Summary
DMA articles summary
- Course
- Institution
Digital Marketing Analytics summary of all articles (19/20).
[Show more]
1 review
By: Casparweisner • 2 year ago
Seller
Follow
nikkinuman
Reviews received
Content preview
Digital Marketing Analysis ArticlesWeek 46 – Lecture 2:
Article: PROSAD: A Bidding Decision Support System for Profit Optimizing Search Engine Advertising by
Skiera, B., Nabout, N.A. (2013)
Abstract PROSAD means PRofit Optimizing Search engine ADvertising, this system maximizes an advertiser’s
profit per keyword without the need for human intervention. Field experiment demonstrates that PROSAD increases
ROI by 21 percentage and improve yearly profit potential for SoQuero (online marketing agency) and its clients by
€2.7 million.
Introduction SEA (search engine advertising). How mechanisms supporting SEA works: consumer types a
keyword into a search engine and receives two types of results 1. The lower left-hand portion of the page shows
unsponsored (free of charge) search results; whose ranking reflects the relevance assigned to these different results by
a search algorithm. 2. The right-hand side, sponsored search results appear (advertisers pay for each click on their
ads).
Rankings and prices paid per click depend on keyword auctions; in auctions advertisers submit bids for a specific
keyword by stating their maximum willingness to pay for each click. The search engine provider weights the
submitted bids according to the ad’s quality (measured using a proprietary quality score QS) and displays the
sponsored search results in decreasing order of weighted bids. If a consumer clicks on the ad, the advertiser pays an
amount equal to the next-highest weighted bid, divided by its own QS.
SEA campaign managers use rules-based approached to determine their bids for keywords to make automated bidding
decisions.
1. IF keyword profit after acquisition costs is greater than €10, THEN increase bid by 30%
2. IF rank is worse than rank 5, THEN increase bid by 20%
3. IF keyword profit after acquisition costs is smaller than €0 AND the number of clicks is more than 100 AND rank
is better than rank 3, THEN decrease bid by 20%
Disadvantages of using these rules:
1. The number of rules can grow quickly and become difficult.
2. Different rules might offer contradictory bidding suggestions.
3. The choice of parameters in the rules is rather arbitrary
A decision support system, as PROSAD, can set bids to maximize profit automatically.
Description of Bidding Decision Model 3.1. Basic Idea: The bidding decision model links bids to profit and
therefore reveals which bid maximizes profit. The bid and ad quality determines the ad rank which influences the
number of clicks the ad receives. Conversion rate = number of clicks which is affected by the number of users, the
clickthrough rate (= number of clicks/searches) and conversion rate. Finally; the difference between profit
contribution per conversion – the acquisition costs per conversion x the number of conversions = the keyword’s profit
after acquisition costs.
3.2. Determination of Optimized Bid: Advertiser aims to maximize profit after acquisition costs, by determining the
optimized bid per keyword with the constraint that the bid must and positive and lower than the highest possible bid,
which achieves rank 1.
The advertiser’s profit after acquisition costs is the sum of transactional profit and branding profit.
Transactional profit = the difference between profit contribution per conversion – the acquisition costs per
conversion x by the number of conversions.
Branding profit = captures the perception generated by the display of an ad in the sponsored search results.
The number of conversions that an advertiser acquired by bidding on a keyword can be calculated by: multiplying the
number of clicks and the conversion rate. The number of clicks is simply the number of consumers searching for
keyword multiplied by the clickthrough rate.
,The ratio of the bid to the conversion rate yields the acquisition costs per conversion.
To calculate the number of searchers: it is assumed that the click response function is continuous and depends on the
rank of the ad. The clickthrough rate increases at better ranks and reaches its highest value at rank 1.
The rank function is the inverse of the price function. Rank 1 is the highest possible rank; boundary solutions also
exist (if the optimized bid is lower than rank , the resulting rank is worse than rank 1).
3.3 Optimized Cost per Profit: As the optimized bid is lower than the bid at rank 1, the optimized costs per profit
depend only on percentage increases in prices per click and clickthrough rate within ranks. The optimized costs per
profit decrease with ascending percentages increases in prices per click. The optimized costs per profit decrease with
ascending percentage increases in price per click. The intuition is as follows: if the percentage increase in prices is
high, the advertiser can bid substantially lower for worse ranks, which strongly reduces the acquisition costs per
conversion and results in lower optimized costs per profit.
An increase in clickthrough rates within ranks has opposite effects: if the percentage increase in clickthrough rates is
high, the number of clicks diminishes substantially within ranks, which makes better ranks more attractive. Then
advertisers submits higher bids, leading to higher optimized costs per profit.
Calibration of Response Functions 4.1. Estimation of Price and Click Response Functions: As automatic
calibration of the price response function and click response function is challenging as it requires fast estimation
procedures. Instead, regressions that estimate a response function for each keyword separately are used to deal with
unreasonable parameter values (competitors). Search engine providers inform advertisers about prices per click, ranks,
number of consumers searching for a keyword, number of clicks, number of conversions, and the QS of the campaigns
but do not provide competitors’ data.
4.2. Outline of Heuristic: Missing data, too few observations, measurement errors can produce such unreasonable
values, and thus poor bids. If the calibration of the response function yields reasonable values, the heuristic that costs
per profit should be 50%, because the percentage increases in clickthrough rates and prices per click are frequently
equal is used. This heuristic is simple to use and guarantees that the acquisition costs per conversion do not exceed the
profit contribution per conversion.
Application of PROSAD The results of the field experiment include weekly information about numbers of
searches, clicks, conversions, average cost per click, average rank, acquisition costs and the profit after acquisition
costs for each keyword. Experiment shows that without PROSAD the advertiser bids too high which diminished the
profit after costs. PROSAD, because of simplicity, does not take competitive reactions into account nor model the link
between clickthrough rate and QS.
Summary PROSAD can automatically determine optimized bids that maximize the advertiser’s profit. Field
experiment shows that PROSAD increases the ROI by 21% and better captures the trade-off between number of
conversions and the money spent to acquire those conversions.
,Questions:
1. What is being studied? The effectiveness of using a program, Profit Optimizing Search engine Advertising
(PROSAD) to optimize profit per keyword and its advantages. PROSAD determines per keyword the bid for the
search engine advertising which optimizes the profit after acquisition.
2. Why is this important to study? (managerially and scientifically) To be more effective in creating profit while
minimizing costs.
3. How and where is this research conducted? Field experiment is used to test the automated bidding versus bidding
for keywords the traditional way. Field experiment with a small client, conducting a nationwide campaign for toys and
kindergarten materials to rely on PROSAD for 20 keywords over a period of 10 weeks. In this experiment the client
first managed the campaign itself without PROSAD and then SoQuero managed 20 keywords using PROSAD.
Results of this experiment included weekly information about the number of searches, clicks and conversions.
4. What are the main findings?
1. PROSAD increases ROI with 21% and improves yearly profit for its client by €2.7 million.
2. PROSAD results in significantly lower bids and ranking, lower click-through rates and a decrease in conversions,
but due to the fast decrease in acquisition costs does result in a positive ROI
5. What are the strengths and weaknesses? (e.g. room to improve, things to add, etc.) I would rather see more
programs such as PROSAD as this seems more like an advertisement for this program. Moreover the ROI and yearly
profit are not really explained how it is improved; results could be more in depth.
, Week 47 – Lecture 3:
Article: Buying, Searching, or Browsing: Differentiating Between Online Shoppers Using In-Store Navigational
Clickstream by Moe, W.M., (2003).
Abstract In virtual shopping environment, the underlying objectives of shoppers are hardly to see. By using page-
to-page clickstream data from a given online store, visits are categorized as buying, browsing, searching or
knowledge-building. Each type of visit varies in terms of purchasing likelihood. Categorizing visits allow e-commerce
marketers to design more effective, customized promotional message.
Introduction Availability of internet clickstream data allowed marketers to examine consumer search behaviour in
a large-scale field setting. Visit-to-visit basis = examining store visit and purchase decisions over time. Page-to-page
basis (examining navigation within a single store-session). Most interstore studies focus on search behaviour. Studies
of consumers’ intrastore behaviour over time more closely examine and predict purchasing behaviour. Customers
‘flow’ describes the general customer experience online. Mandel and Johnson (2002) showed that preferences, and
purchasing decisions, are often constructed online while navigating through the store. Therefore, the content of the
pages viewed can be very important both in determining the type of shopper involved and in predicting purchases.
Typology of Shopping Strategies Search behaviour can be either goal-directed versus exploratory search. Goal
directed = having a specific or planned purchase in mind. Search patterns are focused and directed toward the goal of
making a purchasing decision. Exploratory search = consumer is less deliberate and focused and perhaps not even
considering a purchase. Search tends to be undirected and stimulus-driven; type of search is called browsing or on-
going search. Both types of search behaviour may result in a purchase. Exploratory stimulus, which is stimulus-driven,
may result in a purchase when exposed to the right stimulus. Moreover, activities in exploratory search visit may
influence future purchasing decisions.
Search behaviour
Directed Exploratory
Purchasing Horizon
Immediate Directed buying Hedonic browsing
Future Search/deliberation Knowledge building
Directed buying The shopper intends to make a purchase and is not lacking any substantial information before
making that decision. Store visits are driven by directed-buying strategy and likely to result in an immediate purchase.
In-store behaviour is focused and targeted. Little information search outside the simple availability and pricing
information is typically required. If the search process is nearing an end and little information remains to be gathered;
later visits to a story may result in an immediate purchase. In online shopping environment more product-level pages
rather than category-level pages are viewed as category pages provide a broader level of information and product
pages provide more targeted and detailed information. As directed buyers are near purchasing, directed-buying
strategies would involve deep deliberations as indicated by high levels of repeat product viewings.
Search / deliberation also goal-directed with planned purchase in mind; difference is timing of that purchase.
Directed-buying results in immediately buying whereas search / deliberation visits are motivated by a future purchase.
The objective of visits is to acquire relevant information to make an optimal choice. The strategy is designed to build
the consideration set and evaluate these items. Characteristics of this strategy would include a focused search within a
product category.
Hedonic browsing dominated by exploratory search behaviour. In store behaviour tends to be more stimulus-
driven and occasionally results in impulse buying, depending on the nature of the stimuli and countered. In-store
behaviour is less focused and therefore more of the session is spent viewing the broader category level pages than the
product level pages. Hedonic-browsing sessions should exhibit a lot more variety, in terms of products and categories
viewed. We expect very little repeat product viewings and limited drill-down on a particular product.
Knowledge building another motivation of exploratory search. Shopper’s objective is to increase product and/or
marketplace expertise. Information gathered may influence future purchasing decisions. Knowledge-building shoppers
focus on informational pages in an online shopping session (advice columns, discussion areas etc.).. Shoppers tend to
spend more time processing informational content on the site than other shoppers. Thus, we expect these shoppers to
have longer page-view durations.
The benefits of buying summaries with Stuvia:
Guaranteed quality through customer reviews
Stuvia customers have reviewed more than 700,000 summaries. This how you know that you are buying the best documents.
Quick and easy check-out
You can quickly pay through credit card or Stuvia-credit for the summaries. There is no membership needed.
Focus on what matters
Your fellow students write the study notes themselves, which is why the documents are always reliable and up-to-date. This ensures you quickly get to the core!
Frequently asked questions
What do I get when I buy this document?
You get a PDF, available immediately after your purchase. The purchased document is accessible anytime, anywhere and indefinitely through your profile.
Satisfaction guarantee: how does it work?
Our satisfaction guarantee ensures that you always find a study document that suits you well. You fill out a form, and our customer service team takes care of the rest.
Who am I buying these notes from?
Stuvia is a marketplace, so you are not buying this document from us, but from seller nikkinuman. Stuvia facilitates payment to the seller.
Will I be stuck with a subscription?
No, you only buy these notes for $4.31. You're not tied to anything after your purchase.
Can Stuvia be trusted?
4.6 stars on Google & Trustpilot (+1000 reviews)
67474 documents were sold in the last 30 days
Founded in 2010, the go-to place to buy study notes for 14 years now