categories

HOT TOPICS

Subscribe to our Feed

Thought Leaders in Big Data: Interview with Tim Minahan, Network Strategy VP for Ariba (Part 4)

Posted on Saturday, Mar 30th 2013

Sramana Mitra: Talk about that a little bit. Our community is full of small companies. If they want to try out the Ariba network, what would you advise them to do?

Tim Minahan: I would invite them to go and get discovered with a service on the network we call Ariba Discovery. It is at discovery.ariba.com. It operates just as if you were registering on eBay or Amazon. Go in, setup your account and provide a little bit of information on your company. We provide guidance on how to build a robust profile, what categories and regions you support, what services you provide, etc. Then when opportunities arise – when one of the buyers on the network is looking for a good or a service that meets your qualifications – you will be alerted, and you can begin being exposed to new business. It is that simple.

SM: What kind of volume of data are you dealing with in this network?

TM: We are dealing with over 45 million purchase transactions that went on over the past 12 months. Sixty-five million invoice and financial settlement transactions went on in the same time. You repeat that back 11 years, and you begin to understand the volumes that go on there.

SM: What technology stack are you on?

TM: We are [SAP], but it is all available in the cloud by its very nature. By its very nature it needed to be architected as a single instance multitenant environment. That is exactly what it is. It is a multitenant platform that regardless of whether you are a Fortune 100 or Fortune one million company, you merely plug in once to access it from a simple web browser.

SM: I was wondering more about the back end of your data infrastructure.

TM: We are J2EE, backed up to your typical Oracle database.

SM: That is interesting. You are inside SAP and you have an Oracle database.

TM: We are only about four months inside SAP.

SM: What is in the future of this network?

TM: I think, as you page forward, the first year of value for networks was creating efficiency – providing a scalable way to connect ones to transact and collaborate with many. But the next phase of network business is going to emanate from mining the insights from all the interactions, transactions, and unstructured engagement and commentary that occurs within the network.

This is what some might see as big data. We are already seeing it in our personal lives – whether it is Amazon that harvests the buying patterns of its customers to recommend complimentary products for upselling, or Twitter or Facebook that mines unstructured comments to develop psychographic profiles for highly targeted advertising. The potential of analyzing the hundreds of billions of dollars of financial transactions in relationship history tied up in certain business networks is staggering. You are going to look at things like offering transparency to relationship patterns and historical payment trends that can lower risk not just for buyers and sellers, but also for banks and other service providers. A good example in the Ariba network is that we now have other service providers, namely banks, coming to offer up supply chain financing, since it is a low-risk proposition for them because they know who the buyer is and who the seller is, and how they performed historically.

SM: Do you anticipate that the network is going to become an advertising platform as well?

TM: There is certainly a potential for that, but it is not a direction we are taking right now. However, we have begun to harvest some of the information on the network to make recommendations – not for a fee, but for free. If I am a buyer looking for web development services, we have not only the transactional history of what other buyers have bought, we also have the relationship and community-generated information, and we run it through an algorithm that asks, “What are your criteria? Who are the suppliers who have the capabilities?” Then, based on what others have bought, here are the two, three, four or many suppliers that we would recommend – not just based on their capabilities but based on the idea of “did other people who were looking for the same good service also buy from them?”

SM: So, it is an unbiased organic recommendation system that is generated by the system as opposed to an advertising system.

TM: Exactly.

This segment is part 4 in the series : Thought Leaders in Big Data: Interview with Tim Minahan, Network Strategy VP for Ariba
1 2 3 4

Hacker News
() Comments

Featured Videos

`