As it is the end of the year I’ve decided to devote this blog post to some thoughts on where we see the analytics market going in 2018.
Unlocking the analytics potential of unstructured data
Enterprises are collecting ever greater quantities of unstructured data. A recent survey by Forbes showed that the number of organisations with over 100 terabytes of unstructured data has doubled in the last two years. However the same survey showed that only 32% of organisations have successfully analysed unstructured data. We see this changing in 2018 as the power and sophistication of text analytics platforms develops to be able to really handle unstructured data effectively. Older generation text analytics platforms can be complex and unintuitive, and analysts with the skills to get value from them are thin on the ground. This will change and as text analytics platforms become more accessible so organisations will begin being able to really unlock the value in their unstructured data.
Move over data scientists, time to make way for the data engineers
Judging by changing trends in job postings it looks as though the ubiquitous data scientist may be in decline, and we’re seeing more and more organisations advertising for data engineers. A data engineer is, at heart, a software engineer who designs, builds and integrates data from various different sources, ensuring that it can be accessed easily and works effectively. This change has come about because organisations have started to appreciate the truth of the old maxim, ‘garbage in, garbage out’. This means that all the sophisticated models in the world won’t help you if your data quality is poor. The shift in focus from data science to data engineering recognises this and puts the emphasis more at the start of the analytics process, acknowledging the importance of developing effective data architecture and ensuring this architecture will support the needs of the organisation.
More analytics outsourcing and growth of insights-as-a-service
Forrester is predicting that up to 80% of organisations will outsource at least some of their business insight capabilities to a third party service provider in 2018. This trend has come about because so few organisations have the internal capabilities and resources needed to be able to effectively use the vast quantities of data that they’re collecting. Providers of insights-as-a-service use a combination of the client company’s existing data, usage data and syndicated third party data to generate actionable insights for the organisation. It’s this focus on actionable insights that differentiates insights-as-a-service from more traditional software-as-a-service providers.
Growing number of analytics partnerships between academia and enterprises
As already discussed, organisations often struggle to get value from their data because they don’t have the internal skills and resources to really make the most of it. The problem isn’t lack of data, it’s the lack of analytics capability. In academia things are the other way around. There’s no shortage of skilled analysts but it can be difficult to get access to valuable dataset. We’re seeing a growing number of partnerships between academics and businesses to address these twin issues to the benefit of both sides.
Many more organisations will adopt a cloud-first big data strategy
Forrester is predicting that half of organisations will embrace a cloud-first data and analytics policy in 2018. This is another response to the incredible growth in the quantities of data that organisations are generating and the prohibitive cost of hosting these volumes of data in house. Cloud solutions also offer far greater flexibility than is possible with in house storage options.
Further development of natural language querying
We’re growing increasingly used to interacting with devices using natural language querying in our everyday lives. Siri, Alexa and other virtual assistants are all designed to reply to natural language querying and so this approach is now commonplace in the home. In 2018 we’ll start to see natural language querying used more and more as a way of interfacing with analytics applications.
Real time deployment of AI
Organisations are going to be making much more use of real time AI deployment in order to make business decisions and provide employees with instructions in real time. We’re probably also going to start to see more sophisticated applications of intelligent self-service analytics. It’s the growing sophistication of natural language generation that will enable this development to really take root as it enables analytics results to be translated into easily understandable instructions and readable information.