A Conversation with Steve Elfman, President, Network Operations and Wholesale, Sprint Nextel

A Conversation with Steve Elfman, President, Network Operations and Wholesale, Sprint Nextel

This interview with Steve Elfman is excerpted from the 2011 Mobile Future Forward Book. We are delighted to have him join us for the 2012 Mobile Future Forward program.



Running a multi-billion dollar operator business requires the best of operational experience and foresight. As the mobile business is becoming more data-centric, there is a significant emphasis on running a tight operation to manage the margins. Driving the operational efficiency is as important as revenue generation to help manage the margins. Steve Elfman, President, Network Operations and Wholesale, Sprint has a very keen eye on operations and how to make them better. He has done that repeatedly throughout his illustrious career. To gain a better understanding of how operators think about operations and how the ecosystem should think about the same while coming up with relevant solutions and products, Mobile Future Forward (MFF) sat down with Steve to gain some insights on what drives operational efficiency.

MFF: When you get into a new situation, what’s the process you follow to figure out the opportunities for operational efficiency? Is it the process or the technology decisions that drive your thinking?

SE: It is actually yes on both counts and then a third. The first thing I look at is the overall financial picture of the company, the really high cost items of the company. Inventory is a good example. How much are we spending on inventory? Additional things to take a look at are – what are the big CAPEX and OPEX items, the network, the IT organization, building solutions, the subsidy, the logistics, and the distribution. So, start with the financials.

In the financials, I look at the two key metrics – the quality and the cycle time (and waste etc.) and then go into a deep dive and quickly determine where you can get operational efficiency with processes and change those. Then you jump to technologies to make sure we got the most efficient technology to not only automate but also to figure out if we are getting the cost down by using the technology. It is equally important to assess if it is impacting the go to market strategy.

The third thing I get into is the ecosystem and the partnerships like outsourcing and where we can work with the domain experts and partners to bring in the efficiency, the time to market advantage, and also minimize the cost out of the equation.

Finally, it is important to establish the benchmarks so you can gauge the progress.

So, that’s the kind of the process I go through.

MFF: Can you give us some examples in each of these categories?

SE: Sure. When I got here, I looked at the whole network operations. Our network cost and the margins were not in sync compared to of our competitors and when I looked deeper, it was easy to explain – we had more labor costs because we had lot more cell sites and switches per subscriber and then we looked at investments we have been making in the processes and in the technology. It drove us to look at the outsourcing opportunities to improve efficiency and also to automate a bunch of things. And we found a partner in Ericsson which has been operating networks in various parts of the world. This allowed us to focus on customer service while Ericsson helped us with automation and improving the processes, in the use of best in class technology so we didn’t have to invest in best practices.

Second example – in pursuing the Network Vision, we said that in addition to the operational efficiencies, we need to leapfrog our competition and go to the next generation technology. That’s why we went to multi-mode technology which is lower cost. Like many, we saw the data tsunami coming and we needed flexibility and a lower per unit cost.

MFF: How has the focus on operational efficiency changed as we are migrating from the voice world to a data-centric universe?

SE: Indeed it has changed and the reason is quite obvious. Data uses greater bandwidth, especially, rich media, and the usage characteristics are so different than voice. The pricing equation is challenging as we try to be efficient and at the same time give customers what they want, where they want it. The increasing demand of data has driven us to be more operationally efficient. Additionally, the go-to-market strategy has to be different as well.

MFF: Do you see the segments of postpaid, prepaid, connected devices differently? Since their cost and the revenue structures are different, does it mean that the strategy for operations is different as well?

SE: Yes there is but you have to look at not just the operational efficiency and here is why. Connected devices space is very nascent and immature and I will call that the segment is in the early part of the S curve and so you are making the investments, some will work out and will not. But you are trying to have other businesses fund the investment just like a new business.

In terms of prepaid and postpaid, both are fairly mature businesses, prepaid a bit less so. So, the cost structures are mature as well. The prepaid business can ride on top of the postpaid business since the basic components are the same – operational efficiency in the network, many of the same underlying IT systems, same distribution and logistics, warehousing etc. The prepaid segment has low lifetime value (LTV), so you need to get all the costs out to make it an operationally efficient business.

MFF: In the mobile world today, things are changing so fast as new devices are coming onto the market and the consumption itself is changing dramatically, how do you go about planning for 2-5 years out when the forecasts get exceeded on a continuous basis?

SE: I think you hit the nail. Forecasts by their nature are wrong. The only way you can be effective is by getting more discipline in that forecast and it doesn’t mean that it has to be right the first time but you are forecasting more regularly as you get more data and information about how things are selling, what the usage patterns are by subscriber, media, and data type. So, you are updating the forecasts more frequently, almost like an MRP-type system.

At the start of the year, you start with what devices you are going to put out, what kind of promotions, what capabilities, what applications, and do the micro segmentation of the customers to figure out the best strategy. Additionally, you are spending lot of time with network gear providers and device OEMs and other companies in the ecosystem to look at offload and optimization strategies more so than it used to be in the voice centric world. We also work closely with the major content providers to make sure that the multimedia content is optimized for the network. So, tighter planning, more data analysis, and finally look at the revenue side as well and try to differentiate the services and how you price services and products.

MFF: From managing the competitiveness in the market, does it come down to maximizing the margins per bit? What metrics do you keep an eye on that tells you how the business is doing?

SE: I think there can be a gazillion metrics but at the end of the day, it is the EBITDA. You have to make sure that you are delivering on that metric and it drives a whole bunch of behaviors whether it is subsidy, operational cost, negotiating deals, offload, roaming, access costs, etc. The EBITDA is a central measurement, week over week. Others are revenue, net-adds, churn, gross adds – day over day because then you take your precious EBITDA to invest so that you keep more customers and focus on higher value customers. It helps with pricing and with competitiveness. Managing your unit cost per GB is important in driving it down to the lowest possible cost.

Every day in this industry there is something going on which nullifies what you thought was true yesterday and I will give you an example. We went into the year not expecting that AT&T will put out a $49 iPhone. The day it came out – big surprise from the competitor – looks irrational, how do we respond to that. For a company like ours, cash flow and EBITDA are critically important. Over the last 3 years, we have managed the cash very well.

MFF: Operators obviously understand how the business is run. Do you think the ecosystem understands how the operator business is run and what metrics are important so they can better design solutions for you?

SE: No, I don’t think so. However, where we are in sync are with the larger vendors as you spend a tremendous amount of time with them, spend lot of money with them, so you tend to go through the underlying metrics in great detail with them. At the smaller and the mid-level, historically, we have spent less time since we spend less. And what we have tried to do is to recognize that a great deal of innovation comes from outside the operator. It is our job to seek out and support these innovators to help us out.

We recognize that by being open, our partners can come up with solutions as they are nimble. There has been a mismatch in the past but it is getting better. We are interdependent on their success. Traditionally, the culture in operators has been more of a supplier/customer relationship vs. a true partnership and I have tried to shift that dramatically here at Sprint.

MFF: As you look towards the next three years, is there going to be any fundamental shift in strategy?

SE: The time is upon us when making money from other than subscribers is essential. Over the course of the next three years – mobile advertising, mobile commerce, VAS, where operator can play a key role, can help us make a lot of money as long as they are not greedy.

MFF: Does the operator model shift to empowerment and enablement?

SE: Yes, we should bring the assets we have to the table, to the developers and to the ecosystem – things like the network, the back office, and the distribution and show that the ecosystem can generate revenues for everyone. This helps in building new revenue streams and share fairly with our partners while enriching the customer experience.