Amol Sarva: How To Make An MVNO Work

Last week Disney announced that it was abandoning Disney Mobile, its second failed attempt to launch its own "virtual" mobile network (MVNO). The question we asked: Can anyone make an MVNO work? Yes, says Txtbl CEO Amol Sarva, who was also a co-founder at MVNO Virgin Mobile, which is about to go public. We asked him to make his case; he takes us up on it below.

Not every MVNO ever created in the US has flamed out--just the ones that don't know how to run an MVNO business.

Disney Mobile announced their departure from the market last week. It's a nice bookend to the even higher profile shuttering of sister-company Mobile ESPN last year. And who can forget the flameout (and helicopters) at Amp'd? Or, for that matter, the scarily large sums of money being invested by Helio for not so many customers. It makes for a tidy story: MVNOs don't work.

Well, there is a "lone exception," of course: Virgin Mobile -- with nearly 5 million customers, over $1 billion in revenue, and over $75 million in EBITDA for the first 6 months of the year -- is going public next week.

But do a cursory bit of research and you'll find that Virgin is only the 2nd largest MVNO in the US. The biggest is Tracfone, owned by America Movil. Carlos Slim Helu -- the world's richest man and the guy behind AMX -- picked up Tracfone back in 1999 and has since grown it to nearly 10 million customers and nearly $200 million EBITDA in 2007.

And then there is Boost Mobile, the original Virgin Mobile copycat, launched in partnership with Nextel back around 2002 and now wholly owned by Sprint Nextel. It has nearly as many customers as Virgin.

Qwest, the 3rd largest phone company, has an MVNO too, so it can offer bundled landline and wireless service. It has more than 800,000 subscribers. Embarq, the Sprint spinoff, launched an MVNO last year and is adding customers quickly. PagePlus runs a big MVNO on Verizon's network; it has about 1 million customers.

And as far as startups go: Kajeet's kids MVNO raised about $40 million last month, Greatcall did the same for their old people's MVNO, and Movida did about that for their Latino prepaid MVNO. And there continue to be some slightly adventurous new MVNOs marching bravely off to battle: Sonopia's affiliate-driven MVNO and Europe's Blyk which offers free minutes for ads.

In all, there are about 20 million U.S. MVNO subs -- nearly 10% of the US mobile phone market.

So much for the great MVNO debacle. So what's the difference between the helicopter-riding disasters like Amp'd and diligent, patient machines like Tracfone?

Success as an MVNO means playing to your strengths. You are buying wholesale access to someone else's network, so there will be things you can't do well.

For example, a badly conceived MVNO does the following:

1. Targets customers that carriers themselves are trying hard to get (e.g., young tech geeks who love sports and music, and spend a lot).

2. Launches a not-very-distinctive product proposition that focuses on all the same shiny things that carriers are themselves offering (e.g., NFL and ESPN on Sprint, Verizon VCast with Madden, MySpace on Cingular)

3. Builds hugely expensive new product platforms with loads of capex and opex (e.g., build a "mobile media company" from scratch complete with studios, Jack Black on retainer, and so forth)

4. Ignores retailing 101. Launches big glitzy media... with no or little store availability (Amp'd, Helio, Disney, and Mobile ESPN all did this -- Superbowl and MTV ads before they were even available in 100 stores nationally)

5. Therefore acquires customers expensively. Industry standard is $300-500 per customer. Amp'd, ESPN & Co were closer to $500-1000.

6. Gets distracted by the "amazing ARPU!!" of people spending $100 per month with your service. Most smartphone users spend around that much. The "average" user in the US spends $50/month. Treo, Blackberry, high-end-type users spend quite a bit more.

7. And so doesn't realize that the revenue per megabyte or revenue per minute is nowhere near where it needs to be. MVNOs are wholesalers. They pay a fixed price per minute. The industry average is $0.07 per minute of revenue. When Amp'd was selling 1000 minutes for $40...they were making $0.04.

Smart MVNOs do, well, the opposite:

1. Target "bad" customers that carriers ignore, like the low-credit, prepaid customers that Virgin, Tracfone, etc. focus on.

2. Launch a differentiated product like... prepaid. Carriers have always made prepaid extra horrible.

3. Build cheap platforms. Virgin's phones original phones had black and white screens and zero music. They were just simple phones.

4. Focus like crazy on retail -- like selling in big national chains from day one.

5. Acquire customers cheaply. If you read Virgin's S-1, they spend about $100 per customer.

6. Focus on the right metrics: Revenue per megabyte or per minute. For example, if Virgin's frumpy prepaid customers spend $20/month to use 150 minutes... they are spending $0.13 per minute.

Now you know the secrets. Execute properly and you can buy your own helicopters.

See Also: Another MVNO RIP: Disney Mobile Going Down, SK Telecom Throws Helio A $270 Million Lifeline; EarthLink Backing Off, Txtbl: Mobile Email For The Masses?