The problem with data analytics and what 2015 holds Posted on : Mar 02 - 2015

The world of data analytics has experienced huge change in recent years, and 2015 is the year that may see the most revolutionary changes yet.

But in order to make those changes effective and useful, business leaders need to confront an urgent problem: for most organisations, data analytics is too slow.

According to Alteryx research, 72 per cent of business leaders aren’t satisfied with how long it takes to derive the insights they need from data. Nine out of every ten of those dissatisfied point the finger of blame at the inability to efficiently combine data from multiple sources.

Effective data blending is, therefore, crucial to business analytics, but analysts are chronically underserved by the tools and processes they currently use. Many analysts have become a bottleneck as they spend their time blending together data from multiple sources, and preparing rather than analysing it.

Rapid blending of information from data centres, cloud apps, and even Excel spreadsheets, is only possible when managers provide analysts with the proper tools and processes. They can then deliver insights in minutes or hours, rather than days or even weeks. This will unleash a company’s analysts to provide responsive analysis and visualisation while reducing the burden on its technical team.  This in turn can unlock business possibilities that would otherwise have lain unnoticed. 2015 will be the year that agile companies realise this.

But this year is also set to see other changes in the analytics space. Below are just some of the changes we’re set to see over the coming months…

Predictive modelling that relies on coding will become a minority sport

Gone are the days when data scientists and technical experts were the only people who could interact with and analyse data.

We’re at a tipping point where analysis and predictive modelling are being taken back from the IT department and given to users who truly understand the business’s questions and goals. Previously, preparing and analysing data meant spending hours writing line after line of code, churning out reams of SQL.

The latest generation of tools out there embrace usability as one of their most important features. They aim to make impactful analysis possible for non-technical people, integrating drag-and-drop user interfaces, amongst other user-friendly features. By reducing the need for coding – and therefore a technical expert – analysts can interrogate information directly and more swiftly. In the coming year, we’ll see coding-based predictive modelling used less and less frequently.

Everyone will become an analyst

Not only will analytics transfer from technical experts to line-of-business analysts, it will make its way even further out into the open: 2015 will be the year of the data hobbyist.

These people work with data because they love to. They’re naturally curious and thrive on discovering new insights. They may be analysts, but they could be from any part of an organisation. They may spend their spare time analysing data on Premier League clubs, and this should be encouraged too:

They’re training themselves in analytics skills that their company can benefit from, and their enthusiasm can be a vital asset in encouraging wider adoption of data analysis tools and practices. Regardless of exactly how they’re practising their data hobby, the factor that unites them is their love of working with data.

Data blending and data discovery take the sting out of Big Data analytics

Understanding the main pain points around Big Data is not difficult. First and foremost, it’s a matter of volume and scale. Huge quantities of data mean more storage is required, more powerful traffic management tools are needed, more cable bandwidth, more, more, more.

And this obviously stretches to the data analytics. More data to process and analyse means that analytics tools have to be faster and more efficient. Speed is of the essence, and they need to be returning insights within minutes, not days.

If tools can cope with faster, near real-time data blending and analytics, then one of the major concerns of Big Data is taken care of. Quicker analysis not only means that more data can be processed but also that analysts have more time to explore their data and deliver business insights.

The cycle can be ongoing: a business query arrives; relevant data is blended from multiple sources; patterns and insights are discovered and filtered; analysis takes place; and it all starts again.

More questions can then be asked of the data, meaning additional insights that might not have been possible before, ultimately resulting in better informed business decisions. Source