Next up a customer panel with cspire, xo communications and Purdue University.
First question was what new things can you do with Watson Analytics?
- Easily convert data into a story in the context of a business
- The cloud supports a very distributed workforce allowing more people to access more advanced analytics
- Empowering people who would otherwise have asked an expert to do something to do the analysis for themselves.
- More success with predictive analytics tends to result in more demand and this would allow line of business to participate more, letting data scientists engage with more complex problems.
What was most compelling about the announcement?
- End to end speed for the complete data to insight process
- Powerful features delivered through a totally new, more accessible user experience will make those features more accessible
- The data refinery, the ability to quickly recode data for instance, offers the chance to get past blockages and work on insight
- Collaboration tools help break down the barriers between the business, operations, IT and analytics people
Do you worry that people can get in trouble using these more powerful analytics?
- No, the level of knowledge is there and when supported by visualization this won’t be a challenge
Why does the natural language matter?
- Really eliminates the “where to start” problem
- Especially when combined with the no-blank-slate interface that Watson Analytics presents
Is it OK for Watson Analytics to empower business users directly over the cloud?
- Yes as it will get more people to think more deeply about data which is great in the long term.
- Will put some pressure on central analytics groups as it will set expectations for this kind of interaction and will raise the bar.
- Yes as it will support the creativity of business users to solve their problems.
- The ability to provide trusted datasets, to have analytics/IT staff collaborate with business users will also help.
Where will you start with Watson Analytics?
- Sales organization because sales and marketing are the folks who have the most questions and cause the most backlog. Exposing the data they need to them and teaching them to answer their own questions will be a powerful tool
- Smaller, more local problems that might not have bubbled up to the central group. Especially those outside the administrative realm where domain expertise is really important.
How important is speed?
- A bias toward action means that a predictive model is just the start – have to change behavior too. The more rapidly every step can be done the better and the faster you get to action.
Would have been great to hear from actual Watson Analytics users but nice to see what these very experienced customers see as the value proposition.
Comments on this entry are closed.