Monday, November 19, 2007

Will there be a 3rd Pipe?

In the top 30 MSA's (metro service areas), I do not think there is enough demand for a 3rd broadband pipe in the consumer market. M2Z, Frontline, and Google have designs on a nationwide wireless network (based the 700 MHz spectrum to be auctioned off in Jan., 2008).

EarthLink has failed at the Muni Wireless play. Yes, it was mainly free with a premium offering that about 3000 people subscribed to, but some of the issues that marked its demise will be felt by other providers as well.

  1. Wireless Network Design / Build demands RF Engineering even for licensed spectrum.
  2. 700 MHz doesn't offer much more than the 900 MHz or the 2.4 GHz spectrum, except maybe less interference.
  3. It always costs more than planned!
  4. What's the marketing pitch? What Pain are you trying to fix? What pain is left that Clearwire, Sprint / Xohm, EVDO, 3G, T-Mobile, TowerStream, et al don't already try to address?
  5. How will you differ from a WISP like Clearwire?
  6. Free only gets you so far. At some point, you need revenue to get ROI (and to pay to operate and maintain said network). How will you acquire paying customers?
  7. What handsets or devices are you going to capitalize?
  8. How will you handle customer care / tech support at $30 per call? When Joe calls with his Zire 72 that can't get a signal, how do you keep him from becoming disgruntled?

MSOs and Telcos offer triple and quad play. So you have wireless broadband and VoIP. I hear all the buzz about IPTV, but Aloha couldn't get it off the ground, how will you? (Content costs big money. And if you thought ILEC's were a PITA, try Hollywood!)

The way I see it, here are the niche plays:

  1. The 3rd Pipe is still needed in rural and suburban areas.
  2. Businesses are always looking for a more flexible communications partner in addition to a redundant link.
  3. MDU (apts., condos, resorts, subdivisions, greenfield) and MTU (office parks & bdgs) - will give you a captured audience in a specific geographic area that you can interact with and offer flexible services to meet demand. (The PCO model, kinda).
  4. A really cheap offering, maybe done with a rechargeable pre-paid model.

With Embarq, Sprint, T-Mobile, Cbeyond and others offering FMC, UMA and femtocell, the 3rd pipe gets less likely. Once this gets better footing (and marketed so the average ID10T can figure it out), this will satisfy most consumers, SOHO, and very small businesses. A riskless combo of convenience from a trusted brand.

The consumer broadband, voice, and TV markets are a Bloody Red Ocean, that will drain revenues and cash reserves for the next couple of years for telcos, MSO's and DBS. (Ask DISH how it is doing against FiOS).

Unlicensed spectrum in the major MSA's is almost "full". The noise and interference increases daily with more and more WAP's and unsecured WLAN's popping up. The addition of FON, Meraki, Muni wi-fi plays, HotZones, and other wi-fi technology doesn't make the case any better.

That doesn't mean the 3rd Pipe won't work in regional pockets. I think it can work regionally. It will be very hard to do on a nationwide basis. It gets harder if Clearwire gets its act together -- or if Clearwire partners with Google or Sprint or T-Mobile.

The business case by the numbers:

One city can cost $10M to $27M to build out. Let's take $12M. Not backhaul or backbone or tower leases or operation. Just the MAN network installed.

Let's say we offer BB Lite at 1MB for $19 per month and unlimited VoIP for $25 per month. That's $44, which is average MRC, right now. The offer needs to be less than current solutions to get traction (or market share). At $44 and a 3 year ROI, 7600 users are needed per MSA. ELN only captured about 3000 combined. And why 3 years? 3 years is a long time.

Now let's say that our VOIP costs are $8 per user. Our internet costs about $6, like dial-up. That's $14 in hard costs. Rent for AP's. Power. Internet Backbone. OSS. Billing. Admin. Tech Support. NOC overhead, like collocation, hardware leases, labor. Lets call all that $15 for now. We are at $29 from $44, leaving $15 of gross profit. But not at customer # 1, nor at customer # 100.

What about customer acquisition (CAC) costs? Advertising, Salespeople, Commissions, Referral fees. VoIP provider SunRocket saw acquisition costs of $200+. That's $15 per month for a 14 month "break-even". Now, at $15 gross profit, it will take 36 months to pay back the $200 CAC plus get your ROI on the $12M. that means you will need ((12M/(36-14))/15 = 37,000 users. There are very few Independent NSP's with 37,000 users in one MSA. (And that is if all of them are acquired in the first 6 months or the ROI falls off.)

Is there a CPE cost? Or will a cheap wi-fi adapter work? Installation required? Tech Support required to walk the user through install or sign-up or use? If yes, add costs. (More customers needed).

Take Rates.

In its first year offering triple-play, in 2006, Comcast had a 6% take rate. [dslreports]. That is an established giant with a big advertising machine. 6%. VZ's Quad-Play has a 5% take rate. Have you seen the marketing full-court blitz VZ applies? 5%. Those numbers would be great for a third pipe, but that take rate took an established brand millions in advertising to achieve in its first year. Oh, yeah, and how many were already customers that just added addition service? How many consumers are locked into a multi-year telco bundle?

As more users utilize the network, it will need to scale. Access Points can only cater to a finite number of users and can only offer finite bandwidth. Even the backhaul has a finite limit. Internet backbone will also need to scale with usage. Latency and jitter (QOS) will need to be maintained if users will be using it for video and VoIP. This escalates costs. Quickly.

And we didn't factor in if the 3rd pipe had to BUY spectrum. Add another $5B for that. Plus the network build costs increase, because 700 MHz equipment isn't mass market yet. (CPE will be required as well, since current wireless, EVDO, handsets are not equipped to take advantage of 700 MHz). Increased acquisition costs.

The other factor the business plan will have to take into account is that broadband penetration has almost flattened out. In other words, every household with a PC that wants broadband has it. There are about 30% of the households without a PC. And still millions that think dial-up is fine. (Or that broadband is more than dial-up or unavailable in thier area). In a flat market, take aways are required. Take aways are expensive. (Or you have to develop a Purple Cow, like a $100 wireless internet device.)

It isn't just me either. See this breakdown. Even Om doesn't think big pocketed Google will try to build a 3rd pipe. Or buy Sprint. I'll finish on this quote from GigaOm:

AT&T Mobility CEO & President Ralph de la Vega is not too worried about Google and its wireless ambitions. “Running a wireless network is a capital-intensive business,” he said in a chat earlier this week. “It’s not a business for the faint of heart.”

No comments: