From: Linda (joelinda1@home.com)
Date: Thu May 25 2000 - 11:25:40 PDT
From the May 2000 issue of Hambrecht & Quist: Connected:
http://connected.hamquist.com/
Deciphering the Riddle of Wireless Data
Edward F. Snyder
edward.snyder@chase.com
Nokia had set a lot of records in the last three years - the most phones
sold, the largest market share, the highest margins
- so naturally, we anticipated another string of impressive stats at the
company's Analyst Day in Dallas last December.
Still, when Jorma Ollia stepped up to the microphone, few expected he'd
swing for the bleachers. But swing he did, raising
Nokia's subscriber forecast to over a billion by 2002 (Exhibit 2). This
was bold, even by Finnish standards.
[Exhibit 2.]
Source: Nokia, Ericsson
His comments may have surprised many pundits, but they were simply
further confirmation of what the faithful have
understood for sometime: Wireless is no longer a niche technology
relegated to the "other" bin of access solutions. Like the
Internet, it has quickly become an alternative, ubiquitous channel for
reaching consumers. To be without a wireless
strategy is to be without a connection to the fastest-growing market
anywhere. Indeed, once heralded as the last word in
communications, legions of dotcom'ers are hastily beating a path to
mobile Mecca.
As with many pilgrimages, early believers that were once labeled crazy
have been vindicated. One of the best examples of
this occured in 1985. Moves once thought to be insane are now termed
insightful. Tongues wagged over Craig McCaw
purchase of MCI's paging and cellular holdings for $120 million (MCI
decided growth prospects for wireless were limited).
The purchase raised the bar, and more than a few eyebrows, given
comparable valuations. But what was an incredible sum
in 1985 seemed a bargain in 1994, when AT&T bought McCaw Cellular for
$18 billion (Exhibit 3). Clearly, in the ensuing
seven years, the pilgrims had moved a lot closer to Mecca. AT&T
cultivated its holdings, expanded its footprint, and with
McCaw's spectrum, emerged with the largest mobile footprint in North
America.
[Exhibit 3].
Source: Dow Jones, Company Reports, Chase H&Q
The company parlayed McCaw's gains of 1994 into a $73-billion market
capitalization when it floated its wireless tracking
stock, AWE on April 27 (Exhibit 4). Clearly, MCI was correct; the growth
prospects for wireless were limited, but that limit
was probably somewhere in the trillions of dollars.
[Exhibit 4. Chronological Appreciation of McCaw's Spectrum]
Source: Company Reports, Wall Street Journal, Chase H&Q estimates
Twenty-eight Years of a Three-hour Tour
Squeezing value out of spectrum will require more users, more content,
and more technology, but most important, it will
require a recognition that the scarce commodity here is spectrum, not
data. MaryAnn and Ginger notwithstanding, the
benefits of using spectrum to transmit reruns of "Gilligan's Island"
pale in comparison to those achieved using the same
spectrum for mobile telephony. When consumers got a taste of those
benefits, they flipped off the tube and beat a path to
the carriers' door. Regulators then beat a path to the auction blocks.
Subscriber growth boomed, and more spectrum was
cleared for sale. The problem, of course, is where the former is
accelerating, the latter will soon end.
Accommodating the ever-increasing demand for service using finite
bandwidth requires ingenuity, heavy investment in new
technologies, and most importantly, a salient recognition of the value
of spectrum. Clearly, value will increase as better use
is made of bandwidth. So much so that we believe the $73-billion market
cap for AWE will seem paltry a few years hence.
After all, most of the service and nearly all of the revenue for AT&T
Wireless is derived from mobile telephony, not data. In
fact, this is the case for nearly all carriers outside of Japan.
Subscribers are essentially paying to walk and talk - never
mind the promised benefits of mobile data, packet systems, and
location-based services, each of which could dwarf mobile
telephony in popularity and market value. Still, a phone in the hand is
worth 2 Mbits on the drawing board, and subscribers
are not waiting for the wonders of the wireless Web. This means that
when data finally is deployed, the frantic pace of
subscriber growth could actually increase.
Things Are More Like They Are Now Than They Have Ever Been Before1
It is no surprise then that history repeated itself in April when the
price for United Kingdom Universal Mobile Telephone
Service (UMTS) spectrum far exceeded expectations. The auction gavel had
barely sounded before tongues began wagging
again; $36 billion seemed a bit pricey, even considering the enormous
gains in subscriber growth posted over the last year.
UMTS spectrum may be overpriced, but only if it is used exclusively for
voice. If that isn't the case, then the Law of
Unintended Consequences suggests it's too difficult to say what the
total impact and true value of spectrum will ultimately
be. The rapid pace of development precludes us from seeing very far into
the future, except perhaps to catch a glimpse of
something big. Very big. Why? Because as fluid as our day-to-day lives
may be, our access to information is not - we sever
connections when we move. Twenty years ago, a trip to the supermarket
entailed complete isolation. Forget remote
downloading of a forgotten grocery list - we didn't even have voice
connections. Shopping was performed in a
communications black hole that included radio silence. Closing the gap
between our lifestyle and our access to information
holds enormous economic benefits that are reflected in the huge sums
being paid for spectrum. The fact that thus far
we've only bridged this gap for voice suggests there's a lot of money
left on the table in the form of wireless data. The
problem, of course, is determining what form the data will take and
which technologies will be deployed to deliver it. The
business case for wireless data is still a matter of some debate and a
topic we touched on in our October 2000 Connected.
Give Me Ambiguity, or Give Me Something Else
As with many prognostications, our view is obscured. However, it is our
opinion that technical chatter from some of the
hardware OEMs has gotten well ahead of the reality at the carriers, and
more important, we believe, the desires of the
subscribers. For as much sport as we derive from debating the merits of
Wideband Code Division Multiple Access (WCDMA)
versus CDMA2000 versus Enhanc ed Data for GSM Evolution (EDGE), it is
important to remember that packet service hasn't
even been deployed yet. Moreover, debates about data rates, modulation
schemes, and carrier bandwidths are really
ancillary to the core issue, and that is, subscribers are buying
benefits, not technologies. Users want the benefit of being
connected to their office e-mail client while mobile. Whether such a
service is delivered using narrowband CDPD, the RAM
mobile paging network, or wideband HDR is of nearly no consequence to
the customer - other than the form factor and
cost the technology choice forces upon them.
The salient point too often lost in technical banter is that subscribers
don't buy technologies, they buy service, or more
specifically, they buy mobility. Subscribers are signing up in record
numbers (Exhibit 5) in spite of the fact that per-minute
costs for mobile service are significantly higher than for landline.
[Exhibit 5.]
Source: IDC
Mobile cellular fetches a premium not because it has good voice quality
- too often it isn't - but because it's mobile. This
means that when faced with a choice between upgrading voice quality or
increasing coverage, carriers are likely to get the
biggest bang from their dollars by expanding footprints. Coverage
enhances mobility, which increases the value proposition
for the subscriber. Improving voice quality over a minimum acceptable
level does not (all else being equal).
The same dynamic holds for wireless data. Desktop services shoved
through a wireless connection to a mobile subscriber
will have little economic value. Content tailored to provide utility in
a mobile environment on a mobile device could fetch a
premium. Moreover, services that combine information on the user's
location with push data filtered through a personal
profile from the Internet could prove wildly popular. But this
popularity will be derived from the time and location value of
the information, not the bandwidth of the connection. Understanding why
can help decipher the riddle of wireless data.
Wherever You Are, That's Where You'll Be2
Information only has value if you can access it. As obvious as that may
seem, it's important to state it explicitly because it
points to the reality that the value of information depends on the
physical distance between the information and the
consumer. For example, a printed travel itinerary sitting on the desk
two feet away has more value than it does six blocks
down the street at the travel agent's office, and a lot more value than
1,500 miles away on a counter in the airport from
which you departed (Exhibit 6). Clearly, the further its distance from
the user, the less its value. The corollary is also true;
the closer to the location of need, the greater its value, with the
user's physical location being the point of maximum
benefit.
[Exhibit 6. Potential Value of Location]
Source: Chase H&Q
This, in essence, is the value proposition for cellular systems. They
put data in very close proximity to the subscriber where
it can be of greatest benefit. In doing so, they increase the value of
vast amounts of information that would normally be
worthless given it was too far removed from the consumer to be of
practical use. An interesting take-away of this is that
although we speak of the mobile user, it's really information that's
gaining mobility, not the individual. People have always
been on the move, it's data that has been stationary. Cell systems
essentially make the location of maximum value and the
subscriber's location one and the same by making data mobile. Seen in
this light, it is easier to understand that the value of
mobility is the location value of information and that opening all of
the Internet to subscribers may not be of any value if
the information is not easily retrieved or of particular use in a mobile
environment.
It's Always Later Than You Think
The value of information also varies with time. And just like with
location, there is a point of maximum value that
corresponds, in this case, with the moment of need (Exhibit 7). The time
value of information can be quite high and is the
primary driver behind the explosion in Internet use. Because it provides
quick access to large amounts of data, the Internet
can deliver data much closer to the user's moment of need when the
benefit is near or at its maximum. This is one reason
the service is so popular.
[Exhibit 7. Potential Value of Time]
Source: Chase H&Q
One important distinction between time and location is the asymmetry in
the time value of information. Unlike with location,
the time value of information drops rapidly after its peak. Once our
moment of need has come and gone, the importance of
the information we didn't have declines rapidly. The conclusion is that
it's much better to be early than late. If it's early,
the information value increases as we approach the moment of need. If
it's late, it's worthless.
It is also important to note that in its current form, the Internet
provides no location value because we have to go to it
instead of it going with us. Accessing the Internet is like going to the
library - only instead of a few thousand books, there
are multiple-millions, all with multimedia content and the ability to
access all the other books in all the other libraries in the
world. That and every tenth page has something to do with Pamela
Anderson. But just like a printed book, the Internet
does not provide any location value.
Getting More From Cookie Dough Than Cookies
If mobility delivers location value and the Internet delivers time
value, what can be said about the mobile Internet? Well, if
the two fastest-growing markets in the world (perhaps fastest ever) are
being fueled by location value and time value, it
seems plausible that their combination will yield unprecedented, if
perhaps more focused, growth.
[Exhibit 8. Delivers Value of Time and Location]
Source: Chase H&Q
This is because data that is relevant in both time and location can
deliver value unequaled by either the Internet or cellular
systems alone. The form of this data can vary widely, but the value to
subscribers will be uniformly high. Imagine walking
off a delayed flight to a notification of a gate change for the
connecting flight, and then receiving real-time directions to
the boarding area. As shown in Exhibit 8, Location-Based Services (LBS)
combine fast access to a wealth of Internet
information filtered through a personal profile with the subscriber's
location to amplify value well beyond that of the moment
of maximum value (time value) or the location of maximum value (location
value).
In addition to amplifying value, we believe LBS will cast a wide net
across a variety of utilities. Dominant applications will
likely include travel, eCommerce, public safety, entertainment,
education, recreation, and engineering. Typical of these is
the use of LBS to notify subscribers of sales on preferred merchandise.
Using the subscribers' personal profile and current
location, the LBS notifies the subscriber of a favorite item in close
proximity and can even forward an electronic coupon to
close the purchase. This approach enables the subscriber to avoid the
delay and shipping cost associated with the
Internet, derive nearly immediate gratification, generate
revenue-sharing income for the carrier, and result in a sale for the
retailer. Where the Internet flattens distribution for users that have
access to a computer, LBS does the same for users of
mobile terminals. Shortening the distance between supply and demand
improves the efficiency of transactions and benefits
everyone involved.
Still, to be successful, content and services must be crafted to exploit
the time and location value of information for the
specific user. Clearly, this will require some personalization, which is
already common in the Internet world, but not in
mobile systems. It will also require technology for determining
subscribers' actual location. Although there are several
systems in trial that provide location information to the whole terminal
or the base station, none of these technologies
have yet to be commercially deployed.
Conspicuously absent from these requirements is the need for an air
interface that will support high-data rates. Nowhere in
our discussion have we assumed broadband applications, since the mobile
Internet does not require high-data rate
applications to succeed, just as the mobile cellular system and the
fixed Internet have not. In fact, as the two
fastest-growing markets in the world, both channels have flourished with
effective average throughputs well below 100
Kbps. In the case of cellular, it's well below 10 Kbps.
Things Should Be Made as Simple as Possible, But No Simpler3
So where does this leave us then? What are we to make of all these
proposals for multimegabit air interfaces for a market
that's technically agnostic? What performance is required to make the
wireless Web a reality?
Forget about 3G, what about 2G, or more to the point, what about a
low-data rate IP technology that provides subscribers
basic packet service? At this stage, we believe the arrival of the
wireless Internet is being delayed more by the lack of
packet service than by low data rates. This is one reason Nokia's
High-Speed Circuit-Switched Data (HSCSD) system did
not find wide application. While it addressed throughput issues, it
lacked packet switching. Conventional wisdom suggests
that remedying this will require GPRS in the TDMA world and IS-95B 1XRTT
in CDMA. And given that these technologies are
currently being deployed, we'd expect demand for wireless data, and
content for the wireless Internet, to increase rapidly
soon after they become active.
This isn't to say that higher bandwidth systems will not be deployed or
prove popular. On the contrary, we expect EDGE
and 1XRTT to eventually be deployed throughout much of the existing
infrastructure. Both approaches provide more voice
capacity, support a greater number of data users, and do not require
forklift upgrades. Nevertheless, we believe the
success of the wireless Web has more to do with tailoring content and
services to the mobile users than is does with raw
bandwidth. And we expect most of the utility and much of the value of
the wireless Web can be provided using low
data-rate links, just as it is today in the wireline world.
That being the case, we do not see a compelling reason other than
perhaps regulatory for deploying new systems capable
of multimegabit data rates. Moreover, we expect the 80/20 rule to come
to bear on carriers and scribers who will be the
first to lay down real money to revel in the glories of the wireless
Internet. If 80% of the benefits of wireless data can be
had for only 20% of the cost of a full blown WCDMA or CDMA2000 system,
how compelled will subscribers be to spend
more? Nothing we've seen so far has suggested they will, and if
subscribers don't, carriers won't, at least not on a large
scale. Still, there are a lot of moving parts to the story and the jury
is still out. The final verdict may not be clear at this
point, but the evidence is: Time and location hold inherent value to
subscribers. Solutions that focus on technical
performance at the expense of these benefits will likely fail.
Epilogue
The opportunities presented by the advent of the wireless Internet are
by no means exclusive to carriers and subscribers.
Equipment OEMs will also benefit from a massive build-out or, at the
very least, upgrade cycle to enable packet data
service over what today is exclusively a circuit switched voice network.
As we saw in the mid-1990s during the transition
from analog to digital cellular, hardware OEMs have risen to the
challenge of fielding solutions for what they hope will be
the next mobile standard. When the stakes are high, the entrants are
many, leading to a deluge of competing technologies.
Raw performance aside, we believe cost-per-Kbit and time-to-market will
be especially important to carriers given the scale
of the build-out and the revenue opportunities available from wireless
data. Although a detailed review of each of these
solutions is beyond the scope of this analysis, we thought it useful to
include a summary of the various proposals by air
interface standard for a clearer understanding of the tenor of the
debate.
Exhibit 9. Proposed, Existing, and Imagined Wireless Data Technologies
View Exhibit 9 in PDF format here
Source: Yankee Group, Chase H&Q company reports
This archive was generated by hypermail 2b29 : Thu May 25 2000 - 11:23:46 PDT