Download - Pricing Digital Marketing

uridavov - Arun Sundararajan

Découvre YouScribe en t'inscrivant gratuitement

Je m'inscris

Obtenez un accès à la bibliothèque pour le consulter en ligne
En savoir plus

40 pages

Français

Obtenez un accès à la bibliothèque pour le consulter en ligne
En savoir plus

A propos
Informations
Extrait

Description

Download - Pricing Digital Marketing

Informations

Publié par	uridavov
Nombre de lectures	89
Langue	Français

Extrait

Pricing Digital Marketing: Information, Risk Sharing and Performance Arun Sundararajan

Leonard N. Stern School of Business New York University 44 West 4th Street, KMC 8-93 New York, NY 10012-1122 asundara@stern.nyu.edu

Abstract:A unique value proposition of Internet-based digital marketing is the availability of precise measures of the actual performance of individual campaigns, which makes performance-based advertising pricing schedules feasible. These pricing schemes are studied in the presence of competition, performance uncertainty and asymmetric information about the quality of the clients content and the eﬀ Theectiveness of the publishers technology. papersndings challenge the dominant practice of CPM-based pricing for digital marketing by establishing that performance-based pricing is always prot-improving for publishers, even when publishers are constrained to oﬀer CPM-based pricing in parallel. It is shown that performance-based pricing cannot screen out clients with lower quality content/creative quality, and describe when to choose between low-end coverage and full coverage performance-based pricing. However, publishers can use performance-based pricing to credibly signal superior technological eﬀvalue of risk pooling for publishers results highlight the ectiveness. These of digital advertising, and the strategic role of pricing in signalling technological quality. Managerial guidelines are also provided for how to strategically respond to varying competitive intensity, client size and changes in outcome distributions.

Keywords:advertising, information asymmetry, adverse selection, screening, signaling,online electronic markets, digital goods, online marketing, banner advertising, CPM, performance-based pricing, Internet, World-Wide Web, WWW.

Internet-based digital marketing has evolved from Web banner ads and text-based email marketing to include a variety of sophisticated formats, technologies and delivery mechanisms. Many digital campaigns use Shockwave and Java-enabled animations, full-screen superstitials supporting video and sound, shoshkeles whichoat over a web page and then collapse into a small clickable icon, and rich media email incorporating streaming audio and video. Many publishers allow marketers to place deep-linked advertising text into the content of their web sites, and most search sites sell keyword-based advertising, wherein all consumer searches for a specic word or phrase trigger paid marketing messages related to that subject. Spending on digital marketing has risen steadily over 2002, and the medium isnding its place as an integrated part of corporate marketing initiatives. Current estimates indicate that by 2006, companies in the United States will spend a total of $16 billion on online advertising, and a further $19 billion on other electronic marketing initiatives, such as email marketing and targeted promotions (Gluck, 2001). Thus far, pricing for digital marketing has primarily been based on total quantity of impressions delivered  this impression-based method of pricing is commonly referred to as the CPM (cost per thousand impressions) model, and is a well-understood standard in the advertising industry. However, one of the unique value propositions of digital marketing is that a client can get precise measures of the actualperformance Examplesof each campaign that they run. of these performance measures include clickthrough rates on banners and icons, click rates on embedded web page hyperlinks in email cam-paigns, stream rates on rich media marketing messages, and response rates on keyword advertising or targeted promotions1. Delivery technologies like the DART system from DoubleClick support tracking detailed performance information of this kind, for both banner ads and for more sophisticated formats. Newer email marketing technologies, such as the RadicalMail suite from MindArrow Communications, can track the number of times a customer who received a rich media email streamed the embedded video. This ability to measure performance is an attractive feature to most media buyers  accord-ing to the Gartner Group, for todays cost sensitive marketers, ...accountability is more important now, and the Internet is the ultimate medium for this because of its ability to target [consumers] and measure [results]. (Gaﬀney, 2002).

1These measures may not be an exact indicator of total eﬀinstance, click-through rates do not measureectiveness  for how many people saw a banner without taking a measurable action. However, so long as they are correlated with actual eﬀectiveness, they have value as as basis for pricing.

1.1. Performance-based pricing issues

The performance measures associated with digital marketing provide a basis for pricing models in which a client pays an amount proportionate to the actual measured performance of a marketing campaign, rather than simply paying a price based on the number of impressions delivered. The feasibility of these performance-based pricing policies clearly represents a signicant opportunity for both the buyers and sellers of digital marketing. However, while some new media companies like Google have implemented sophisticated performance-based pricing systems for their advertising (Tillinghast, 2002), they are still the exception rather than the rule  in 2001, only about 20% of online advertising included some performance-based component. The limited use of performance-based pricing may be partly because designing protable performance-based pricing structures can be challenging. Recent industry reports supports this contention, suggesting that certain kinds of pay-for-performance digital marketing are substantially mispriced. For instance, Grahn (2001) indicates that when banner ads are priced based on performance, even when using the most sophisticated tracking techniques, the delivered banner advertising inventory is undervalued by about 35%, relative to comparable CPM rates. The complexity in developing and implementing such pricing models arises primarily from per-formance uncertainty, information asymmetry, risk sharing issues, and competitive pressure. These issues are elaborated on below. Performance uncertainty and risk sharing:There is performance uncertainty associated with any marketing campaign; this stems from both the incompleteness of measurable metrics like click-through rates in actually capturing total eﬀectiveness, as well as the inherent variability in consumer response to any specic marketing message. Any performance-based pricing structure shifts part of this performance risk away from the client and onto the publisher. Since this variability is partially determined by the composition and scope of the portfolio of campaigns that the publisher is currently exposing their audience to, publishers understand the risk better, and those with a diversied port-folio can pool performance risk. However, the extent to which these publishers should leverage their performance-based pricing models to insure client performance risk is unclear to most of them. Variable creative quality:A critical determinant of the performance of a digital marketing campaign is the quality and eﬀectiveness of the creative, or the actual text, images, sound and message delivered to the audience. The quality of creative is under the control of the client (or their advertisingrm), who designs it based on their knowledge of what actually comprises high-quality marketing content for their product. It is largely unknown to the publisher at the time of pricing  even

if the content is available for inspection, publishers typically lack suﬃcient in-house expertise to assess what constitutes good creative quality (their clients are from a wide variety of product industries). As a consequence, publishers may need to design their performance-based pricing schemes to screen between high and low quality creative content, and if possible, provide incentives for the latter group to self-select CPM-based (rather than performance-based) pricing. Eﬀectiveness of marketing technology:is also considerable variability in how eThere ﬀectively the publisher delivers a clients marketing messages. This stems from diﬀerences in technological and analytical capabilities  apart from the traditional direct marketing capabilities such as target selection (Bult and Wansbeek, 1995) and the use of purchase histories (Rossi et al., 1996), there is variability in the precision and quality of targeting technology, the ability to assess the consumers browser and email software viewing capabilities, and the ability to time and sequence the delivery of the messages for maximal response. In addition, diﬀerent publishers have access to target audience sets of diﬀerent quality. A fraction of these capabilities can be assessed from repeated usage of the same publisher  however, the digital component of a marketing campaign is often bid out by large advertising rms (who own the client relationship) on a campaign-by-compaign basis, to one of a number of competing Internet marketinglack of direct and repeated interaction between clients rms. This and marketingrms increases ex-ante quality uncertainty. publishers with superior Consequently, capabilities sometimes view performance-based pricing as a way of signaling the superiority of their delivery systems, but need to ensure both that their pricing schemes makenancial sense for them, and that these pricing schemes cannot be protably imitated by lower-quality competitors. Competitive substitutes:Online advertising and marketing is a competitive industry, and even when oﬀering performance-based pricing, the pricing power of publishers is limited by the competitive CPM-based price, which is always an outside alternative for their potential clients. This complicates the design of an optimal pricing schedule  a problem which is further exacerbated by the fact that the marginal cost of an additional digital impression is zero. In addition, many clients continue to be uncomfortable with purely performance-based pricing schemes, because the advertising industry norm is the CPM model, and the client often has axed budget, or rigid media spending constraints. Consequently, in order to maintain viable business models, publishers are often forced to oﬀer the option of a CPM-based price, which is typically the competitive rate. This means that they need to design their performance-based pricing scheme with the understanding that clients could observe, infer from, and yet bypass their performance-based pricing, while still purchasing from them by paying the competitive CPM-based price.