What are the problems organisations are facing today when dealing with data?
“Data is the new currency in today’s digital economy, but data economics are broken. Data volumes continue to double each year, while the number of data workers triples, yet IT budgets remain flat—making it impossible to keep pace with growing business demands for data processing and analytics. In parallel, it’s becoming increasingly common for disparate business units to use their own tools to collect and integrate data, leading to more information silos, or ‘shadow IT’, which ultimately prevents companies from reaching their data-driven potential.
“The amount of data that companies are generating today has no precedent. Companies can derive substantial benefits from big data analysis. On the whole, big data brings many benefits, but it does come with challenges. In many cases, there is a divide between business requirements and IT departments with both teams using different tools and processes and a lack of collaboration between the people preparing, processing and analysing the data, and the people providing the necessary software platforms to do so. This leads to companies spending too long to put data to work for the business as processes are not clear.”
How can businesses improve their access to insights?
“In order to get actionable insights, businesses need to ensure they are keeping up and replacing old, obsolete data with up-to-date, accurate data as fast as possible. With data informing every aspect of business functions – from marketing through to strategy – ensuring data is of a high enough quality for analysis is paramount. Building a platform that’s adaptable to change from the beginning is fundamental to this. Too often organisations focus on the database selection, cloud vendor, Hadoop vendor, whereas these all have the potential to lock the data into a new silo. CIO should spend much more time consider the data infrastructure. Invest in cloud and build a live dashboard of the ROI of your data. This should include a daily cost of the data ingestion, the processing and then who’s using it accessing. Data is a business, not an IT cost centre.
“Cloud-based, big data streaming technologies play a key role in helping companies use the cloud to scale data-driven decision making, while ensuring their most valuable asset—data—is governed, trusted, managed and accessible to a variety of employees. With this technology, companies can get up in running in minutes, ingest large volumes of data and simplifying data integration process significantly.”
What lessons can CIOs learn about turning unstructured and inaccessible data into business intelligence?
“There are two key tenets of enabling real-time analytics that incorporate unstructured and inaccessible data: artificial intelligence (AI) and human intelligence. Augmented intelligence - the combination of artificial and human intelligence - is complementary to human intelligence. This mixture of human and AI can benefit businesses if used correctly. It’s about helping humans become faster and smarter at the tasks they’re performing as well as finding faster and more intelligent ways of analysing data.
“Analyst firm Gartner predicts that one in five workers will have AI as their co-worker in 2022. Therefore, it is vital to deploy both AI and human brainpower appropriately. The end result will be more time for employees to focus on tasks which require human rather than machine-based intelligence, such as critical thinking, and creativity.
“AI is particularly useful when it comes to structured data that has defined parameters. This frees up time for data analysts and scientists to tackle unstructured data analysis and define and implement processes to make inaccessible data accessible.
“Companies can leverage machine learning and advanced analytics to guide users in their data journey by suggesting next best actions. This improves developers’ productivity and empowers non-data experts to turn data into a team sport, further engaging line of business decision makers with data analysis and insights. Machine learning also allows companies capture tacit knowledge from business users and data professionals.
“To facilitate this, cloud and data integration tools deployed across the entire data ecosystem, pulling in data from disparate sources, helping to sort it, clean it, and prepare it for analysis are essential. We see cloud based systems as the most appropriate as these can give the right employees access when they need it, and have the capacity to handle the ever-increasing workloads of modern data driven businesses.”
What are the potential pitfalls they should look to avoid?
“There are certain instances when data analysts need to know they are fighting a losing battle. For example, if it takes too long to access distributed data for the insight to be gleaned in a timely fashion, data analysts may want to reassess their strategy to determine whether obtaining this data for analysis is essential or not.
“Similarly, businesses need to ensure they are not holding onto gathered data for too long. At present, 55% of a company’s data is not accessible meaning that an analysts spends more than 80% of their time preparing the data. Not only is this a waste of time and money for a company, under the General Data Protection Regulation (GDPR) which came into effect on 25th May 2018, businesses need to use data in a responsible fashion to justify holding it at all.
“One opportunity which requires some structural and cultural changes towards data management within an organisation is moving from big data to ‘smart data’. This means creating a repetitive, predictable and sustainable process for turning raw, unstructured data into critical insight for the business.”
How do you balance technological investment with human capital to best leverage data insight?
“It’s all about planning, measuring, evaluating and replanning in this scenario. Advances in cloud-based data analytics tools are helping data analysts and businesses increase their productivity, but in some instances, the human element is irreplaceable.
“Our advice to CIOs seeking to implement a robust data analytics programme that combines disparate, diverse data sources is to consider the skills of their human team and how software can complement them in their planning. Once a plan has been successfully implemented, there should be continuous measurement and evaluation of inputs, processes and outputs to ascertain what can be improved and where. While this sounds like a lot of work, the benefits of cloud-based analytics solutions, the enhancements of algorithm based analytics, and the advancements in AI, mean that it is now possible to create a data analytics programme that is constantly evolving to the improve itself and meet the needs of the business.”
Jean-Michel Franco, Senior Director, Data Governance Solutions at Talend
Image Alexskopje / Shutterstock