Estimating Operating and Capital Budgets

Many excellent books on budgeting for business enterprises have been written, so I will not
attempt any extensive description of the process here. Instead I will indicate the major items
that should appear in the capital and operating budgets of the wireless broadband operator.
The capital budget will include the major infrastructure equipment that will be necessary
to launch the network and will be amortized over time. Often, perhaps most often, the actual
radios will not constitute the biggest capital expenditures.
If the network operator elects to use licensed spectrum in part or entirely, then licensing
fees may well constitute the biggest capital expenditure, and these will vary enormously
according to the spectrum sought and the circumstances under which it is purchased. Participating
in government auctions for first rights to newly opened spectrum can be ruinously
expensive, and obtaining spectrum as a distressed asset from the unfortunate participants in
prior auctions can be a steal. In general, licensed spectrum for use in public broadband net- works is not too generally available, but one never knows what will turn up in a given market.
It always pays to check federal records to find who has what spectrum in the locale one has
chosen to operate and then to explore the availability of such spectrum.
If the network operator is attempting to function as an ISP as well as an access provider,
the network operator may have to invest in large routers, black-box encryption devices for creating
virtual private networks (VPNs), server farms for Web hosting and Web site mirroring, an
IP voice softswitch, multimedia servers for distributing specialized content and service offerings,
and various suites of element management software for running the individual devices,
plus the additional servers and workstations to host such software. In addition to element
management software, the network operator will need provisioning software, customer relations
database software, network-attached storage, billing software, and perhaps a global OSS
system to tie it all together. And because everything has to be super reliable and redundant, at
least some of the operational facilities will have to be mirrored, and backup power must be
available that is sufficient to run the facilities, not for hours but for days. All of this together
is apt to run well into the six figures. Table 4-1 lists typical capital expenditures of wireless
broadband.
Today much networking equipment is priced on a per-port basis so the network operator
can add facilities incrementally and cost effectively, but the capital expenses can still be considerable,
particularly if one is attempting to emulate or surpass incumbent wireline operators.
And that’s not even counting purchases such as office furniture and equipment and business
condominiums.
Many of the companies attempting to establish wireless broadband networks in the past
focused on providing basic access to underserved rural markets and thereby avoided really
heavy capital expenditures, but that strategy may be inadequate for the future. Increasingly, all
successful networks will offer converged services, and those operators who will not or cannot
do so are going to become irrelevant.
Obviously, the network operator will have to secure backhaul, which will either entail
building one’s own links, wireless or otherwise, or entail leasing capacity. Wireless links will be minimally several thousand dollars apiece apart from the cost of securing roof rights, and millimeter
microwave links may run into the several tens of thousands of dollars to purchase.
Free-space optics are apt to run somewhere in between.
Operating expenses will include staff payrolls, normal business expenses such as utilities,
office leases, insurance, auditing, and costs more directly associated with the core mission of
the business such as maintaining equipment, advertising service offerings, maintaining a fleet
of vehicles for field operations, and so on. If the network is maintaining a large data center,
electrical requirements may be considerable because data and telecommunications equipment
is extremely power hungry. Table 4-2 lists typical operating expenditures for a broadband
wireless network. Type of OPEX
Existing Operator:
Investment Required?
Greenfield Network:
Investment Required?
Roof/tower lease Yes (partial) Yes
Base station and O&M Yes (partial) Yes
CPE O&M Yes Yes
Network O&M Yes (partial) Yes
NOC O&M Partial Yes
Leased line rental Yes (partial) Yes
Office expenses Partial Yes
Advertising/subscriber acquisition Partial Yes
Facilities Partial Yes
A million U.S. dollars is not an excessive sum of money just to get started in a single municipality,
and funding an ambitious build in a second- or third-tier city requires a war chest of
many millions. One may hope to bootstrap a wireless broadband business with the revenues
from one’s customers, but it is difficult to cite many examples where this has been successfully
accomplished.
Countless underfunded wireless ISPs (WISPs) have come and gone in the United States
over the course of the last five years, probably on the order of several thousand altogether.
Taken in aggregate they demonstrate powerfully that one cannot launch a public network simply
by taking out a second mortgage on one’s house and expect to succeed. Instead one must
secure serious investment and patient investors who do not expect sudden salvation from a
timely initial public stock offering.
I cannot outline in detail here procedures for courting such investors. I will say that
network operators proposing to use unlicensed spectrum generally have a far harder time
attracting investment capital than those holding licenses because unlicensed spectrum itself is
not valued and is not an asset that can be disposed of if the network fails. The best course to follow
in securing capital is to devise a highly detailed business plan that goes into specific service offerings and the necessary expenditures to support them and that indicates convincingly how
they will be delivered profitably over the wireless infrastructure the network operator has
selected. Avoid speculative leaps about “paradigm shifts” in subscriber practices or unfounded
projections concerning demand for bandwidth, and try to base as much of the plan as possible
on demonstrable fact. The plan must be flexible simply because high-speed access is such a
volatile business, but at the same time it must embody clear objectives, not just in terms of
achieving revenue goals but in building a business that will attract and retain desirable
customers.