Studies & Opinions
of Warner’s “Branded Domains vs. Generic Domains”
April 19, 2007
I confine my review to the article’s concluding section, the
part dealing with valuation.
one were to believe the claims made about revenue multiples, buyers
and sellers would be clueless about value and there would be no
marketplace. The article suggests that most sold domain names are
undervalued, with market prices failing to incorporate the value
of brand names.
the article’s criticism of the use of revenue multiples in domain
name valuations has neither empirical nor theoretical foundation.
Moreover, the claim that branding is the major factor behind domain
name value is unsubstantiated, and the article fails to suggest
an alternative valuation model or to acknowledge widely used alternative
The article claims “the
domain sales aftermarket places a high importance on the value of
the revenue which domain traffic generates. In many cases the only
valuation considered is a revenue multiple of the domain traffic
revenue.” It is not clear how this claim is arrived at. Our statistical
estimation results show a wide range of brand-to-traffic
(B/T) ratios, suggesting that the branding value outweighed the
traffic component in a large number of domain sales. But the article
fails to mention such studies, let alone address their methodology.
“Of course, [revenue multiples are] a very poor way to value domains.” Why
is it so? The statement criticizes the methodology of estimating
value, the way, not measurement issues related to the use
of the multiple. However, the use of multiples
as, say, equity valuation benchmarks — namely, the price-to-earning
(P/E) ratios — varies by industry and country.
is conceptually only the rent collected from the land, as domains
represent the land of the internet.” This statement, at best, uses
a bad analogy of domain names with land. For, say, commercially
zoned land, the expected rent is the most important determinant
of the revenue side of value. If two identical adjacent
pieces of land are to be developed into, respectively, Trump Plaza
and 101 Bay View, the value of the former may well be higher because
of the Trump brand effect. As a result, any future buyer who is
able to retain the brand name will pay a multiple of the premium
rent. Hence, revenue multiples can be meaningfully used for branded
land, whether physical or virtual.
valuing real estate buyers do not just look at the road traffic
near the land they are considering buying. They also look at the
neighborhood, previous use, and what the zoning is — which are all
reflections of the potential future use of that land.” True, but
what does this have to do with the appropriateness of the use of
multiples? It seems to support my earlier argument that different
real estate markets have different revenue-multiple benchmarks.
“It takes an exhaustive and lateral view to
understand the true value of a domain, and traffic revenue is only
one of the key criteria.” Recommending an exhaustive view to determine
value is, at best, unnecessary. One needs to use valuation models
based on systematic factors, namely, branding and traffic for commercial
domain names, rather than include all the idiosyncratic factors. An exhaustive street map
for a car trip from San Francisco to New York does not add value compared
to one with only the major highways. To be useful, a model also has
to be elegant, beautiful in the way that mathematics, the ultimate
science, is. Furthermore, when one takes an exhaustive view, would’t
a lateral view be a subset?
greatest value attributed to a domain is not its traffic revenue;
a domain’s greatest value attribute is its brand value.”
Based on the B/T ratio, this is an inaccurate statement. It is true
that, other things being equal, a domain name representing a brand
name will be more valuable. But, from an investor’s perspective,
the return on investment is more important. Moreover, one can argue
that using a well-recognized name does less for brand value than
creating a strong brand name does. Furthermore, for the branded
company to establish revenue growth it has to generate growth that
is unexpected; otherwise its value will, at best, stay constant.
“Respectively the problem with the sales value
of branded domains is that they generally have very confusing or
poor attributable brand value.” It is not clear what is being referred
to as a problem. The sentence appears to say that branded
domain names are undervalued, and/or that the estimation error of brand value is higher
than that of traffic. It is possible, however, that this is an empirical
question and needs to be validated.
A large portion of domain names sold are generic traffic names for
which appropriate multiples reflecting domain name risk can be used
to obtain a reasonable measure of value. Moreover, if the brand
name in a domain name is well established, the revenue generated
by the associated business will manifest itself in revenue. Thus,
revenue multiples are also a reliable measure for such domain names.
problematic valuation issue is associated with a domain name’s brandability
potential, which is a million dollar question. This is where advertising
and naming agencies use serious science and art to come up with
names that have high brandability potential. Hence, one should not
expect a $20-40 appraisal or the use of any multiple to determine