Monday, 20 October 2014

Salesforce Wave and the future of data analytics

Forbes Magazine concluded a lengthy report on Dreamforce - Salesforce's mammoth conference held in San Francisco last week - with the comment that neither of the major product announcements - data analytics tool Wave and mobile app development platform, Lightning - were "as finished, or nearly as polished, as the four-day corporate love-fest that Salesforce has become the master of hosting."

I'd come to the same conclusion. Like everything else at Dreamforce the demonstrations of Wave were slick and seamless, but I also came to the conclusion that what was demonstrated was without doubt the future of data analytics and that, however far short of that Wave today is in reality, it will get closer, and probably quite rapidly.

Firstly, Salesforce has developed it Wave the basis that the primary user interface will be a mobile device, not only because mobile devices are now the favoured means of interfacing to cloud based IT services but because it believes that access to analytics capability needs to be available in real time to the people at the coalface: sales people trying to win deals, distribution managers trying to determine delivery volumes, etc, not just to data analytics specialists in back rooms.

Secondly, ease of use is a number one priority. Data analytics needs to be available directly to those who need the answers, not just to specialists and the results displayed in easy to understand graphics in any one of many different formats and flicked instantly from mobile to tablet or desktop.

Thirdly there needs to be access to data from many different sources to support a wide range of queries, many of which might be hard to anticipate in advance.

In one demonstration a rep from a financing company convinces a builder of luxury boats to take on millions of dollars of finance to build more of a particular model by showing (a) inventory of that model is very low and (b) demand for it is likely to be high based on the volume of positive chatter about it on social media sites.

Salesforce maintains that none of the demonstrations were set up, that all were real systems. That's no doubt true but what determines the usefulness of tools like Wave is not just the user interface - and Salesforce seems to have done a very good job of making very easy to use - but the data sets that have been integrated into it.

The choices made will inevitably be determined in anticipation of the types of queries likely to be made and costs would make it a case of diminishing returns to cater for infrequently needed queries.

But the vision is spot on. Being able to easily ask any question of any relevant data set, even those that might previously never have seen relevant, comparing the answers with queries on other datasets and getting quantified results presented graphically will revolutionise many aspects of many industries.

The realisation of that vision will do for the world of quantitative data analysis what Google it for qualitative information research. Remember the pre-Internet and pre-Google era where online research was the exclusive domain of librarians and researchers who understood the arcane technology of online databases like Dialog?

Launched in 1966, Dialog claimed to be "the world's first online information retrieval system to be used globally with materially significant databases." Its usefulness was limited by the range of its datasets and the skills needed to use it. And that is pretty much where data analytics is at today.


No comments:

Post a Comment