Sramana Mitra: That is a pointer to a lot of entrepreneurs. That is an area that needs work. The other thing that is interesting, specifically in this topic, is that you offer this technology, which is kind of Ferrari-level technology in analytics. Inside organizations, there aren’t trained business analysts who can really take advantage of these technologies, use them, and run them to their full capacity. If you are trying to sell to a finance organization or a risk organization inside a credit card company, they may not have the analytical talent or the kind of talent you need that has business, mathematics, statistics as well as communication/presentation capabilities to make full use of these kinds of technologies.
Michelle Chambers: Productivity tools that are aimed at business analysts as opposed to data scientists or developers really don’t exist today, and there are no techniques that help leverage the incumbent. When they have taken approaches to aim at these groups, it seems to me they did so such that it was packaged in the wrong way. What I am saying is that if they had automatic model fitting, which is aimed at business analysts, then they would embed it in a tool that is aimed at data scientists or statisticians. So, there is a bit of a misfit there. There is definitely a market for tools in that area.
SM: I was actually talking not about productivity tools, but about people.
Dave Rich: We agree with you on the problem, and I think Michelle has set up the way to a possible solution – “R for dummies.” I think what you are saying is that there are not enough data scientists in the world, and there are not enough business analysts, who frankly need to be the fusion. Think about what they need to be. They need to be part data scientists, part computer scientists, and they have to be experts in their industry or the functional domains they are supporting to know what would be an interesting way that department makes decisions.
It is an interesting dynamic in regard to my role back at Accenture – I created and ran Accenture analytics globally. I saw a wave of change coming. What business process re-engineering was for the 1990s, decision process re-engineering was the next big management wave. This is why a company like Accenture was interested. As in those previous waves, you didn’t do business process re-engineering in isolation. You needed the types of productivity tools that made it easier for people who are in the business to do their day-in and day-out of jobs.
This is a big wave for entrepreneurs to take advantage of. We are going to see a proliferation of decision support tools and applications that we customize. When I come in in the morning or when I pull up my iPad and I am a credit risk analyst, a chief marketing officer, or somebody who is located within the bowels of a supply chain, and I am making critical decisions relative to inventory levels, or I am a trader for an energy company or an agribusiness, I am making some pretty big bets.
SM: The point I am making is that there is a proliferation of technology developments right now in this area. But none of those technologies is going to reach full-scale adoption unless there are people inside of these organizations who can use these tools. As you rightly said: Think about what these people need to be. The fusion of skill sets that these people need to have is a rare phenomenon.
DR: It was not that different back in the 1990s. Back then, we were taking advantage of emerging technologies to deliver on the promise that business process re-engineering had. This is pretty analogous, actually. You are right. This is an opportunity for revolution. This is why Michelle and I are here. I am from Accenture, she is from IBM, we have former SAS people here, we have people from Infomatica and other organizations like SPSS and Norman Nie, the founder of SPSS, who is still on our board. We have some of the best minds on the planet around building what I call the rules, tools and schools, to allow for mass adoption. If we look at all of the key stakeholders and the key roles in any given transformation of any industry or company, it takes a lot of people. If we had, for example, a design repository, you have the 80 percent for fraud detection or marked-down pricing if I am a merchandiser, and you have a place to go that would give you the jump-start kit, so you are building something from scratch. This is why we are all here, we believe that this is our opportunity. Our product strategy is to build the repository for the 80 percent solution.
MC: We concur with the perspective that there are not enough data scientists in the world to leverage the business value that is out there around big data and analytics. There are a couple of different strategic paths you could go down. One of them is taking that and leveraging it into applications, because then it makes it consumable – you don’t have to have that in-house ability, yet many people can use it as applications and apply it in their unique business strategy in a way that will help them to get the business style they are looking for.
That is one path. Another path is having productivity tools, and those productivity tools are getting to a point where you can predefine models around certain common use cases. It will perhaps start with people using them in wholesale and tailoring them to their specific business needs. Today you don’t feel a lot of the analytic applications because people come to business in a generalizable model. I think it is because most of the people who are doing analytics today don’t have the computer science or engineering kind of background, but I think that as this class of data scientists emerges, they will figure out ways to make those generalizable. Once they do that, you are going to see their discoveries leveraged into either apps or widgets that can be built into apps very easily.