RealTime IT News

How the Wireless Frontier Will Be Won

We heard Francis McInerney, co-author of Futurewealth (see our review of the book here) and venture capitalist with North River Ventures, speak at the Stupidnet conference, where we also picked up a copy of the book. A few months later, having read the book, we called him to see what he might predict for the future of wireline and wireless telecommunications. We published the wireline interview (The Telecoms Future) yesterday.

Whereas in the wireline telecoms industry, McInerney feels nobody "gets it," he feels that one company is doing wireless right: Verizon Wireless (and it's not just the really good TV ads).

The key reason is a very basic one. "Cellular companies don't own cell towers," McInerney notes.

On the other hand, the companies that do own towers, like American Tower, are laden with debt, much like the RBOCs they resemble. But McInerney expects their fortunes to improve in the long run.

A wireless boom

He expects demand to take off. "I used to say that every home's a TV station, and when I did I was hauled across the carpet. It could be true today, but there are no uplinks."

When demand takes off again, the wireless companies will be better situated to take advantage of the new boom. In the book, the authors write, (p.238) "digging ditches still costs more than ever, placing wired companies at a disadvantage."

At the moment, McInerney believes that WiMax could trigger the demand surge, if it's done right. "If WiMax allows a distributed antenna technology, you will see street lamps backhauled to every cell tower."

At that point, the local WISP, supplying access from every street lamp, suddenly becomes the key connectivity player. "It's complete telco bypass. Existing incumbents might not exist at all in the future. If the finances were available, it could begin in 24 months. There are plans in some places (cities, not nations) to make it happen in 12 months."

Once it starts, there's no stopping it, he says. "One the conversion starts, and the technology and business plan are proven, you get a land rush."

The key concept here is what McInerney calls the "Moore Curve." Moore's Law is a prescient observation by Gordon Moore, Intel co-founder, who, in 1965, said that CPU capacity would double every year. More recently, it has doubled every 18 months, but the principle holds true.

McInerney says that the effect of Moore's Law is that it forces companies and even entire industries to innovate. Those that don't get left in the dust, like Kmart trying to catch up with Wal-Mart, or Compaq trying to compete with Dell.

It's the customers, not the technology

WISPs will not succeed because of their understanding of the technology, McInerney suggests, but because they understand their customers. Many of the companies we admire for their engineering skills, like HP, will fall to companies that are better at sales, like Dell.

At ISP-Planet we have looked down on Cisco as a buyer of technology instead of an inventor. McInerney tells us we're wrong to do so. "It takes Cisco 8 days to turn a sale into cash. At the time of the telecoms crisis, it took 39 days. John Chambers is brilliant. Everybody knew he was a great leader during growth times, but telecoms now is in a no-growth phase. He's really tightened up on operations. You can only get down to 8 days if the organization is tuned so that the responsiveness is almost immediate."

He says Cisco's competition (with the exception of Juniper) takes far longer to turn a sale into cash, citing a few numbers: Nortel takes 69 days, Ericsson 92 days, Alcatel 86 days. "What [these numbers say] is that if you want to know how Cisco is beating you, it's because they manage inventory, payables, and receivables better than you."

McInerney grades the carriers in the same way. "I grade them ABCDEF on this metric. The only one that's a crack carrier is Verizon Wireless."

"But can Cisco keep buying," we ask.

"Yes. There are two schools of acquisitions. There's the Chemical Bank schoolbuy a company every 18 months or two years, and consolidate to gain market share. But Cisco's not acquiring for market share. It's acquiring expertise it doesn't have, almost always in a specific product area."

McInerney argues that as long as the marketing machine works, acquisitions are safe. "Assuming that two parts work well, the CFO part and sales and marketing, acquiring technology is relatively risk free."

Oddly enough, that means that wireless equipment vendors that are doing sales and marketing right will acquire those that are stronger in engineering. Look at the market, and you'll see that this has been happening for some time, as when Floware bought BreezeCom and called the combined entity Alvarion.

Similarly, WISPs need to do anything but buy each other. One successful WISP told us, not for attribution, that he stopped buying neighboring WISPs. "I just build out until they're on the edge of my coverage area, and then I wait for them to fail." Then he either builds out himself, or buys some or all of the WISP assets out of bankruptcy court.

Don't expect consolidation on the Chemical Bank model. (We expected this, see US Wireless Online, AIR2LAN Merge and A Big WISP Merger, but we were wrong.) Do, however, expect fierce competition. There will be big rewards for the survivors. McInerney provides some rough numbers.

"Take the number of consumers in the world, say 1.2 billion, and divide by four to get 300 million households. Maybe there are 1.5 TVs per household, so there are 450 million TVs. But if everyone is a TV station, and the Moore Curve puts devices into peoples hands, people, including children, could have four or five devices each. We may be talking about the low hundreds of millions to tends of billions of devices sold each year in a short period of time."

If that happens, those WISPs with a local reputation for customer service and the ability to respond to change will win like Cisco won in the big wired buildout. It is not clear which equipment vendor would win, but McInerney suggests that the smaller, more nimble companies like Alvarion and Trango and Orthogon and Vivato and Aperto should not ignore the bigger players like Motorola and Cisco if those bigger companies prove to be nimble.