Focus
The concept of the data value map arose from the observation of the difficulties associated with organisations in communicating and achieving a value from data initiatives. I will use some of my work during placement at Dell Technologies without releasing confidential information as an example. The data value map is split into separate categories which are Business Value, Acquisition, Integration, Analysis, Delivery and Governance. – [1] (Tadgh Nagle & David Sammon, 2017, 1441).
Business Value – this is focused on the outcome benefit of the data initiative. A figure of 83% of companies rarely calculate the cost of bad data. – (Tadgh Nagle & David Sammon, 2017, 1441). The focus of the data value map is to emphasise that data is a massive asset to the organization and can improve its position through the benefit of generating revenue and reducing costs.
Acquisition – This phase of the data value map is the gathering stage of data from activities and entities. The sources from which the data is gathered from has to be thoroughly examined as if the identification of good quality data sources is rectified at the start of the process it makes it much more affordable than later on in a project. On major problem of this phase is that many organization do not actually know what data they have or how much of it do they truly possess, how critical it is, the degree of redundancy of the data or where it is even stored. – [2] (Khatri and Brown, 2010).
Integration – this is the stage where data is combined and merged from the data sources. From the study of Tadgh Nagle and David Sammon, it was seen that 45% of organisations saw this as their driving point in providing funding for analytics. (Tadgh Nagle & David Sammon, 2017, 1441). The point of this phase is to align the data from the organization and observe how well integrated their data is, meaning of which of all their data how much of it actually aligns to the one goal or truth.
Analysis – during this stage the analytics is carried out on the sets of data. If data analytics is carried efficiently and properly then organisations should be able to reap the benefits. Such advantages include better strategic decisions, improved asset management and better risk mitigation.- [3] (Davenport and Harris 2007, Davenport et al. 2010).
Delivery – this is where it is decided on how will the data be delivered and seen by its users. This will determine the way a user will receive their business data to hopefully resolve any issues that the organisation had faced. The purpose of this stage is for the user to acquire high quality data in a timely manner. It is also recommended that the delivery of data should be aided with graphic visuals. This will help the data users interpret the data easily and make it more user friendly. – [4] (Few 2006).
Governance – understanding and combining the four above sections (Acquisition, Integration, Analysis and Delivery) is known as governance. According to Tadgh Nagle and David Sammon, governance is “the promotion of behaviours for good data practise” – (Tadgh Nagle & David Sammon, 2017, 1442). This means that if a firm was to use these steps from the Data Value Map it would be an efficient use of data practice. From a studied survey from Tadgh and David they gathered that 53% of firms believe that governance can take less then 3 months to implement however only 10% actually have governance in place. – (Tadgh Nagle & David Sammon, 2017, 1442).
The aim of governance is to create an understanding of what data governance firms may have in place, how mature it is and how it can become integrated part of the organisation’s data initiatives.
Methodology
I will describe how I used the Data Value Map and will show visualization of the analysis following the methodology used. When working at Dell I had come to learn that with such a huge company there was so much data in terms of customers who were looking for Support. They would be broken down into customers who were “In Warranty” and “Out of Warranty”. Both sets of customers had their own respective departments and I was based in the “Out of Warranty” (OOW) department as an analyst intern. For these customers Dell had 3 tools which were used for logging data for OOW support. Without naming these tools for confidential reasons let us call them X, Y and Z where X would be the tool where all the relevant data in terms of the process of purchased support would be logged, Y would be the tool where the relevant data in terms of communication between Dell and the customer would be logged and tool Z would be where any incidents raised for a certain case would be logged. Now where it gets messy is that Dell had a new web application set up for tech support agents to log customer’s system data while on the call / chat. This was going on for the Europe, Middle East and Africa (EMEA) regions. This data would be stored on SQL databases with various amounts of column’s and about 45000 rows as of 31/8/2018 and increasing. In this setup the Data Creator would be the Tech Support Agent and the User would be Senior Management.
In terms of the business value of the DVM I believe the number of column’s recorded in the Database were too many highlighting bad data. The business value as well would be that management could see how many customers of all the entries actually bought support and on the contrary how many customers turned down support. From that management may also take note of why or what the reason is for customers not purchasing out of warranty support.
We know that for the EMEA OOW support Dell has 4 sources for data in the tools X, Y, Z and the newly developed Web App. Therefore for the acquisition stage these are our sources of data. The one thing that should be common in all these sources of data is the Service Tag of the system which can be found on a label underneath the system and by an identification number for when a customer contacts OOW support. During this stage all the unnecessary data should be removed in terms of the business goal. Once all the relevant data is needed we can move onto the integration stage.
During Integration all of the data from the 4 data sources have to be aligned and reflect the truth. For example what this could mean in this scenario is that lets say one incident regarding a specific service tag has multiple support identification numbers. We must find the identification number that is relevant to that specific call. This would be done by cross examining dates of the identification numbers. If the system came in on a specific date and was logged it would have a unique identification number. Any other identification number for the same system would mean that number would be relating to a separate problem with the same system. So in this stage a lot of cross examining would have to take place along these 4 sources and integrated aligned toward the end goal required by management.
At analysis senior management at Dell in the OOW Department should be able to see multiple benefits from the integrated data. From my time spent at Dell and carrying out a similar project there were many observations made. Such benefits could be seen as to what line of systems had the most common faults that were out of warranty. Other benefits that could be seen was which countries were generating the most revenue in EMEA for OOW repairs and which were not. Another was that we could see that some customers were refusing to purchase repairs because of the high costs associated with the service. Some other observations made was that the third-party repair center which I cannot name due to confidential reasons was taking from a day to several day to diagnose customer systems.
Finally, at delivery all the gathered data has to be delivered to senior management in a visual and easy to understand method. The most common way to do this would be to present a PowerPoint presentation with visual graphs and pivot tables showing the most beneficial findings from the gathered data. As from some of the benefits as talked about above management could make the decision on which countries to promote OOW repairs in where revenue may be lower, workaround how to make the cost price of repairs more appealing to the customer or find methods in reducing price to increase demand and sales. Upon delivery it can help reduce overheads such as the time spent in diagnosis at the repair centers or how if common faults are occurring with certain systems, if engineering can fix these issues before releasing the next wave of systems.
With all 4 above put into practice Dell would have good data practice for the OOW department. Now from my understanding if that I was being asked to carry out this analysis during my time there as an intern it may show a lack of resources for this department to undertake this type of analysis as the company is so huge, but the benefits can be seen.
Visual
Resulting Impact
From the gathering of the data it can be seen the beneficial effects of using a Data Value Map for companies. However, would the likes of the much bigger companies it would require Data Value Map based on per department or per project. A potential downside to this is that some organisations may not have the man power or hours allotted to accommodate the design and implementation of DVM’s. This would mean that firms would either have to spend on recruitment of extra resources or either outsource man power. This is all in the conditions that the organization does not have the required man power to carry out DVM’s.
When the DVM is carried effectively and efficiently as possible the number of benefits to the user can be vividly seen. The immediate reaction I believe from this example is that senior management should learn from the data findings and try come up with ways to benefit from the data and increase growth for the firm. This could be by increasing customer satisfaction or generating more revenue for the organization where possible.
In terms of future planned action Root Cause analysis can take place on systems with reoccurring problems to prevent any large known issues. A sample can be taken from forums that a known issue with Dell XPS systems is that usually after a year of use they can be prone to swollen battery issues, resulting in damaged battery’s, palm rests and keyboards of the line of systems.
This my analysis of using the Data Value Map and how an organization like Dell who I completed my third-year placement with, can benefit from the implementation from it.
References
[1] – Nagle & Sammon, T., 2017. THE DATA VALUE MAP: A FRAMEWORK FOR DEVELOPING SHARED UNDERSTANDING ON DATA INITIATIVES. 1st ed. University College Cork: University College Cork
[2] – Khatri, V. and Brown, C. V. (2010) Designing data governance. Communications of the ACM, 53(1), pp. 148-152.
[3] – Davenport, T. H. and Harris, J. G. (2007) Competing on analytics: The new science of winning, Boston: Harvard Business Press.
[4] – Few, S. (2006) Information dashboard design: The Effective Visual Communication of Data, California: O'Reilly.