Page 63 - The European Business Review

models give different channel valuations and average channel reward values showed a
than last-click and whether these channel VLJQLÀFDQW GLIIHUHQFH EHWZHHQ WKH FXUUHQW PRGHO
YDOXDWLRQV GLIIHU VLJQLÀFDQWO\ EHWZHHQ WKH PXOWL (last-click) and the other models (e.g. rule-based
channel models used. We develop predictions and statistics-based) explored. As we value each
that examine at what stage in a consumer’s medium for its contribution to the end purchase,
journey different online channels feature most our study understates the value of some key
SURPLQHQWO\ IRU DQ RQOLQH EXVLQHVV WKH ÀQDQFLDO emergent media in the chain, particularly social
importance of these channels under last-click; media. It appears that the attribution models
and the effects of moving to the rule-based currently do not fully value social media, which
multi-channel attribution models (i.e. time- often do not directly lead to purchase but can
decay, uniformly distributed and position-based) have a strong behavioural impact, for example,
and statistics-based multi-attribute models. As by shaping the consideration set. While the value
our study of multi-channel models show, display of social media does improve as the sophisti-
is the biggest loser as we move from last-click cation of the attribution models increases, the
to other multi-channel attribution models. In consumer behaviour implications need to be
fact, this is our key result; when multi-channel fully accounted for in any study of advertising
models are used online marketing tools such as and attribution.
organic and search will receive higher credit. ,Q FRQFOXVLRQ RXU ÀQGLQJV UHYHDO ZLGH
In our empirical tests, we use the last-click differences between the online channels
model as a default attribution strategy, which investigated. There is a striking drop in the value
means that we could focus on assigning value to of display ads when we move away from the
display ads and compare the effects of moving current last-click model to the other multi-channel
to multi-channel attribution models. We focus on attribution models. Online marketing tools such
the extent to which display ads generate higher as organic and search instead receive higher
average order values than other online marketing credit under these multi-channel models. The
tools; and whether display ads generate more ÀQGLQJV SURYLGH LQVLJKWV LQWR WKH FRPSOH[LWLHV RI
revenue under the last-click model than it would attribution modelling, and how one can choose
when using the other attribution models such an appropriate model based on its underlying
as time decay, linear, position-based or Shapley assumptions and stability characteristics. Our
Value-based model. Our questions are moti- results also shed light on the convergent validity
vated by the observation that display advertising of the multi-channel models, as well as the
is likely to act as a converter in a purchase funnel; predictive ability of the statistical model.
and the last-click model attributes 100 percent
credit to a convertor. About the Author
Our findings show that the last-click Tahir M. Nisar is an Associate
generates the most revenue for the converter Professor in Southampton Busi-
– in this case display ad – and delivers the ness School at the University of
highest average reward. However, when Southampton, United Kingdom.
comparing the last-click model against each Dr. Nisar has published numerous
one of the other models, the results show a articles in distinguished academic journals,
VLJQLÀFDQW GLIIHUHQFH LQ WKH DYHUDJH FKDQQHO including Journal of Retailing and Journal of Adver-
reward value, with the multi-channel attribu- tising Research. His current research is on Digital
tion models assigning increased value to search, and Social Media Analytics and Big Data.
organic and other ad formats; both the revenue
References
1. Manchanda, P, Dube, J.P., & Goh, K.Y. (2006), The effect of
There is a striking drop in the value of display ads banner advertising on Internet purchasing, Journal of Marketing
Research, 43 (1): 98-108.
when we move away from the current last-click 2. Nisar, T.M. & Yeung, M. (2017) Attribution modeling in
digital advertising: An empirical investigation of the impact of
model to the other multi-channel attribution models. digital sales channels, Journal of Advertising Research.

www.europeanbusinessreview.com 63