Nokia Improves Analytics of Big Data

Lee Feinberg, Sr. Manager of Decision Planning and Visualizations

Nokia is a leading global supplier of mobile phones, telecom networks, and related customer services.

Lee Feinberg
Senior Manager of Decision Planning and Visualizations
I would say the biggest change is that people are actually thinking about, “How do I do something different in the company?” They’re considering Tableau as a tool to change how they work.

Tableau: How did you discover Tableau?
Lee Feinberg, Sr. Manager of Decision Planning and Visualizations: Actually, I’ve known about Tableau since about 2005. I was working for a consulting company and looking for a way to get out of using PowerPoint and Excel. I came across Tableau just through constantly searching on the Internet for something better.

Tableau: What was the process of rolling out Tableau at Nokia?
Lee: We started pretty small with a couple of Desktop licenses just to prove to ourselves that it was something useful that could have a future in the company. Then we started rolling it out to a few more people, and eventually it got to the level where we wanted to implement Tableau Server. So, we put that in as well and again started with a handful of users to show the performance. Then we grew on top of it.

Tableau: What kind of data are you looking at?
Lee: Some of it is user data, about how customers use our products and our services. Some is data about surveys that we take across our user base. We have panels, so we use it to understand where we should be going with our products. It’s a really wide range of information.

Tableau: Have you been able to quantify the time or cost savings with Tableau?
Lee: We haven’t really been able to quantify it yet. I would say the biggest change is that people are actually thinking about, “How do I do something different in the company?” They’re considering Tableau as a tool to change how they work. They realize that now they don’t have to rely necessarily on some of our other tools that we still use. They’re great too, but Tableau is an option where they can do things in a new way.

Tableau: So how is Nokia doing things differently with Tableau?
Lee: I think the key is that people who have their own sets of data are able to take that and do their own analysis. The data is not all locked up in a database. Before, they either didn’t have the data available at all, or they had to use Excel or PowerPoint to do the analysis. It’s pretty clear that using Tableau is a really fast way to get through large amounts of data.

Tableau: What kind of feedback are you hearing from users?
Lee: You hear people frequently say, “I want this report that I used to have in Tableau,” but they don’t necessarily know what that means. They just know they want it in Tableau now because they’ve seen the work that other people have done with Tableau. In some ways, it’s speaking for itself in the quality of the work that’s produced.
People that have been using it more for analysis than just for traditional reporting are starting to see some of those benefits. They’re able to get information out into their teams a lot faster than they may have been able to do before.

Tableau: What kind of “aha” moments or new insights have you had using Tableau for analysis?
Lee: I’d say almost every week we see something unexpected going on from a couple of different perspectives. One is just the data itself. We have so much data. We’re able to see through Tableau when there might be something happening in the data that we didn’t expect, just in the overall data set.
Then when we get into the analysis phase, it’s easier for us to see trends, whether it’s with our devices or with our services group or in a certain country. We see behavior, for example, where we may have a lot of users coming in from a marketing campaign, and then we’re able to see trends about how long those users stay with us and interact with us. If you can figure it out once you get a customer how to keep that customer, that’s really important to the business.

Tableau: Tell me a bit about your data environment.
Lee: We have lots of different systems that are connected, and some people use Excel as their database. We have Teradata, and we also have Oracle and many others in the company.

Tableau: Do you use Tableau against all those different systems, or is there a primary way you use it?
Lee: We mostly use it for larger data sets. We use the extract capability. We’re not going live to the data because there’s just so much. And it also doesn’t change frequently, so we don’t need to be connected live. But we make great use of the extract capability.

Tableau: Do you ever work with two different data sources in the same workbook?
Lee: Yes, we do. We have some instances where there’s data coming out of one of our main databases, and we’ve supplemented with Excel files to create some mappings of some custom data. That’s really been very easy because we didn’t have to go back to the system developers and ask them to create one field for us. So it was a really easy way to move the project ahead.

Tableau: What kind of best practices would you recommend to another company that was looking to manage a lot of analytics with large data, like you’ve been doing with Tableau?
Lee: I’d say in the long run, even though we started small, you have to partner up with your IT group if you really want to grow it. The earlier you can sit down with them and tell them what you’re thinking about, and have them experience Tableau, the better.
There are so many solutions out there. If it’s a large company, they likely have several other existing implementations already. There’s always the question of “why is this one different?” I think you really just have to experience it. You can’t do it by putting check boxes on a paper and trying to do a comparison that way. I think Tableau changes the game.

Also, in terms of how do you get people to actually use it, I would say go fast. Don’t spend lots of time in meetings and planning the way you would go through your normal business intelligence processes. Take a very small project, use Tableau, and get something out that people can start using. Then they’ll see that it’s different.

Tableau: That’s great advice. Do you have any final thoughts?
Lee: I’d just say that the key, especially in a large company, is to prove it first in a small way. Then go and find somebody who has a need that you can help them fix. It’s not your traditional BI route. It’s really about finding that internal customer who has a problem, and you can show them how to fix it.
That was our key to success. We had one executive who was just wowed by the possibility, and he said, “I’ll give you the money to go and start building this out.” Once he did, that was the start of our success.

You might also be interested in...