Cloud Analytics - the story so far...

If you thought that the Hype curve was steep when it comes to Analytics, watch it spike now with Cloud Analytics! The marketeers must be delighted to be able to merge two of the most hyped IT concepts. Now, if only they could bring in a third, like Big Data ….. oh, wait…!

Organisations are beginning to get to grips with Analytics, by taking time to understand their key information before investing heavily in tools. They have learnt lessons from sometimes horrific datawarehousing projects, which have often delivered little at significant cost.

Yet there is still a need to deploy agile solutions to the business, which offer rapid ROI, without either significant capital outlay or long deployment timelines.

Software companies like Qlikview and Tableau have capitalised from this need. They are creditworthy tools and may just have nudged the big players enough to make some smart moves.

Oracle purchased Endeca some time ago and now positions this as a rapid start Analytics tool. IBM promises much with Watson and the messaging coming out of the SAP has changed recently with the introduction and rollout of HANA to many of its ERP and non-ERP customers. Salesforce has also recently made a play, by introducing the imaginatively named Wave, ‘a beautiful user interface designed to make interpreting and using business data easier and more intuitive than ever before’.

So what is the Cloud Analytics story, how will it affect organisations in the near and far-term and where is Oracle playing?

Cloud Analytics is of course described differently by each vendor to suit its own strengths, but essentially boils down to a Cloud-based deployment of Analytics tools, a data warehouse or both. It is typically purchased on a subscription basis, so has a low capital outlay and it is managed by the vendor (or partner), so there is no need to employ expensive technical skills. End users can then have full access to the data, in a way they understand and they can then build and deploy complex reports very quickly.

The downside for organisations could be that:

  • The tools in the Cloud may be cut down or non-standard versions of the on-premise tools.
  • There is little evidence yet of this approach working.
  • An organisation’s data may need to be held in the Cloud, with its consequent security and privacy concerns and
  • They may have little control about how the data is transformed in the Cloud, for instance adding new fields into the warehouse or changing the way a metric is calculated.
  • A lack of IT involvement may lead to a lack of governance around the data, as each departmental fiefdom loads and interprets data in its own way.

It is therefore clear that an organisation should consider carefully what it wants to do about Cloud Analytics. There is nothing remarkably new about the technology, so the risk is not great. The subscription model may be attractive, along with the time to deploy. However, much of the proposition can be implemented with existing tools on premise, possibly using a different, more agile approach.

[For more about how Oracle is playing in this area, please see my next blog.]

My feeling is that Analytics will be the very last thing organisations deploy in the Cloud. It contains much of their competitive advantage, it needs to be massaged and manipulated in a way Cloud providers rarely offer and, as business requirements change and visualisation capabilities improve, organisations need the option to utilise different tools against a consistent well designed information model. The promise for Cloud Analytics is clear, but organisations will certainly benefit from evaluating their current state before embarking on a journey with any Analytics tool, let alone Cloud Analytics.