Survey: Carriers Plan More Optical

Carriers are mulling marked increases in spending on DWDM and optical crossconnects, according to results of a recent survey by KMI Research.

KMI says it got Web-based responses from 141 carriers around the world to a July 2003 survey aimed at probing spending plans for a range of products, including Sonet (Synchronous Optical NETwork) and SDH (Synchronous Digital Hierarchy), DWDM, digital crossconnects, and Optical Crossconnects.

Carriers were asked to indicate whether or not they'd increase spending in 2004 on the different product areas. Results were compared with those from a similar survey KMI conducted last year at this time to test carriers' plans for 2003. The firm received about 150 responses last year.

As the following chart shows, spending's on the rise overall in the areas of equipment mentioned above. But in some areas the rise is more marked than in others.

When it comes to Sonet/SDH equipment, for example, 35 to 40 percent of last year's respondents said they planned to increase spending in 2003. This year, 40 to 45 percent said they'd increase spending. For digital crossconnects, 23 percent of 2003's respondents said they'd spend more, while this year, 43 percent said they'd spend more.

The biggest changes between last year's and this year's results were for DWDM and optical crossconnects. In 2002, 35 percent of KMI's respondents said they would increase spending on DWDM gear. This year, 59 percent said they'd increase such spending. Last year, 20 percent of respondents reported they'd increase spending on optical crossconnects. This year, 49 percent said they'd spend more on these devices in 2004.

In some ways, the results don't meet expectations. KMI analyst Michael Arden is puzzled by respondents' indication that they'd continue to spend on digital crossconnects, despite research that shows that segment of telecom equipment dwindling in sales.

Overall, the results indicate the carriers surveyed want to spend more this year than last. Whether they actually will or not remains to be seen, however. Further, there's another odd quirk in the results: Based on the responses to multiple questions, carriers say bandwidth growth on their networks year-to-year has continued to tumble since 2001.

KMI says the information isn't inconsistent, because even though growth is slowing, carriers are reporting on increases in bandwidth across all their networks. And they may be planning to spend more on gear for data and video services.

— Mary Jander, Senior Editor, Light Reading

solver 12/4/2012 | 11:23:45 PM
re: Survey: Carriers Plan More Optical I am not sure what the result says at all.
It seems to me that a pie chart for each equipment group showing % of people who want to:
spend more
spend the same
spend less
makes more sense than just showing % of those who want to spend more.
The message to me is that in 4 out of 5 categories, people expect no growth since less than 50% of those responded were positive about growth.
But this is uncertain unless we know what % think spending will be less in 2004.
materialgirl 12/4/2012 | 11:23:44 PM
re: Survey: Carriers Plan More Optical It would be helpful for these folks to correlate their survey findings from last year, about 03 spending plans, with YTD 03 actuals. While the mix data are interesting, the uncertainty over the overall accuracy renders these findings pretty useless. Also helpful would be context, as in what services the newer devices would enable.
dwilip 12/4/2012 | 11:23:43 PM
re: Survey: Carriers Plan More Optical I agree with Solver. The only thing that can be concluded from the data is that most carriers will NOT spend more for all categories except DWDM. Without knowing how much more or less they plan to spend, it is impossible to say whether overall spending will increase or decrease.

KMI's survey is seriously flawed, so any conclusions they draw are undoubtedly wrong.
lembergjames 12/4/2012 | 11:23:40 PM
re: Survey: Carriers Plan More Optical I've not read the actual report, but have browsed KMI's promotional pages. It appears that they might have done a bit better job than can be inferred from LR's article.
milliman 12/4/2012 | 11:23:38 PM
re: Survey: Carriers Plan More Optical I would be leary to invest any money in this report. What is it indicating? What these people would like to do? What it tells me is that carriers would like to spend more for DWDM equipment and less for all other categories next year. These are relative figures. I cannot plan a business on relative figures. I can plan a business knowing that cap ex may be up 7% next year.

The quality of market research still hasn't improved. If we continue to make decisions based on this quality of research we will continue to be doomed.
nbwaite 12/4/2012 | 11:23:36 PM
re: Survey: Carriers Plan More Optical Let's see:

Suppose there are 100 carriers and that
99% of them say that they will spend more
on widgets in 2003 than in 2002.

Okay, if widgets were in a popularity
contest or in a democratic election, then
from some such survey data we could
believe that widgets would win.

If we are selling widgets and want to know
if our revenue will increase in 2003, then
we are not told so much.

For example, suppose of the 100 carriers,
99 of them spent $1 on widgets in 2003 and
plan to spend $2 on widgets in 2003.
Suppose one carrier spent $10 million on
widgets in 2002 and plans to spend $0 on
widgets in 2003 (the $10 million caused
the VCs to put in a new CEO that doesn't
like widgets!). So, total widget
expenditures in 2002 was $1,000,099. Total
widget expenditures in 2003 will be just

So, for 2003, we could say with great hot
air bubble blowing excitement: "For 99% of
carriers, widget expenditures will be up
by 100% in 2003!" Still, total widget
expenditures stand to fall from $1,000,099
to just $198. So, the widget market fell
like a VCs portfolio after the bubble

Moral: When we ask survey questions, then
need to have in mind what real, useful,
solid purposes we intend to do with the
data. Asking questions just because they
are common in surveys, and possibly have
been useful somewhere, e.g., in analyzing
democratic elections, promises not to be
very useful.

That is, to be useful, market research and
survey analysis really cannot be just
throwing around various methods but need
some sense to them.

For some sense, in business, we should
likely start with our customers and begin
to understand what they need; we can start
by asking them 'where it hurts'.

Okay, maybe our customers want to know
what the widget market will look like in
2003; our customers are not interested in
predicting the outcome of a democratic
election; instead, maybe our customers
want to make decisions on marketing,
production, development, research,
ventures, M&A, etc. in the widget market.

For something really simple, maybe our
customers want a good estimate of total
widget expenditures in 2003. Okay, we
will want to address some questions: (1)
What is a 'good' estimate? Do we want
unbiased, minimum variance, maximum
likelihood? (2) What data do we have
already and what data might we get? (3)
How can we use such data to make our
'good' estimate?

But, not everything powerful and valuable
is simple: In principle, we might want
more than just an estimate of widget
expenditures. E.g., we might want an
estimate of the distribution of widget
expenditures. We might want estimates of
earnings from different candidate
decisions based on the data. Etc.

Broadly, our customers will be getting our
data to contribute to business decisions.
The framework of 'statistical decision
theory' could be relevant here.

Norman B. Waite, Ph.D.

Network Architectonics

9 Fox Run

Wappingers Falls, NY 12590

[email protected]

Sign In