Ten tips for driving value from big data

Wednesday, 17 February 2016 00:00 -     - {{hitsCtrl.values.hits}}

By Vicky Falconer

Just think what $ 48 billion could buy. In the private sector, that could buy a lot of R&D and innovation; the lifeblood of a successful and growing economy. In the public sector, think of the boost it could give to education, healthcare and defence. What is that figure and why is it relevant? DFT-12-5

According to a new report from PricewaterhouseCoopers (PwC) that is the sum of money the Australian economy left on the table last year and wasted, due to its inability to fully leverage the potential of data-driven innovation. $ 48 billion! That’s the equivalent of 4.4% of gross domestic product and about the same as the entire (and struggling) Australian retail industry.

Little wonder then that big data is top of the CIO priority list on almost every report you read. So why then do people still question the value of big data? 

The challenge is that, while the principle of getting value out of all the data that is now available in the world in order to gain new insight and so better serve customers and find new markets seems simple, actually doing so is actually hard in practice. 

So just how do you realise commercial value from your data and not end up on a hunt for fools’ gold?

1.Realise you’re on a journey – When you begin to seek value from big data you will need a healthy degree of realism; it is hard to get to a revenue stream right away. More likely it will be a journey, starting with projects that add value to the existing business before arriving at opportunities that create whole new revenue streams. A good starting point is to leverage how you already make money. Consider your current revenue streams and consider how you could approach them differently to add new value and what insights will be required to achieve that. 

2.Do you have the right team – It is common to get caught up in the excitement of big ideas, but are you ready for it? It’s not to say the concept won’t become reality, but in many instances, it may become apparent that you are actually six to twelve months away from it becoming viable. From experience, those companies that have gone at projects too early have failed to realise the value, and then stalled in their big data activities as a whole due to the failure, or have given the idea away in their attempts. A key element of being ready is having the right team in place; that is tenacious, will persist in the face of various obstacles, brave enough to take the initiative, and inventive being able to create new value from data. 

3.Selling your data could be like selling your soul – As big data skills are in short supply some companies decide to sell their data and achieve revenue from it that way. Unless you understand how you could monetise your data yourself, you are at risk of commoditising your own information. Before you go this route, focus on figuring out how your data can augment your unique value proposition, and don’t give up on creating new value from your data just because the first experiment doesn’t work. Also be aware of the potential risks of selling your data or the insights into it. These include issues around the security of the data once it is no longer under your control, the chance of re-identification of anonymised information, and the potential impact on your reputation if it is used for unintended purposes. 

4.Is there a wider ecosystem you could be part of? There are, however, increasing instances of companies and industries collaborating around, or selling data. For example, in the pharmaceutical industry we have seen organisations working as a consortium around the creation of new data sets from clinical research, as a way to overcome the prohibitive cost they would have faced doing it alone. So the advice here would be not to rule options out – especially if they might enable you to do things with data that you couldn’t do before, and as a result, move up the value chain or closer to the end consumer. 

5.Don’t chase fools’ gold – use data, and especially social data, wisely. Whilst providing great insight into the digital DNA of customer decision making, developing accurate models for sentiment analysis is hard, due to the large amount of false positives that exist. The nature of social means that many companies, at any time of the day or night, can have somebody saying something negative about them. How do you know when is this out of the norm? This level of understanding is something that often develops over time, and is an enrichment and maturity process in your analytics. Get is right and you can make money by developing an understanding of the soft signals, but unless you have a historic wealth of data in that area, gaining that sort of intelligence is hard to come by.

6.Understand the customer contract – You need to know what your trust relationship is with your audience. At the advent of the concept of 1-1 marketing, the customer understood the idea that by giving a bit more information they could get a more personalised treatment – and they didn’t mind it. However, with the arrival of big data, some customers are feeling that companies are going too far in collecting and using intimate details gathered without their prior permission. It also depends on the company. Customers expect the likes of Google and Shazam to use data to make recommendations. You just need to understand where the line needs to be drawn between where and when it is beneficial to use data to add value to customers and going too far. 

7.Realise the goal posts are always shifting – change is inevitable and rapid. Your key competitors today may well be superseded by companies you never dreamt would fit in your market. However, it’s not all about the big guys. The Internet of Things is going to be a great leveller, particularly in the field of ‘controls’ which is enabling new, smaller players to nip in and seize customer data and ownership. This is giving them the power to disintermediate traditional providers. 

8.Be prepared to unlearn – in some cases you might find that the data shows that your assumptions are incorrect or that your activity is not the success you hoped it would be. For example, many companies are seeking to make money from content, but increasingly analysis is showing that it is a highly crowded market and there is little money in it other than for the likes of Google and Facebook. So companies need to understand what the implication is for their business. Does content add value to your customers, is it expected by your customers, or rather than there being money in the content itself, is there value that can be derived from better understanding the digital signature of your end users; in being able to see what caused someone to click on an ad or what sort of people are visiting your website. 

9.Don’t confuse perfection with monetisation – This is extremely important. Particularly when programmers and IT people talk about data, they often talk about perfection because we are very deterministic. We want to say ‘if this, then that’; if you get this data, then you will achieve exactly that result. The challenge is that it is hard to achieve perfection. Consider the case where you find you can cut the cost of a process by 30 percent and in a relatively short time frame. Or you might have the potential to cut your costs by double that but it would most likely take you several years. Is the wait worth the lost opportunity cost? In a big data world, the best practice approach would be to experiment – try something and iteratively improve it instead of trying to get perfection out of the gate.

10.Remember, David doesn’t always beat Goliath – while new entrants can outmanoeuvre established organisations, most often David doesn’t beat Goliath. In fact, as often, the incumbent can use data to create barriers to entry, due to the fact that they have a significant advantage from the large volumes of historic customer information and transactional data they hold. However, to realise the benefit, established operations need to digitise or datify all of this information before their rivals do, and potentially seek out new data streams to compliment what they already have. For new entrants, many of the key business opportunities exist where there is a breakdown in a process, or supply chain, in an area that really matters to the customer – think Uber or Airbnb. Look for what is ‘broken’, in areas that are not already heavily digitised. 

The reality is that there is a whole range of data out there, offering new ways to get insights, drive value and compete. It is essential that you understand the potential and get excited about the opportunity. So think really broadly about the data that is out there, both inside and outside of the organisation, see what there is that could add value, without ending up on a hunt for fool’s gold.

(The writer is Big Data Specialist at Oracle Australia and New Zealand).

COMMENTS