Is the timing right for sponsored data?

	Oct 2014
	Strategic Perspectives

Oct 23^rd, 2014

Subash MANDANAPU,

Executive Summary:

In November 2007 Amazon introduced their Kindle line of products, with a seemingly magical Whispernet functionality where the content can be delivered without a subscription or monthly fee. Whispernet was one of the contributing factors of the original Kindle sales as it removes the friction of connecting to Wi-Fi networks. Amazon had to go great lengths, working with Sprint, AT&T, Qualcomm, and others to set-up the network to support this functionality. A process that proved challenging for Amazon’s competitors to replicate, due to heavy capex and technology complexity.

Today, mobile networks across the globe are transiting to 4G/LTE, smartphone adoption is increasing rapidly (70% in the U.S), and billing subsystems are evolving to support dynamic charging. Industry verticals like health care, finance, automotive manufacturing and others are embracing this mobility wave. We are entering into a connected intelligence era, wherein ubiquitous connectivity will enable contextually-aware objects (machines, medical pills, cars, appliances, food, etc.) in our everyday lives. This has an impact on how technology is instrumented, and potentially more importantly how incumbents embrace new business models to capitalize on a digital economy.

While telcos have done a tremendous job embracing modern technologies (4G/LTE transformation, cloud computing, ubiquitous coverage, smart devices, etc.) their business model remains little changed. Most Telcos are still responding tactically, not strategically (e.g.: Shared plans and data caps as opposed to sponsored data or dynamic charging). Telcos continue to be focused, almost exclusively, on mass market plans as a primary source of revenue. It’s time to apply strategic thinking to existing customer models as we transition in to the connected intelligence era, where consumers are at the heart of a myriad of contextual objects and access is critical to facilitate communication among those objects. Toll-free data or sponsored data presents a strategic opportunity for long term revenue growth, where consumers and/or other entities (enterprises, SMB’s) share costs to create new contextual experiences. By associating consumers and enterprises/businesses to the Telco value chain, Telcos can reduce the cost of delivery and increase value for the consumer.

The state of sponsored data in U.S

As one of the host networks for Amazon’s Whispernet, AT&T understands the need and potential for new business models like sponsored data. After mulling for years over net neutrality issues and waiting for the market to catchup, AT&T recently launched an offering at CES 2014 with 3 partners (United health, Aquto, and Kony). Their announcement received split reactions from the press. Negative campaigns were quick to point out how AT&T stands to make money from both sides (enterprise customers and their end customers). They also touched on some regulatory concerns. Nevertheless, here are the two key takeaways from AT&T’s launch of sponsored data:

a)     Taking a platform approach (where startups build the momentum around sponsored data) is less risky vs going to market directly. Startups are ambitious to bridge gaps between content providers and Telcos, and are interested in connecting to multiple Telcos to provide better coverage for their solutions.

b)     While sponsored data is not a tactical business model for most carriers and enterprises today, it will greatly impact a variety of industry verticals like connected cars, Healthcare, energy, etc, where Telco’s may not own the end customer relationship, but may power new connected devices and services for existing business customers. When Telcos begin to thinks beyond mass market monetization, sponsored data fits very well into the future of the telecommunications industry.

In part from their recent announcement, it is clear that AT&T is emerging as a leader of sponsored data efforts in the U.S.—by powering startups to go after the market as well as applying the model on other industry verticals. It is less clear what direct, bottom line contribution sponsored data has made to their overall earnings, but we can expect to see the results in next couple of years.

Verizon’s CFO Franc Shammo touted sponsored data at various occasions including some analyst conferences by hinting that content providers like ESPN were asking for sponsored data solutions. Verizon did work with a start-up called Aquito, who provided Verizon’s pre-paid subscribers with the Kickbyte app. Subscribers were rewarded with data when they consumed Kickbyte content (Ex: Surveys). Today, Aquito expanded their program to rich media advertising on AT&T network. Verizon is rumored to be evaluating a limited number of partners and content providers (like ESPN). Overall, Verizon’s network can support sponsored data models but they are looking for suitable content partners to launch a sustainable business.

T-Mobile USA is also very aggressive in terms of mass market offers, but they currently do not directly offer any sponsored data. Instead, they are zero-rating all the music streaming from different content providers (including Pandora, Slacker, Spotify, etc.). Essentially, T-Mobile is using a lighter version of sponsored data to gain market share in the U.S.

Sprint is losing ground and recently became # 4 in U.S. They have launched unlimited 4G plans to reduce churn, but have reduces investment in emerging areas like sponsored data. Given their focus on unlimited data plans, and continued losses, sponsored data may not be relevant for Sprint at this point. But, on the wholesale side, Sprint currently hosts more than 35 MVNOs. They are rumored to be evaluating startups which can enable sponsored data as a re-sellable product to their MVNO partners. Sprint-enabled MVNOs could then use such an offering to attract more customers by zero rating the content by region or demography.

Over the last 2 years, there has been increased activity in the startup ecosystem when it comes to sponsored data. There are two types of startups: (a) Enablers; and (b) Differentiators. Enablers build platforms to support content providers with sponsored data initiatives. Differentiators use sponsored data to differentiate their business model to compete.

The enablers like Syntonic wireless (a Seattle based startup that recently raised a $1 million seed round) partnered with AT&T, and announced toll-free data app called ‘Freeway’ on the  Android market place as well as ‘OnRamp, ’ an education program to enable sponsored data on tablets that were given to children at schools. Syntonic plans to work with other Telcos in the U.S to go after other verticals like IoT, enterprise services, and health care. Syntonic’ s competitor, ‘DataMi’  (Also an AT&T sponsored data partner), recently spun out of Princeton University and launched a limited trial on a subset of AT&T subscribers who use a social news application called ‘Trove’.

Differentiators like Aquito and Hipcricket are using sponsored data in advertising and marketing solutions.  Aquito launched a product targeting advertisers to sponsor high quality media content as well as credit additional data to users as an incentive to consume media. Hipcricket plans to use sponsored data to increase customer engagement with brands.

Market opportunity:

The market opportunity for the sponsored data model can be classified into two:

(a) Connect the un-connected; where Telcos target consumers who are not subscribed to data plans by introducing sponsored content; and

(b) Beyond mass market, where Telcos target premium offerings like mobile video for content providers like ESPN as well as other industry verticals like Auto, health, education, energy .etc.

In the table below, the International Telecommunications Union (ITU) indicates the opportunity presented to telcos to increase data adoption in emerging markets.

Key ICT indicators for the ITU/BDT regions (totals)

Mobile-cellular subscriptions(in Millions)	Active mobile-broadband subscriptions(in Millions)
2011 2012 2013 Africa          507          582          629 Arab States          379          399          410 Europe          743          766          780 The Americas          994       1,036       1,059	2011 2012 2013 Africa                74              117              172 Arab States                58                75                92 Europe              305              356              399 The Americas              401              494              577

Sponsored Data presents a considerable opportunity to engage new customers. As of 2013, mobile broadband penetration in Africa is only 13% and is only projected to grow by 19% at the close of this year. According to IDC, 18% of the African population has purchased a smartphone. This number is expected to double by 2017. With increased smart phone penetration, sponsored content, or subsidized content, can of great value to African consumers. It may be challenging to convince local content providers to sponsor data due to the economics, but internationally recognized social networks and sports content could be great way to start.

In most developed markets, mobile broadband has reached affordability levels, and with increased competition, Telco’s are forced to increase monthly quotas and reduce prices. At the same time, Mobile video usage is growing exponentially, and consumers may be attracted to sponsored data models where content providers pay for a consumer’s access to their preferred video content. In the U.S, many startups are focused on sponsored advertising content like video trailers, rather than an entire video episode or film, where the unit economics may not work for the sponsoring business partner.

As Telco networks are built to handle peak demand, there is significant availability during non-peak hours. Network demand based sponsored data can be another model where the network periodically checks for availability and engages with content providers dynamically to offer subsidized content paid by the content provider. Along with edge network caching, this model can be used to sponsor video content or rich media advertising.

Although not a single, definable industry vertical, the new wave of Internet of things (IoT) will impact many industries like energy, health care, automotive, and insurance, to name a few.  Ubiquitous connectivity will play an important role in IoT, and there will be challenges dealing with sub-dollar ARPU devices. Sponsored data models are very well suited to support various IoT solutions across multiple industries.

Sponsored data business case – Facebook

Facebook’s ad revenues jumped 67% in Q2 2014 year over year. This was primarily driven by a 151% increase in mobile ad sales. Facebook is now delivering 1 billion video views every day.  They also announced premium video Ads in international markets, including France. But going forward, it may be challenging for Facebook to monetize the video ads, as consumers may take a cautious approach towards video viewing, as it will be reflected on their monthly mobile statement. This is especially true for prepaid customers, who keep a more accurate measure of their data consumption.

By offering sponsored video ads, advertisers can expect increased engagement, and consumers can see direct benefit where it doesn’t impact his/her data limit. This, in-turn, can benefit Facebook to increase the click through rate (CTRs) of video Ads. Operator may initially cannibalize some revenue from consumer data plans, but over the long term, sponsored data will open up new markets. Sponsored data will also allow operators to differentiate their customer offering and diversify their revenue opportunities.

Facebook business case for Premium Ad per Million Subscribers

	Non Toll- Free	Toll-Free 1	Toll-Free 2
Facebook Click Thought Rate (CTR)	1%	5%	10%
video views	10000	50000	100000
Data cost to FB (Operator Rev) per ad ( @ .2cents fixed)	0	$1,000	$2,000
Facebook rev (@ $25 eCPM)	$250	$1,250	$2,500

Based on the above table, Facebook can double their revenue with 10% increase in click through rates with sponsored data option, more importantly, Operator can generate up to 90% of what Facebook can generate in ad revenue which otherwise is a lost opportunity.

Technology/Architecture choice

There multiple solutions that can enable sponsored data for a Telco like Operator. Each solution has its own advantages and disadvantages. Telco may need to consider ROI when picking the solution. Here are few solutions that a Telco can consider:

1.    In-Network solution: The traditional way of enabling sponsored-data is via an in-network solution where the Telco has full control and visibility over the packets that are intended to be toll-free.  The In-network solution reduces chances of fraud, as it will be implemented with-in the Telco network perimeter using existing network blocks. Unfortunately, this solution requires changes in various network elements including the policy layer, OSS/BSS layer, and the Home Subscriber Server (HSS). This solution may take time to implement and the associated costs are high. This solution can be implemented when the Telco can expect to generate commensurate revenues (or discounted cash-flow).

2.    Wrapper based solution: The wrapper based solution moves some of the policy, metering aspects from the network to the device to reduce the complexity on the network side. The native application can be wrapped to provide the same functionality and synchronize the metering and policy information to an independent server sitting outside the Telco infrastructure. By exposing an API (similar to billing credit), the server can credit the user in real-time or near real-time for the associated toll-free content. The big advantage of this solution is that the app developer or content provider can enable toll-free option without making any code changes.  While this solution may be comparably cheaper the In-Network, It has its own limitations including development and management of wrappers for each handset Operating System and there is possibility that wrapper can be compromised and become vulnerable to security attacks. This wrapper based solution takes whole application into toll-free context, hence it cannot support sponsoring sub parts of the app(ex: only Video advertisements)

3.    SDK/Libraries:  By providing sponsored data SDK’s /Libraries and an open source server to the developer, the Telco can not only reduce risk of managing toll-free content, it gives greater flexibility to the developer or content owner owner to enable whole or part the application sponsored. The Telco need to expose an API to content provider to credit the user’s account for the sponsored content.   Some of the limitations include the engagement during development time, managing SDKs across multiple languages (Ruby, Javascript, android , iOs, etc). By introducing an open source project, the Telco can mitigate some of the fore mentioned risks.  By combing device based SDKs with exiting OSS/BSS infrastructure (PCC rules) , Telco can create feasible solutions.

It is highly recommended to investigate the SDK based approach due cheaper development costs during the initial phases of market adoption. The open source approach may attract numerous developers, content partners and Telco community.

.

Execution Strategy

APIs: There are multiple ways to ignite the market with sponsored data models. Exposing sponsored data APIs is one of the easiest and least risky ways of approaching the market. This can build a startup ecosystem where the startups who are hungry for success can work with the content providers to make their content free to the end user. Small IoT manufacturers are prone to be attracted to this model as it might be less risky and with less friction to set up. Developer reach can be done via Operator’s existing partner program. The only risk with this model is that it is not guaranteed to produce results within a specific time frame. The success of this model greatly depends on startups that are willing and capable of executing. In U.S, AT&T is following this model by exposing APIs to startups and manages the relationship with a few partners, rather than hiring a direct salesforce and going after sponsored data contracts.

Big Launch:  Launching sponsored data by partnering with big content players like media houses, social networks, Music services, OTT messaging services and sports networks can catch attention of public as well as other content players and developers. Since sponsored data is a new proposition, it’s often challenging to convince big players to commit. But, if achieved, this is one of the best ways to create the momentum in the market. Popular events like mobile world congress, along with big brands, will drive attention from the industry. After a big launch, the API model for growth is more realistic. Due to the business complexity of sponsored data, showcase examples play a significant role in educating the ecosystem (startups, business customer, end-customers) about the value of sponsored data.

Buzz Maker: As the networks are planned for peak usage, there is often a lot of unused capacity. Telcos could generate good will by allocating network resources to good causes. By dynamically detecting network demand, subscriber profile information and demographic data, Telcos could offer a very niche group of services to a very niche group of customers, for free, in exchange for the goodwill generated. For example, telcos could use spare capacity to “sponsor” data for apps providing:

-        Education: Free online education at certain times for certain students/customers.

-        Healthcare: Free access to healthcare education at certain times for certain customers.

-        Government: Free access to government services at certain times for certain customers.  

All of these arrangements increase the telco’s brand reputation, reduce churn, and improve customer satisfaction.

Cross-carrier: One of the challenges with sponsored data is solution coverage for business customers not wanting to address a sub-set of the population. Content providers are often more attracted to sponsored data models when it is available across multiple carriers. In a given market, the #1 or #2 Telco could provide a sponsored data business partner customer with a master agreement, and then work with other telcos in that market (on a wholesale basis) to provide majority coverage. Based the history of Telco collaborations this is likely one of the biggest hurdles for operators to execute on a population wide sponsored data initiative. Startups may have a better chance of interconnecting multiple operators to create a one stop shop for content providers.

Regulatory concerns

When AT&T launched sponsored data in U.S, The digital rights groups like public knowledge, Free Press and other pro open internet groups criticized that sponsored data falls under net neutrality laws by encouraging operators to establish agreements with  established content providers(for faster lanes) and hinders the entry of new players. AT&T responded that they are not prioritizing the traffic when the content is sponsored and they are only shifting the billing from consumer to content provider. FCC chairman Tom wheeler told the press that he would investigate potential implications but so far no action was taken. It is highly recommended to check with local regulatory bodies to understand and assess potential regulatory hurdles.

Summary

We are entering into an era of connected intelligence, yet half of the world is not subscribed to data plans, and Telco business models have not evolved to accommodate new ways of connecting consumers. Traditional ARPU from mass market customers continues to decline, while the capex required modernizing and maintaining best in class networks continues to increase. It’s time to connect the un-connected, and diversify revenues, by targeting more innovative business models for data delivery.  Sponsored data presents a very flexible business opportunity and can be adapted to suit different market dynamics and various industry verticals. By working with partners to deliver sponsored services, telcos can increase the perceived value of connectivity. It’s time to seed new business models like sponsored data in the market for new growth opportunities.