The Future of Telecom Digital Advertising What Telecom Digital Advertisers Need to Know about First-Party Data and Targeted Advertising November 2015
TABLE OF CONTENTS 3 7 10 Why First-Party Data matters in Telecom Digital Advertising Digital Advertising vs Traditional Advertising vs Direct Database Marketing Concerns about Privacy & CPNI EXECUTIVE SUMMARY First-party data use in telecom digital advertising is currently limited to media targeting, through retargeting and lookalike modeling. While media targeting with first-party data is an improvement upon traditional advertising, first-party data is not often leveraged for creative decisioning; its use is often limited to decisioning off of broad service eligibility requirements instead. Digital advertising increasingly resembles digital direct database marketing channels that telecom providers have long mastered. Telecom service providers are leaders in hyper-targeted micro-messaging in direct mail, email and outbound call center channels. New platforms enable telecom providers to leverage database marketing expertise for micro-messaging to digital audiences without the limited reach and high costs of targeted media. CPNI and consumer concerns are real, but do not present obstacles when firstparty data is used correctly with proper opt-outs. 2
WHY FIRST-PARTY DATA MATTERS IN TELECOM DIGITAL ADVERTISING Leading digital marketers across all industries have recently shifted their focus from third-party to first-party data for digital targeting. As reasons for the switch, most digital marketers point to the expense of third-party data and the fact that its common availability diminishes any competitive advantage. First party data is free, and provides a unique competitive advantage as it is owned by the advertiser. What is first-party data? First-party data is data that is gathered on your potential customers interaction with your brand, and actual customers interaction with your products. Purchase and usage history Purchase date, Size, Product ID Service usage volume Cancellation date, Reason Shopping cart activity Date, Size, Product Line, Abandonment Site browsing activity Date, URL, In-page-events Email marketing activity Opens, Clicks The current primary use cases for first-party data are focused on media targeting: lookalike modeling and retargeting. These media targeting tactics require little more than a unique identifier, and they enable a granularity of media targeting that has been impossible with traditional advertising. 3
digital advertisers are able to spend less on targeted media, thus reducing CPM media costs while expanding reach and lifting response rates. Targeting through media in digital advertising is an optimization of the traditional, broadcast advertising model in which the same ad is shown to a mass audience. A much more powerful use case that is rapidly emerging for first-party data is creative decisioning that targets through micro-messaging to digital audiences, not through targeted media. Micro-messaging builds upon database marketing science perfected by telecom providers for direct marketing channels like direct mail and email, where the media costs are less than advertising media. By leveraging first-party audiences for digital micro-messaging in the same way, digital advertisers are able to spend less on targeted media, thus reducing CPM media costs while expanding reach and lifting response rates. Emerging platforms like those from Telogical, PaperG, Spongecell and Anagram are bringing the micro-messaging techniques common in direct database marketing to digital advertising. These platforms enable creative testing of dozens of messages for multivariate audience discovery and then targeting of the right message to the right audience. Until this point, use of first-party data for creative decisioning by telecom digital advertisers has been generally limited to eligibility requirements, as service providers generally have different network capabilities, and different offers and promotions, in different markets. 4
How do you use first-party data? (1) Lookalike Modeling: Many platforms allow marketers to onboard customer lists, and target ads to audiences with similar profiles. These platforms include Facebook Custom Audiences, Google Display Network s Similar Audiences and Twitter Tailored Audiences. (2) Retargeting: While basic display retargeting has been around for a while, advanced retargeting tactics include search retargeting using RLSA from Google and email retargeting using a vendor such as LiveIntent or Triggermail. (3) Creative Decisioning on Eligibility Requirements: Current use of first-party data by telecom digital marketers is generally limited to eligibility requirements, such as market-based offers and network build-out. (4) Creative Decisioning for Micro-Messaging: The first two tactics, like third-party cookies, target through media and are costly. While third-party cookies and lookalike modeling come at a direct premium on top of CPM, retargeting pools typically have high CPMs because they are cookie-rich users. Emerging vendors, such as Telogical, PaperG, Spongecell and Anagram, target through creative micro-messaging, which is less expensive and doesn t require limiting media reach to be effective. 5
In consumer telecommunications marketing, both wireless and wireline digital marketers face even greater lift potential, as well as greater challenges, than other sectors from the use of first-party data. Service providers possess granular product purchase and usage data that they have used for years to target traditional direct marketing direct mail, outbound call center and email. The following sections will explore the potential to realize this opportunity in telecom digital marketing, as well as the unique challenges to use of first-party data for fine-grained targeting in telecom. 6
DIGITAL vs TRADITIONAL ADVERTISING vs DIRECT MARKETING Digital advertising is often compared favorably with traditional broadcast advertising for its granular targeting and attribution capabilities. However, as digital advertising becomes increasingly addressable at scale, the more apt comparison is with traditional direct marketing channels such as direct mail, outbound call center and email. Telecom service providers have been leaders in advancing these latter channels. How Direct Database Marketing Works Statistical Modeling: Database marketing analysts use stats packages (SAS, SPSS, etc) to build one of two types of model. Cluster models combine first and third-party data to identify clusters of similar customers, and then create campaigns specific to each cluster. Predictive models assign a score to each prospect of their likelihood to respond to a campaign. Activate Model for List Generation: Identify contacts in a list management tool to target using the statistical model for the campaign. Run Campaign: Send direct mails or initiate outbound calls or emails to all contacts in the campaign list. Refine Model: While in-house database marketers use response data to refine their statistical targeting models, more advanced platforms like those from Globys and Pontis enable real-time model refinements using machine learning. Globys and Pontis in particular enable personalized contact center offers based on statistical models. 7
How does traditional direct marketing compare to digital marketing? Media targeting vs creative content targeting. Digital marketing has focused on media targeting, while email, direct mail and outbound call center have sophisticated micro-messaging platforms. Direct mail copy, for example, is now sent via XML formats that include content personalization rules and are targeted towards clusters identified by in-house database marketers. Telecom product content personalization platforms like Telogical have been used for many years on contact center desktops to prompt agents with personalized product claims and messaging. Ownership of data and models. Digital marketing does not own data or statistical models of data, while direct marketing has traditionally own and modelled data in-house. The rise of walled garden platforms in digital advertising are heightening awareness of this issue, as telecom marketers have historically invested heavily in their management and modeling of campaign data across direct marketing channels. How can service providers combine their traditional expertise in direct marketing with the broader advertising formats available in digital advertising? Creative Testing for Audience Discovery. New platforms like those from Telogical, PaperG, Spongecell and Anagram enable telecom digital advertisers to test several dozen, even hundreds, of creative variations concurrently. These platforms then leverage first-party data and other data sources to discover which audiences respond best to which creative messaging. 8
Telogical s Creative Performance Dashboard, seen below, enables digital advertisers to monitor the overall lift from each of dozens of creative tests, while also monitoring the highest and lowest performing audiences for each creative variation. In addition, Telogical s platform learns a predictive model that dynamically selects the creative that maximizes each user s propensity to purchase based on their user profile. 9
CONCERNS ABOUT PRIVACY AND CPNI Customer Proprietary Network Information (CPNI) refers to the type, quantity, destination, technical configuration and amount of use of the telecommunication services subscribers purchase from a service provider, as well as related billing information. The use of CPNI for marketing purposes is regulated by the FCC under the authority of Section 222(h)(1) of the Telecommunications Act of 1996. Service providers may share CPNI among their affiliates to market communications-related products and services. CPNI may not be shared with unrelated third-parties for their own use. Service providers must enable subscribers to opt-out of the use of their CPNI for marketing of communications-related products, and all major service providers have long included such opt-out capabilities. This is how telecom database marketers have leveraged user profiles to build statistical targeting models in traditional direct marketing channels. 10
CONCLUSION Telecom service providers have historically been at the cutting edge of direct database marketing practices, leveraging their deep first-party data to finely target their acquisition, upsell and retention messaging via direct mail, outbound call center and email. While digital advertisers are paying a premium for media targeting, such targeting still presumes the mass advertising model of a single creative to a mass audience. Telecom digital advertisers are increasingly looking to build on their first-party data and database marketing expertise to target at the creative content level, not through costly media targets. This is the future of targeted digital advertising, and telecom advertisers are likely to lead the way. 11