Blog

Learn to run with big data – how to move beyond baby steps

learn-to-run-with-big-data

One of the consequences of the hype surrounding big data analytics over the past couple of years is that companies can no longer afford to ignore the technology (little yellow) elephant in the room. With understanding growing about the wide range of benefits this can bring to an organisation, many executives are now worrying that if they do not get on board with and embrace the innovations, they will be left behind by more adventurous and technologically advanced competitors.

Indeed, figures from Gartner suggest that almost three-quarters of businesses have plans to invest in big data by 2016. However, the firm industry analysts found that while interest in big data analytics is growing, there is still a lot of work to do before many firms organisations can consider it as an integral, and valued, part of their operations.

According to Gartner, just 13 per cent of deployments in 2014 had reached the productions stage. While this was a considerable increase over the previous year, it suggested many professionals are still looking to gain understanding of the technology before going all-in on a full enterprise-wide deployment.

So what can firms businesses do to ensure the leap from experimentation and pilot schemes is a success?

Understand the challenges

One of the key things firms organisations need to recognise about big data when looking to make the step up to full production uses is that ‘big data’ is much more than just a simple technology implementation, and is not a one- size- fits- all solution. It’s important for businesses to know what their needs are and how they will be met, so the first priority for any project should be a full audit internal workshops, asking what problems need to be solved and where resources and effort should be directed.

There are essentially two elements to any big data programme – the business side and the technology requirements. A pilot scheme will often seek to address one or two specific technology needs and prove the concept works, but firms organisations need to ask how this will translate into a wider business benefit.

Other key questions surround issues such as how firms a business will gather and secure data at the enterprise scale. Firms Organisations often find that when they take the step up from smaller projects, the amount of data available rapidly becomes overwhelming. Therefore, it’s vital companies that the business are is prepared for this and have has a strategy in place to cope with the explosion in information data and its use.

Take the right lessons

When evaluating initial pilot projects, firms an organisation are is likely to identify a range of issues that will need to be addressed if the technology and process is to be scaled up throughout within the business. But all too often, firms end up taking the wrong messages from these results and so carry poor practices into a wider deployment.

For instance, one mistake is not to fully explore consider the potential use cases for the technology. While many businesses will initially focus on common requirements such as boosting marketing campaigns or detecting fraud, this is often just the tip of the big data information iceberg.; But by continuing to focus solely on one or two uses, they can end up building a system that has limitations and is difficult to adapt.

Therefore, it’s vital to design and build an agile system environment that can cater for different changing scenarios. A pilot stage should inform the businesses as to what they will need to achieve this, in terms of data acquisition, infrastructure resources and staff skills. This is the stage when it should become apparent if there are any gaps in an firm’s organisation’s knowledge or capability, so there needs can’t be filled before expanding to full-scale deployments.

Bring it to everyone

While initial trials may be restricted to one team or department, one of big data’s biggest advantages is that its can have a positive impact on all a firm’s business’s operations – provided companies are able to deliver a platform to provide insights to the people that need them.

This means breaking down traditional silos of information, which may prevent a department from accessing key data that’s stored elsewhere in the business. Because pilot schemes are often run in isolation from the rest of the business, firms may think it will be simpler to persist with this approach when they scale up – but this could be a serious error that leaves users without full visibility of their data.

It’s also important that tools and interfaces are simple enough to be used by non-experts. When rolling out big data initiatives across a company, it is inevitable they will come into contact with people without specialist analytics or data science knowledge – but in order to make the most of the technology, these staff will still be expected to be able to ask questions and interpret the results. So ensuring that investment in extra skills and training requirements are kept to a minimum will also be essential for a progressive workplace.

There’s no one route to a successful big data implementation, but by maintaining a clear idea of your long-term goals, keeping a broad vision view and investing in the resources required, a businesses will be well-placed to enjoy many years of success with their tools riding the big data wave.