**This is the full version of an article I wrote for this week’s Reality Check column in RCR Wireless. To meet the character limit, the RCR version was truncated, and the last part of the conclusion was missing.**
For the entirety of my career in the wireless space, I have always worked for a small company selling something to or through wireless network operators. As such, I have made a good number of friends who work at these various carriers. I have observed that the most reliable way to get their dander up is to casually insert into the middle of any conversation, “Well, it doesn’t really matter because you are eventually just going to be a dumb pipe anyway.” Then I sit back, sip my mojito and watch the ensuing rant. Fun times.
Last week I tried this with a friend of mine who works at a carrier, and he said, “I prefer to think of it as ‘an open marketplace of ideas and innovation.’” Attitudes on this topic are changing. In fact, there is great enthusiasm for what most people agree will be a reduction of friction if we all embrace the word “open.” This got me thinking. First off, my friend is absolutely correct: When cast in a slightly different light and without the derogatory descriptor, a “dumb pipe” has the potential to be a very good thing. That simple realization led me to wonder how good it could actually be, and specifically from a financial standpoint. Could a major wireless carrier flip a switch to full “dumb pipe” mode and in so doing, take massive operational cost out of their equation and increase their value overnight?
What follows is an admittedly rough analysis of that possibility, so the conclusion may only be within spitting distance of the truth, but if it is even that close, I am very surprised by the conclusion.
BASELINE
I started by looking at the financials and structures of some of the world’s largest wireless carriers. There are certainly some differences between them, but at a high level, the structure is essentially the same. I could have used any of them as a model, but for my exercise, I happened to use Sprint’s publicly available financials from 2004 which I found to have a very readable structure and some useful details. (Also, I looked only at the wireless division, and not the consolidated financials.)
At the end of 2004, Sprint had 24.7mm wireless subscribers, 7.7mm of which were also data subscribers. Here is the revenue and margin analysis for that year: (numbers are in millions)
Net Operating Revenues $ 14,647 100%
Operating Expenses
Costs of services and products 7,096 48.4%
Selling, general and administrative 3,406 23.3%
Depreciation 2,557 17.5% (includes amortization)
Amortization 6
Restructuring and asset impairments 30 .2%
Total operating expenses 13,095 89.4%
Operating Income $ 1,552 10.6%
Adjusted Operating Income $ 1,577 10.8%
Adjusted EBITDA $ 4,140 28.3%
These are solid financials, and a 10% operating margin is quite respectable. Here is some other important data:
- 26,288 employees
- $557,000 in revenue per employee
- $2.5 billion in capital expenditure
- ~$62 ARPU
- ~15,500 retail sales outlets
- 800 branded stores and kiosks
- 17.5mm direct customer base
The gross add distribution mix is interesting and also an important part of the analysis:
Stores and kiosks: 38%
Alliance partners: 22%
Other direct: 18%
3rd party national/local: 22%
ANALYSIS
So now let’s make some assumptions and do some back-of-the-envelope math. First, I define “dumb pipe” as “branded access” with the closest analog being an ISP. I also roughly estimate the employee functions into four buckets, and used percentages based on anecdotal information from a few sources at different carriers:
Salespeople 20%
Customer Care 40%
Network Operations 25%
Management and HQ 15%
(The first time I heard that it is not at all unusual for a present-day carrier to have Customer Care represent 40% of their employee base, I was astounded, but it is fairly standard across the industry.)
Now then, how would turning a present-day carrier into what would essentially be an ISP change our key metrics? First of all, all the company-owned stores and kiosks would close. (Have you ever seen an Earthlink store? No.) This is not to say that all the salespeople go away, but there would be a shift to consumer commodity sales practices, relying more heavily on retail channels like big box retailers. Direct sales would be limited to large accounts and “1-800” ordering. Let’s say roughly half of the sales personnel go away. The effect to overall sales would be negative, and I will assume it isn’t effectively replaced. So that means 38% fewer gross adds, which shrinks the subscriber base.
A smaller base would obviously mean fewer Customer Service Reps, but how else might that cost be affected? I will assume a “Bring Your Own Phone” model, where all phones are unlocked, the consumer buys it at full price and chooses whatever carrier they want. It is reasonable that the large percentage of calls to Customer Care having to do with the device itself will go away, or more to the point that the carrier will make them go away through effective communication and education. Furthermore, our “dumb pipe” carrier will push “no-frills” plans for people who are smart enough to operate their mobile phone most of the time and don’t need to call Customer Care. Built into our marketing message will be that our customers are paying less per month for our “Do It Yourself” service. I will further assume the large number of all calls about the application they downloaded that won’t work will go away, because there is no storefront and no editorial function. Billing issues will still arise, so let’s say we cut our CSRs by 50%. This will also mean all the business development and product people who deal with application publishers and content providers will go away. So we can cut a few people out of HQ. Network operations basically stays the same.
The marketing expense of any current-day carrier is very large, and typically consists of an ongoing national television branding presence, an ongoing print presence and channel support. Given our reliance on channel partners now, I think I will keep the national television spend, kill the full-page newspaper ads and shift that part of the budget to in-store support, cutting a conservative 25% of our billion dollar budget in the process. I am also dropping the consumer price and killing the myriad incremental price plans. I like $50/month, all you can eat.
To summarize, we would lose 38% of our gross adds, bringing the total subscriber base down to 24.2mm. Total revenue obviously decreases given our new, lower price and smaller base. I lowered the COGS to reflect the historical 48% range then took our marketing savings ($250mm) off the top. I took our cost savings on sales, customer care and management (32%) out of SG&A. To be thorough, I upped our restructuring to $100mm to cover severance, etc. Here is what our new financials look like:
Net Operating Revenues $ 12,100 100%
Operating Expenses
Costs of services and products 5,558 45.9%
Selling, general and administrative 2,316 19.1%
Depreciation 2,557 21.1% (includes amortization)
Amortization 6
Restructuring and asset impairments 100 .8%
Total operating expenses 10,537 87%
Operating Income $ 1,563 10.6%
Adjusted EBITDA $ 4,126 34%
CONCLUSION
It really is not a compelling case. I shared this scratch analysis with a few friends who work at carriers, and they all agreed that it roughly made sense. I fully expected there would be more cost savings and more dramatically improved margins. The most interesting thing to me is that the bottom line is virtually unchanged in this exercise. Wireless carriers are generally very well optimized. The problem is that this is a very capital-intensive business. COGS and Depreciation are the big costs, and they are due to the high cost of spectrum, network equipment and operations.
For a final comparison, I took a look at Earthlink’s 2007 financials. As I mentioned earlier, as an ISP, they are the best example of a branded “dumb pipe.” At 5.3mm subscribers at the end of 2007, it is a much smaller business than some of the largest wireless carriers, but look at the leverage in their model. They generated $1.22 billion in revenue with just 996 employees. That is $1.2mm per employee, more than twice as much as our carrier example. Their COGS is 35%, so a little lower, but their total SG&A is higher at almost 50%. Operating income was a comparatively thinner 4.1%. Earthlink has swung to a loss in recent years due to the underperforming Helio investment, but even if you back that out, the “dumb pipe” model, even if it does create a vibrant atmosphere of innovation, is less attractive not least because such a company does not capture the value of that innovation itself. On top of it, the expected cost saving does not appear to materialize by handing the greatest driver of value – the services that consumers want to access – to 3rd parties. Lastly, if every carrier embraced such a model, I would also expect to see downward margin pressure as consumers perceive the commoditized “branded access” model as undifferentiated, which would increase competition on price, driving ARPU down.
Like I said before, this is obviously very unscientific, and maybe I missed something. But if the broad brushstrokes are about right, and the assumptions are close to reality, and the comparison to analagous industries is sound, then it is worth questioning whether the enthusiasm for “open access” is a catalyst for very welcome innovation in our industry or just a race to the bottom for the incumbents where the ultimate winners are today’s internet titans. Maybe it is both. Time will tell.
