Spotlight On: Kabel Deutschland
Kabel Deutschland (KD), Germany’s largest cable operator, offers its customers digital, high definition (HD) and analogue TV, Pay TV and DVR offerings, Video-on-Demand, broadband Internet (up to 100,000 Kbit/s) and fixed line Phone services via cable as well as mobile services in cooperation with an industry partner. The publicly listed company (MDAX, MSCI Europe, Stoxx 600 Europe) operates the cable networks in 13 German federal states and supplies its services to approximately 8.7 million connected households. At the 2011 European Tableau Customer Conference, business analyst Giedre Aleknonyte shared how Tableau empowers her team to blend millions of rows of data from a wide range of sources and achieve new understanding in their billing and collections programs.
| Tableau: | How are you using Tableau at Kabel Deutschland? |
| Giedre: | I work on the Billing and Revenue Assurance Team, and we focus on analyzing data related to billing, fraud and collection processes. We have Tableau Desktop licenses, and we also use Tableau Reader. |
| Tableau: | What kind of data are you looking at? |
| Giedre: | We’re working with very large data sets from a variety of sources. Some of them are Excel or text files. Others are Access databases. We also draw directly from other databases and data warehouses based on Oracle. |
| Tableau: | Do you ever work with multiple data sources in Tableau? |
| Giedre: | Definitely. We do projects that involve different parts of the company, so we’ll have sales-related data or marketing-related data that we’ll need to combine with our billing data. It’s impossible to have data sets in the same system or in the same format all the time. The data blending features in Tableau really help us analyze and make sense of all our data. |
| Tableau: | You said you work with very large data sets. How large? |
| Giedre: | With over 8.7 million customer households, even if you just look at the billing for one month, that’s several million entries. |
| Tableau: | And you’re looking over many months or years? |
| Giedre: | Exactly. Typically we’re looking at several million rows of data for every month, and if you analyze trends over time, that can blow up to 10-15 million rows. So, it’s very important for us to have a tool that’s fast and efficient. |
| Tableau: | Who are the consumers of your analyses? |
| Giedre: | Right now we have a couple of people who build the workbooks and dashboards. Some of our work goes out of Tableau – meaning that we build something and then export the graphs or dashboards into PowerPoint and send those out to people all across the company. Our directors and managers from various areas of the company look at this. We also have dashboards that are more for daily use by people in our department. |
| Tableau: | What benefits have you seen from using Tableau? |
| Giedre: | Well, first of all, Tableau has a wow effect. When you first show somebody Tableau, people are very amazed. I used Tableau at my previous company and there was a rumor that our department had something magical, a magical tool, but nobody really knew what it was. I’d get emails asking me, “Hey, we heard that you have this amazing tool for modeling geographic analyses. Could you please share this with us?” or “Could you model our data?” At first I was confused, but then I realized that they were talking about Tableau. Tableau is a geographic analysis tool. Tableau is a fast data engine. Tableau is an analytical tool. It’s all in one. |
| Tableau: | Have you been able to quantify the benefits in terms of time or money savings? |
| Giedre: | Again, at my previous company I used to do reporting, and before we had Tableau we just used Excel. Weekly reporting used to take anywhere from three to four hours in Excel, but with Tableau we automated that and came down to maybe 15 minutes if the network was very slow.
Here at Kabel Deutschland, if we took all of our little data files that we receive from various sources and tried to put those in Excel and build a solid, sophisticated dashboard it would probably take forever. And then to refresh and update every month would take significantly more time than it takes now at Tableau. |
| Tableau: | How have your new colleagues at Kabel Deutschland received the Tableau visualizations? |
| Giedre: | Kabel Deutschland started using Tableau just a couple of months before I joined the team. I originally joined the team so that they could get better in Tableau and improve their visualizations, and they’re receiving it very well. At first, Tableau had the wow effect, but now we’re slowly getting to a point where wow isn’t enough. People see it’s very amazing, and then they want more and more out of it. The appetite just gets bigger. |
| Tableau: | What advice would you give to a company that’s new to Tableau? |
| Giedre: | I would say have a specific project in mind when you start. Don’t expect Tableau to be something you can apply to all cases everywhere. Start small and see if Tableau works for you. See how people receive the projects you do with Tableau. I can almost guarantee they will be delighted. Then you can go from there. |
| Tableau: | Have you had any “a-ha” moments with Tableau? Can you tell us about a time when you learned something from the data that you didn’t expect? |
| Giedre: | Definitely! We use the mapping feature in Tableau a lot because, as a cable company, our regional differences are very important. Using traditional tools like Excel, it’s difficult to get a sense of what the geographic data means without actually seeing it. With Tableau, once you plot the data on a map or create a heat map or use the polygon features to create, it’s very dynamic and interesting. It’s very different from what you would get by just visualizing it in a crosstab or doing a simple line graph, such as bad payers split by regions or bad payers split by different channels in different regions. When you see the data on the map, you can actually see what it means and what the different regions are.
We were analyzing customers who did not pay their bills, and we started out by using Tableau’s mutual mapping features and plotting the non-payers by region. It seemed that we had a lot of non-payers in certain regions. But then we took the next step and created polygon maps for the bad areas that we detected, and when we visualized the regional data one level down to the postal level, we realized that it was more like specific areas in a specific city that were the problem for us. We would not have seen that if we had just stayed at the top level and just plotted regional data onto Tableau’s built-in maps. |
| Tableau: | Any final thoughts on using Tableau? |
| Giedre: | Well, it’s an amazing product. I’m a big fan. I’m one of those people who believe that in the future, you will say “I tableau data” when you analyze numbers, the same way you “google” when you search for something on the Internet, even though you may not be using Google. So, I’m very happy with my use of the product, but this is not the end. I’m looking forward to improvements. Tableau seems to be getting better all the time. |
