April 27, 2020
Satisfying the Hunger for Data LinkedIn Logo

Satisfying the Hunger for Data

There are two main ways that businesses are currently tackling the GDPR and other data privacy and protection laws: either they are not sure how to comply and continue to process data, hiding among many other non-compliant organisations and industries. Alternatively, they do their best to comply, which often results in the application of a strict privacy framework to data governance, processing, and analytics. 

The application of this privacy framework keeps organisations compliant (or at least, attempts to). The problem is that business users then get data that is significantly trimmed down. A business user is anyone within a business that handles data to create value. This includes analysts, data engineers, marketers and campaign planners. This trimmed-down data leaves business users wanting more, and searching for solutions through which they can get their fill of the data they need. In addition to that, trimming down data creates several issues. First, without full data sets and the ability to process all the data, processing results can be of low utility; even worse, they may also be inaccurate, and in some cases biased. 

However, there is a solution that can unlock data utility to satisfy the hunger of business data users: a new technological approach that allows both needs to be met. 

Strict Privacy Frameworks Leave Data Users Hungry

Many companies are continuing to use data without proper controls in place. Business users who have always done things a certain way are not always on the radar of legal and compliance until something “new” is to be reviewed. Who can blame them when historically, businesses have been investing heavily in democratising data so that more people can derive value from corporate data assets. The magnitude of change that has occurred in many industries has left some organisations behind; not knowing how to comply leaves them unable to comply, and organisations continue to drive above the data privacy speed limit. We tackled this issue in our two-part piece on industry compliance and “speeding” in the AdTech industry. You can check out Part I here, and Part II here.

On the other hand, many companies see privacy compliance issues as a major potential business risk. In fact, the Accenture 2020 Privacy Study found that for 70% of their respondents  privacy is a “key material risk” for their firms, and that 1 out of three “lack a clear road map and the resources to address their residual privacy risks.” Attempts to reduce these risks result in the implementation of privacy frameworks that leave data users and businesses asking for more. 

Privacy programs that are currently being implemented typically apply three main levels of risk management:

  • Highest risk data: Delete - We can’t guarantee we will need this data.
  • Medium risk data: Anonymise - We might need to repurpose this data.
  • Low risk data with immediate value identified: Tokenise using static tokens - We are certain we will repurpose this data.

The result is that business users are being fed data from analysts who are starved of the full available data, due to the application of protections and restrictions. The unused data is known as “dark data,” and is data that is collected but never used, whether due to privacy controls, silos within the company, compliance issues, or lack of utility. Even for data that businesses want to keep for usability reasons, there are still problems with actually being able to access it.

In fact, Accenture reports that large amounts of data goes unused, or under-used: “only 32 percent of companies reported being able to realize tangible and measurable value from data.”

McKinsey also observed “that only a small fraction of the value that could be unlocked by advanced-analytics approaches has been unlocked—as little as 10 percent in some sectors.” The problem is that “if there is a gap between the current value derived from data and the potential value to be achieved from that data…  companies can neither improve their core businesses nor lay the groundwork for growth in future businesses.” Depending on the industry, huge amounts of dark data sits idle. One study on Internet of Things (IoT) data found that, for example, “only 1 percent of data from an oil rig with 30,000 sensors is examined.” Business users receive data that is trimmed down to the point where it is no longer useful for their desired goals. This huge difference between collected data and used data means that businesses are missing out on a huge amount of value.

Why Lean Data is an Issue

The scarcity of data coming through the pipeline is also a problem because investments in data analytics and data as an asset do have the capacity for big payoffs: McKinsey notes that there is the potential for firms to “use [data analytics] capabilities to achieve productivity gains of 6 to 8 percent, which translates into returns roughly doubling their investment within a decade.” 

Investment in analytics tools and business analytics services worldwide was projected to reach $101.9 billion US dollars in 2019, and Forbes has also noted that “65% of global enterprises plan to increase their analytics spending in 2020.” All of these new analytics technologies (including things like business intelligence platforms and tools used to interrogate data) are useful in theory, but many businesses are not getting the full ROI. 

In fact, in many cases analysts will start a new job and won’t have work to do for a long time, as no data is reaching them. One survey found that only “11% of front-line employees [are] getting access to analytics reports,” and that for “60% of employees, it takes hours or days to get the information they need, while only 3% can find information in seconds.” In addition, another survey found that of UK companies surveyed, “as many as 70% recently [stated] that their Big Data projects failed to realise Big Data’s full potential.”

The realisation of this potential for data remains elusive: in fact, a study by Deloitte LLP and the American Marketing Association found that the effectiveness of analytics in company-wide performance was rated as a 4.1 on a 7 point scale, where 1 was not effective, and 7 was very effective. Nonetheless, they also found that the amount “companies plan to allocate to analytics over the next three years will increase from 5.8% to 17.3%—a whopping 198% increase.” This shows that despite the current issues with gaining the full value of data, companies are still hopeful that gains will be realised.

Aside from revenue gains, in many industries, a company's data is one of the most powerful assets. Good data and accompanying analytics allow:

  • Development of high-quality products more efficiently
  • Personalisation for customers
  • The business to meet more-specific customer needs
  • Better strategic decisions
  • Improved customer and partner relationships
  • Faster innovation
  • Improved social media and marketing outreach
  • Product safety checking and troubleshooting

A Deloitte survey found that “49 percent of respondents said analytics helps them make better decisions, 16 percent say that it better enables key strategic initiatives, and 10 percent say it helps them improve relationships with both customers and business partners.” 

On the flipside, without complete data, processing can produce inaccurate and sometimes biased results. Some bias in data is inevitable, especially because those collecting and processing the data are unable to remove all subjective bias from any data set. However, biases should be reduced wherever possible, as they can negatively impact both your customers and your bottom line. Especially when an incomplete data set is used for machine learning or algorithm training, biases can become amplified. If you then analyse this incomplete data, your results are inaccurate at best.

When it comes to using data as a strategic asset, inaccuracies from incomplete data can result in simply poor business decisions. Resources may be allocated incorrectly or in a place that is not efficient, such as to a market or customer segment that is actually not the best target for a product or service.

Keeping the Data Plate Full

A light on the horizon does exist. There are ways that the data plate can be kept full, so that data users don’t need to go hungry. Anonos BigPrivacy allows you to unlock data utility, while satisfying GDPR and other privacy legislation requirements for protecting data subject privacy. This makes sure that your analysts can process and protect data that feeds data users, not starves them. The following diagram shows how Anonos Variant Twins enable lawful data use: 

Image describing how BigPrivacy works. data set is loaded via an ETL tool, schema is recognised, you then configure the privacy actions and techniques you want to use for each column, the data is then transformed. K-anonymity is used for re-identification risk management and the final context, use case specific dataset is created. Dynamic pseudonymisation, risk mitigation and anonymisation all combined.

Anonos BigPrivacy unlocks data to provide 99.9% accuracy and 100% lawful repurposing, so that your data users can obtain and process all relevant data for business strategy, product improvement, and faster innovation. 

Without this approach to data protection and use, data users will always be left wanting more, and both businesses and customers suffer as a result. Solutions like this can help to ensure that data analytics and business intelligence tools are not needlessly invested in, and business users can actually get what they need in terms of data on a day-to-day basis. In addition, data can be kept accurate, ensuring that efficiencies are maximised, and biases are reduced. With the hunger for data ever increasing, the need for solutions exists more than ever. 

This article originally appeared in LinkedIn. All trademarks are the property of their respective owners. All rights reserved by the respective owners.