Big Data V’s what are they & how can you use them to formulate a strategy?
Anyone who has ever done any kind of business related course will have heard of the 4 (or 7) p’s of marketing. Product, Place, Price, Promotion. It is a framework to ensure that when you are planning your business proposition, growth or development you are covering off all the key bases your proposition will need to succeed. Misunderstand just one of these and your business can suffer big losses.
When it comes to Big Data, the 5 v’s work in the same way. Often deployed to explain Big Data in general, the 5 V’s model is really useful as a strategy framework to review the data you have access to, how you want use it and formulate your Big Data project objectives.
How much of it is there? This is probably the most defining element of big data… it’s why it’s BIG data after all… the stats on the internet vary but most agree that 90% of the world’s data have been generated in the last 2 years. According to an article on techcrunch.com “We — humanity, that is — created 4.4 zettabytes of data last year. This is expected to rise to 44 zettabytes by 2020.” To give some context, if one byte = one grain of rice, a zettabyte = enough rice to fill the Pacific Ocean. The volume of Big Data is why there are so many tools now built to help you pull together and process it all.
Action: Review your potential data sources, select those that are relevant to your business (while remaining open to the idea that new sources of data become available all the time) and identify how far back you want to go in terms of collection and collation. This will give you an idea as to how much data you’ll need to manage as a part of your project, very important when determining the tools and software you’ll need.
With so much data to consider, it won’t surprise you to find out that another key definition of Big Data is the speed at which it is generated. Transactions, messages, likes, comments, locations, check ins – all happen in the blink of an eye, and all can be processed and used to make decisions at a similar speed. If your app on your customers phone tells you that they are about to walk past your store you might want to send them a message with an offer on something they are likely to be interested in based on their past purchase history… this is no good if you process all the data at the end of the day for analysis. You need to be able to constantly process the information and make decisions on the fly (in real time). Of course this goes back to the data you have and the things you want to do with it.
Action: With the sources of data you’ve elected for inclusion, review their veracity and decide how quickly you want to be able to make business decisions based on the insight they’ll provide.
Key to the challenges around processing Big Data into valuable and usable insights is the variety of formats that data comes in. Technological developments now mean that data can be pulled from unstructured data formats (PDFs, blog posts, website behaviour, comments, tweets, images – the list goes on) and combined with more traditional, structured data sources (transactions, addresses, purchase history, campaign history etc) to generate actionable insights. The type of unstructured data you want to include in your Big Data strategy and processing tools will impact which tools you use and how you process the data.
Action: What format is the data you want to process in? How complex does this make the project? Assign complexity scores to each source of data.
Is it true? How much do you value the source of this data? When you open your data collection nets out to the wider web, you need to consider just how much reliance you’re willing to place on the data you’re collecting. Social media, while a HUGE source of Big Data is fraught with challenges – spelling, colloquialisms, personal opinion & half-truths. Depending on what you’re trying to understand this can be useful or not…
Action: When pulling in data sources a veracity score can be applied to weight your data sources depending on their reliability, that way you can be confident that the decisions you make are based on more solid data sources. If you have less reliable data sources in your mix be clear on why you want them there and what benefit they can add (brand sentiment analysis on Social media for example).
Every data project, Big or small should always come back to this final, and in my opinion, most important V. If you’re not going to get business value out of the data project then why are you doing it? With so much hype surrounding big data it’s understandable that people think it’s where they should be directing their efforts, but Little Data can have a much more significant impact on a business’s bottom line. If you don’t have your Little Data (internal, structured data) in a solid and usable format where you’re regularly gaining and acting upon insights, starting out on a Big Data project is probably not the best use of resources. If you’re here and ready to move on then great!
Action: Review all your data sources and establish what data they’ll provide and what analysis you will be able to do with it. How will you use the data? What insights do you think you’ll get? How will it improve your decision making? What actions will it drive? Use this to assign a value score and combine these with all the other conclusions about each data source.
The size, scope, technology choice and cost of your project will depend on these questions so being thorough and clear about the business benefits of each data source will make sure that you’re driving value from your Big Data strategy.
We’ve just finished the initial stages of a Big Data project for Qantas – you can read more about that here, or get in touch to chat about how Big Data could impact your business.