Businesses are beginning to realise the true value of data as a strategic asset, but to truly leverage trends such as digital transformation and real-time analytics, they must maintain high quality data in the system of record – the ERP system.
Vikram Chalana, co-founder and chief technology officer of data management leader Winshuttle, has redefined Application Data Management (ADM) as a framework to empower business teams to maintain and govern high-quality data in SAP ERP, in his new book, Application Data Management – Empower Business Teams to Get Data Right.
The book was launched at SAPPHIRE NOW 2017, and we spoke to Chalana about the changing paradigm for application data management.
Inside SAP: Tell me about the background to this book.
Vikram Chalana: People are looking at customer experience, how to leverage the Internet of Things, how to leverage machine learning algorithms and real-time analytics – but a lot of the struggle that we see people are facing is around data. They see that the data that they have in their enterprise applications is not good enough. The quality is not high enough for them to exploit these trends. They have been using our software and other technology like Winshuttle to clean up data. There are also other technologies emerging to empower business people to take charge of data clean-up and preparation so that they can use it for digital, big data, and machine learning algorithms. And they have been using our software and other technology like Winshuttle to clean up data. There’s other technologies that are also emerging to empower business people to take charge of data clean-up and preparation so that they can use it for digital, for big data, for machine learning type algorithms.
We find that the importance of data quality is something we find growing. Every customer I talk to is saying the same thing, “It used to be that I could take a week to clean data in the past and then put it in a data warehouse, and that was our approach. Now there’s real-time needs for clean data all the time, so we have to make sure that it right the first time in the system.” So that’s the big driver and we need to identify that as an emerging discipline.
Inside SAP: Data management seems to be increasingly occurring across businesses, not just in IT.
VC: Absolutely. That’s actually one of the critical things that distinguishes application data management (ADM) from master data management or other data warehousing initiatives – it’s a distributed activity. That means it’s not centralised with IT – it’s distributed among business teams throughout the company.
The other trends you see is in the analytics world. We’ve seen Tableau become very popular, Power BI – thigs that a business person can pick up and use. We are seeing that as a trend in data management too. It’s not just for analytics or reporting, it’s also data quality, that becomes more of a business-driven initiative, rather than a central IT initiative.
What we see is a very interesting paradigm. Central IT is managing some very critical data attributes and data objects – they might be managing customer names and addresses, but within a customer, you have many other attributes that are managed by other functions. For a customer, you might have contact and industry information that might be managed by marketing, for example. You might have payment terms and billing information that is managed by accounting. For the same object, you have different people managing different pieces of it, and that’s where the decentralised business team comes in, because in the past, everything was outsourced to IT. We are seeing that the most effective organisations are managing it in a decentralised way and the IT team is focused on very strategic objects and attributes.
Inside SAP: The book contains 10 customer use cases; can you give us an example of how this approach to ADM works?
VC: A global food and beverage company in North America, and they have taken this exact approach where they have a central team who is responsible for a small set of data. The central team sets policies for the rest of the organisation – they decide what good data means, who will manage what data, what tools they will use centrally and decentrally in the organisation. Every person knows what part of the data they are responsible for, making sure the quality is high, and they assign data stores to each of these areas and to each geography. That company is now considered best in class for the quality of the data they have.
Inside SAP: What’s the benefit of taking this approach?
VC: The biggest advantage is you are able to leverage some of these digital business ideas. Digital is not just about creating an app, it is about the different business models that you can go after or different insights that you can get from data. You can do analytics you weren’t able to do before because you weren’t able to trust the data; you can say, when I build models for predicting my sales, I know this is based on accurate data that’s in our system.
That alone is super-powerful. For Pactiv, one of the customers we talk about in the book, because of better management of data, they were able to reduce their cycle time of new products from 28 days. Their factories could be ready in a day, but the data was not ready until 28 days later. They were able to reduce that down to seven days, and that was a huge win, because they could go after new business that they couldn’t even think about before. They make paper cups, and weren’t able to bid for Super Bowl cups, because they had a two-week lead time. By reducing their cycle time, they could bid for that business. So it’s massive if you think about new revenue sources, new insights into data, and new business models potentially.
Inside SAP: For businesses looking to take a step in this direction, what should they consider?
VC: An important part is having data quality managed by business users. The technology should be very easy to use, not too complicated, and it shouldn’t involve master data modelling or trying to figure out what’s going on. Some of the technologies that are emerging around artificial intelligence or machine learning can themselves be useful to make these kind of tools easy for users.
Data governance is also related to this. When somebody identifies that data quality is bad because so many people are manually entering data, and there’s not control, the first thing people think of is setting up a data governance program. We talk about different ways of doing data governance – proactive data governance, reactive data governance – and how these application data management technologies can be used for all aspects. Data governance at its heart is not about technology, it’s about setting the policies and procedures.