stats by IBM).
So what’s the difference between Big Data and traditional analytics?
In an nutshell, traditional analytics analysis is concerned with extracting actionable insights from historical data. Data science, on the other hand, enables you to create machine learning models that will help predict a future outcome. So for an e-commerce retailer, this could mean finding the future value of a customer based on what he/she is expected to buy, or for a finance company it could mean detecting the likelihood of a fraudulent claim.
The other key difference between big data and analytics is the sheer scope of data the two technologies cover. While traditional analysis is heavily dependent on structured data, i.e. one stored in databases and spreadsheets. big data/data science also covers the vast spectrum of unstructured data ranging from emails and phone calls to texts, images, videos, social media data streaming, internet searches, GPS locations and computer logs.
It is this unstructured data that is growing at an exponential rate that we have never experienced before. The sheer volume of such data necessarily means that the traditional means of data storage is no longer viable. This is where big data and use of commercial hardware, in the form of Hadoop clusters, and a map/reduce engine will come into perspective.
What role is Data Science playing already?
Big data/data science is already part of our lives whether we like it or not, and below are a few existing applications to ponder;
Data Science in politics
The concept that big data or data science is here was epitomised when data science was brought into the mix of tactics by the ruthless characters in the fourth season of House of Cards.
The infamous Netflix series is one of the strongest statements about the role data science will play (or is playing) in the politics of the future. Interestingly, the Obama administration has already invested $200m in big data projects to understand and solve complex problems.
Data Science in financial services
The finance industry has been at the forefront of many computer technologies, so it is no surprise that, as an industry, it has reacted rapidly in adopting this changing paradigm in data analysis. A typical application of data science in banking is performing text analysis on structured and unstructured data (a common data science practice) from multiple sources and build predictive models that help detect fraud, financial crime or credit-worthiness.
Data Science in e-Commerce
A great example of the rising influence and implication of data science in e-commerce is the Amazon homepage, which also works as a recommendation engine. The e-commerce giant takes into account all a user’s buying history together with theur buying patterns, and the buying patterns of people that match their profile to come up with future purchasing suggestions. It is a marketing machine and it’s not surprising that other retailers are trying to emulate the methodology.
Data science in the travel industry
Companies like TripAdvisor have millions of reviews and other vital data sets that can be used to augment the user experience. Using existing data sets, travel companies have been creating machine learning models to detect fraudulent reviews or making recommendations for new and often undiscovered locations.
Data Science in digital marketing
While data science is being tested and applied in various industries, marketing (especially digital marketing) is one of the most promising applications.
It is no secret that digital marketers have historically relied on data to come up with digital strategies such as what keywords to target, what pages to promote, what messaging to use in the ad copy, what bloggers/vloggers to engage etc.
The key difference now is understanding the methods to integrate the structured historical data – available through various marketing technologies and tools, with ever-growing unstructured data such as tweets, reviews, likes, views etc, and come up with techniques to derive meaningful insight and plan for the future.
To get an idea of how marketing and data science can collaborate, this video from the Data Science Bootcamp in New York gives plenty of good examples of collaborations between the two.
Where should organisations begin?
The concept of using data to drive future business strategy is not new. But the challenge now is to determine the relationship between multiple data sources available and translate them into customer needs.
Historically the greatest challenge for businesses and brands has been to anticipate the future and proactively make the changes in the present. In that sense, adoption of data science as a core part of any business is a definite step in the right direction.
Here are some of the key points to consider.
- Data types: First, make sure you understand the various types of data that are important to your organisation.
- Knowing the right questions: For any successful analysis, the most important step is to have the right questions to be able to provide the right answers.
- Data storage: The velocity at which data is being generated will almost certainly make you consider whether you have to leverage cloud computing services. So include it in the broader plan
- Data management: Do you need to process data quickly? If yes, then you might want to evaluate in-memory solutions. Or, if you have a lot of data that needs to be processed in real-time, then streaming solutions are worth considering.
- Team: Do you invest in building a team in-house or hire an external specialist? Either way, building a team with both technical and business knowledge will be crucial.
- Tools: Initially, you will be working with low-level tools that typically involve a lot of programming. But as new tools will emerge, you should consider using them for better efficiency.
The possible applications of big data/data science methodology are endless. More and more businesses are discovering the need, and potential, of taking advantage of the vast data sets available to them from internal and external sources.
The technology is still maturing, and as more successful applications surface it is likely that key decision makers in businesses will use it to push the boundaries of their business strategies. If you want to keep ahead of the curve then your Chief Technology Officer, Marketing Directors or Data Officers should begin working together to plan the first steps of your business’s own big data/data science strategy.