Databases Reference
In-Depth Information
sponsored search can generate. Additionally, no online advertising model has been
as effective for so many different types of businesses as sponsored search. Sponsored
search can also be effectively integrated with contextual and social advertising,
enhancing its potential for longevity.
However, the long-term potential for the future success of sponsored search
depends on it continually adding value, both actual and perceived, to its stakeholders.
As long as it does this better than the alternatives, its future is generally secured.
Let us rst take a quick look at these terms, value and stakeholders. From there,
we will examine what changes the future might hold for sponsored search.
Potpourri
: Predicting the future is obviously not easy.
Despite this, scientists are trying hard to do it, from predicting short-term
events with social media streams to predicting longer-term threshold events. The
Internet has aided in this effort with its extreme capacity for data collection.
The problem with predicting the future using data from the past is described
by Hume's Problem of Induction [
2
]. Hume's Problem of Induction questions
whether one can predict that any event in the future will occur just because it
occurred in the past.
The induction problem entered pop culture via the topic,
The Black Swan
, with
the title being a classic example of induction from prior data [
3
]. For many years,
it was widely believed that a black swan could not exist, because no European
had ever seen one. Therefore, the prediction was that black swans cannot exist.
However, black swans do exist, being native to Australia.
Basically, in the end, we cannot prove that something will or will not occur just
because it occurred or did not occur in the past. However, this does not mean that
we cannot do anything with data.
Scientists have gotten around this touchy point by using data only to disprove
something. That is, empirically, we can show that there is evidence to disprove a
hypothesis, but we cannot prove a hypothesis is true. The best we can say is that a
hypothesis is supported based on the data.
We see this in the warnings in the marketing literature of financial invest-
ments - “Past performance is not a guarantee of future success.”
There is the related issue of that one cannot use data to prove a negative (i.e.,
that something does not exist). All we can say from data is that there is no evi-
dence of something.
Potpourri
: Predicting the future is obviously not easy.
Despite this, scientists are trying hard to do it, from predicting short-term
events with social media streams to predicting longer-term threshold events. The
Internet has aided in this effort with its extreme capacity for data collection.
The problem with predicting the future using data from the past is described
by Hume's Problem of Induction [
2
]. Hume's Problem of Induction questions
whether one can predict that any event in the future will occur just because it
occurred in the past.
The induction problem entered pop culture via the topic,
The Black Swan
, with
the title being a classic example of induction from prior data [
3
]. For many years,
it was widely believed that a black swan could not exist, because no European
had ever seen one. Therefore, the prediction was that black swans cannot exist.
However, black swans do exist, being native to Australia.
Basically, in the end, we cannot prove that something will or will not occur just
because it occurred or did not occur in the past. However, this does not mean that
we cannot do anything with data.
Scientists have gotten around this touchy point by using data only to disprove
something. That is, empirically, we can show that there is evidence to disprove a
hypothesis, but we cannot prove a hypothesis is true. The best we can say is that a
hypothesis is supported based on the data.
We see this in the warnings in the marketing literature of financial invest-
ments - “Past performance is not a guarantee of future success.”
There is the related issue of that one cannot use data to prove a negative (i.e.,
that something does not exist). All we can say from data is that there is no evi-
dence of something.
Value and Stakeholders
In Sergey Brin's quote in the epigraph about using technology to make a difference,
there are two concepts that we must define to evaluate whether or not a positive dif-
ference is being made. Namely, we must understand value and stakeholders.
Value
Value examines key questions, such as why goods and services are priced as they are,
how the price of goods and services evolves, and how to calculate the correct price
of goods and services.
