Database Reference
In-Depth Information
Web, there is plenty of scope for revenue generation. Again, this is a business model
that is likely to appear later in the development of the Linked Data market.
8.2.7 S
ervice
b
undleS
and
a
ggregation
Linked Data could be provided as part of a bundle of services to add value to other
offerings or be offered as a sweetener to other data sales. By providing the links to
third parties' data along with your own Linked Data, the purchaser of the Linked Data
gains access to not only your data but also the third party's (in terms of integration,
if not license). In effect, you are selling the fact that you have done all the data integra-
tion work for your customer; taking this idea further, you become a data aggregator.
8.2.8 b
randing
or
“l
oSS
l
eader
”
Companies such as the
New York Times
3
have published Linked Open Data, moti-
vated initially by wanting to understand the technologies and position themselves as
forward-thinking brands. The loss leader idea overlaps with branding as the Linked
Data set can be provided for free as a way of promoting the brand and as an entice-
ment for customers to buy the richer or higher-quality full dataset. This then moves
us into the “free” part of the “freemium” business model.
8.2.9 S
ubScription
a
cceSS
: c
liMbing
the
p
ay
W
all
Subscription is a well-known business model, both for content and services; how-
ever, for Linked Data, if it is completely hidden behind a pay wall, this can reduce
the likelihood of external data linking to it. Encouraging incoming links is hugely
important because they increase the chances of discovery, provide a measure of
popularity and trust of the dataset, and push it further up search rankings. At a
minimum, users or software agents need access to metadata: a description of the
data to know whether it is what they are seeking. The freemium model provides
some data for free and charges for certain enhancements, such as more up-to-date
data (or conversely, archival data) or more detailed, higher-quality, or a wider range
of data. Certain access mechanisms, such as download, might be priced higher in
exchange for the convenience. Charges can be subscription based, for a limited time
period, or “pay as you go.” The last case is interesting as there are of course many
different ways to chunk up data: Could one even pay per triple?
Of all the business model options, we would argue that the freemium model is
of the most interest to a GI publisher since it can provide the openness required to
improve discoverability while also protecting the value of the data.
8.2.10 i
iMpleMentation
of
l
inked
d
ata
b
uSineSS
M
odelS
Cobden et al. (2011) noted that every Linked Data set currently published uses a loss
leader or subsidy business model, and there are a number of technical hurdles to be
overcome before the more attractive freemium models incorporating some form of
paid access can be realized.
