January is a time for reflection and resolutions. There was no shortage of data stories
making their mark last year – from Artificial Intelligence and blockchain to data breaches, with a healthy dose of machine learning and Internet of Things (IoT) in between.
Beyond the headlines, IT managers have been confronted with an alphabet soup of regulatory requirements and, with GDPR on the horizon, the need to unearth patterns and establish links across volumes of data has never been greater.
With this in mind, here are some of the important trends CIOs and IT Managers should have on their checklist as they shape their data strategies for 2017:
Plan for the machine data deluge
Although forecasts for IoT vary wildly, IT managers need to plan for a machine data deluge. I don’t believe we’ll see much of an impact on businesses in the immediate short term, but IT managers need to develop a strategy for managing the colossal amounts of data from IoT, industrial IoT, machine learning and streaming media. Fortunately not all of this data has to persist so understanding what data to keep and what to let go becomes important when managing storage costs.
Turning insight into actions
When a business integrates data from all customer touch points and channels, it creates a 360-degree view that it can use to enhance the customer experience by predicting behaviour, for example in online shopping or TV viewing.
I believe we will start to see tech-savvy companies taking this a step further and acting on this data to increase revenue and customer retention, reduce costs. For example, if you are a frequent flier and you fail to complete a complex flight purchase online, the airline’s call centre will be alerted to text or call you immediately to help resolve any issue. Or, by analysing publicly available figures on the mean time between failure rates for car parts, for example, your garage will be in a strong position to suggest changing your exhaust manifold while it is repairing your catalytic converter.
Show me the NoSQL money?
As VC investors move from the maturing NoSQL market to earlier stage technologies, we will see consolidation in the NoSQL sector accelerate with up to 50% of the dozens of NoSQL vendors disappear in 2017 as a result of having burned through their money. That comes on top of the roughly 10% of NoSQL vendors who hit the dust in 2016.
Open source will lose ground at the executive level, particularly in light of recent security breaches. Open source database vendors will continue to move further along the proprietary, commercial path in 2017 as they charge customers ‘subscription fees’ (aka a software licence) for proprietary add-ons and utilities, for example. This will impact vendor credibility. This move has already happened in the Hadoop space, where customers, become locked in and face substantial costs if they wanted to replace the software with another vendor’s Hadoop stack.
Data wrangling gives way to data integration
Many businesses are failing in their data governance efforts. Some don’t know the provenance of their data. Meanwhile, data scientists typically spend far too much time collecting and preparing unruly data – in some instances up to 80 per cent of their time – because of the inflexibility of relational databases. Businesses are getting fed up with this waste.
If organisations can’t answer key questions about their data – under what agreements was it collected or which pieces are personal information – then they risk significant exposure to regulatory fines, brand damage, leaked customer or employee information, or lost intellectual property.
By untangling the knots of data currently segregated across numerous silos throughout their organisations, companies can profit from getting their data in better shape. In the coming year, we will see businesses spending less time and effort on data wrangling – extract, transform, load (ETL) processes and an upswing in the use of applications and platforms that support all data formats and data sources and offer better data integration capabilities.
Open data is a long slow burn
Open data is taking a long time to catch on, with many organisations unsure of the value of going down this route. Caution over the potential legal implications and security issues may also be slowing things down. As more open data is published in 2017, those businesses that have already deployed a flexible database and are able to integrate their internal silos of data will be able to steal a march on everyone else because they can more easily integrate third-party data to quickly gain new insights.
The rise of the multi-model database
UK organisations will start to standardise on a multi-model database as NoSQL goes mainstream. Now the confusion over Hadoop is behind us and NoSQL is becoming mainstream, we are seeing a concerted drive by organisations in the US to consolidate on one multi-model NoSQL database. As a result, we will see more non-relational database vendors rushing to become multi-model. This post-RDBMS wave will cross the Atlantic and hit British shores over the next 6-12 months.
Conventional wisdom states that sticking to resolutions is easier when you change one thing at a time. A company’s data is a priceless asset but only when you make your data work for you, not the other way round.