Back Office, Front Burner
Imagine having a dream. One of your subscribers is sitting at an airport waiting for a delayed flight. She pulls out a PDA and logs on to her cable email account using the local wireless network, logging on using her email user name. She reads an email from your marketing department listing new content titles available to subscribers. She decides she has time to watch one of the premium titles. She clicks on the "buy now" button, and the content starts streaming to her PDA over the partner wireless network. A billing item is added to her account, and the revenue is reconciled between the wireless partner, the content partner and your company – all automatically.
When you wake up from this dream, are you sweating? Reality check Back office systems are usually thought of as overhead, a necessary evil. They do not deliver services. They are not generally "subscriber facing." They do not represent slick new technologies.
In truth, what the back office does is turn those slick technologies and services into a business. They bill for them. But they also enable the delivery of services by provisioning them, providing customer profiles and preferences to optimize the experience, by entitling them based on subscriptions, by ensuring reliable service delivery (and automated error detection and recovery when necessary), and by making it as attractive and easy as possible for subscribers to buy.
As the traditional cable silos continue to converge and new service areas are added to the mix, the back office will play a central role in taking advantage of operators’ shared networks to provide the sort of cross-platform experience subscribers expect. Unfortunately, today’s cable back office consists of a few large, monolithic components, such as billing and provisioning systems, surrounded by smaller systems that are designed to tackle specific problems that the monolithic systems cannot immediately address. There may also be one or more layers of middleware tying it all together.
By most accounts, this complex infrastructure falls short in its ability to enable the converged services that operators claim they want to provide. Meanwhile, pressure from competitive broadband providers intensifies, and worse, over-the-top content providers are taking revenue directly from subscribers and leaving operators out of the loop.
What follows are a few examples that that illustrate how to begin untying the Gordian knot that most operational support systems (OSSs) have become, how to chart a course toward the delivery cross-platform services and how to take practical steps toward implementing a more open and flexible back office. Content integration In the upcoming challenge over network neutrality, one tool available to the cable industry will be its willingness to work with content providers cooperatively, in return for reasonable compensation for access to the cable networks. Making the most sought-after content products easily available to subscribers will encourage more of them to stay within the portal and use the services it provides rather than roam to over-the-top content providers.
Of course, cable partnerships with content providers are not new. However, even if subscribers can access partner content from within the operator’s portal and use cable login credentials to access them, they most often still have to set up a separate billing account. A given partner may bill based on time used or per download or in other ways that can be difficult to integrate into existing cable billing infrastructures.
Creating a separate billing account is not only inconvenient, it also ties the subscriber relationship to the content partner, not the operator. That obviously reduces desirable "stickiness."
A tighter integration with the content partner, however, can provide valuable information about how subscribers are using the service, which could be included with other subscriber behavior to improve recommendations, advertising targeting and to generally tailor the overall cable experience to the subscriber’s preferences. Making content easier to access also will increase subscriber uptake for services directly affecting revenue.
This level of integration requires a product catalog that can be accessed from within every subscriber access point (online portal, video on-screen guide, wireless portals), an identity management and entitlement system that can be federated to third parties seamlessly, and a transaction management system that will process payment for a la carte content events, ensure that the payment gets added to a cable bill or credit card, and that the proceeds are allocated correctly among the operator and its content providers.
Many of these functions already exist in some form within cable operations today, but they tend to be buried in larger monolithic systems and not available for integration with other services. There are a number of vendors who offer technologies that address these problems quite well. However, it can be hard to sell a product catalog to a business that already has one, even if the one in place is not fully meeting its needs. Impulse buys Each time that subscribers interact with a cable provider is an opportunity to inform them of additional products and services that may be of interest. Yet operators seem surprisingly unable to take advantage of these opportunities.
Buy flows are typically implemented as a script in a customer relationship management system or into the Web application behind an e-commerce portal. Given the significance of those rare moments spent interacting with subscribers wanting to spend money, it is surprising that the approaches taken to maximize the value of those moments are so ad hoc.
What about taking a more strategic approach? Here are some suggestions. First, ensure that whatever investments were made to automate and maximize the buy flow were available to all points of subscriber access. Second, leverage any data about individual subscribers and aggregate subscriber behavior data to best anticipate offers that will most appeal to them when they ask. Third, ensure that whatever technology is used integrates with existing product catalogs so that buy-flow development and product packaging decisions can be made and managed in tandem.
Interestingly, the most challenging aspect of this problem space may well be access to past and present subscriber information. Information on where a subscriber resides and what packages are available (and, in some cases, even what they already have) is often locked in billing systems or other application-dependent databases.
This implies a prerequisite step to this and many other forward-looking back office infrastructure upgrades, namely, a unified data strategy that makes current and historical information about subscribers available through scalable and open interfaces. Effective advertising The concept of targeting advertising is not new. The ability of cable operators to target ads more tightly than broadcast media has made the industry a pioneer of the concept. Yet the processes and systems built to make targeted advertising most effective have been designed toward to the specific needs of video, which creates a challenge.
Project Canoe, which outlines exciting plans for improved ad targeting for video delivery, looks to be an enabler that will support the delivery of even more tightly targeted messaging, but mostly overlooks the most important aspect of the problem: using customer data to actually select the targeted messages. It does allow advertisers to select specific demographics, such as geography or programming. However, compare this with the Google Web ad model that automatically chooses ad placement based on actual subscriber behavior.
Targeted ad delivery is an important technology. But the back office opportunity is to maximize that technology, along with Web-based ad placement and to not only be more effective at the broader-based advertising, but also to go after the so-called "long tail," customers willing to pay premium dollars for exactly what they want.
This, of course, requires data. And it requires coordination among the sources of data from all of the product lines. Much of the data that is useful for ad targeting is the same or similar to that used for maximizing buy flows for impulse purchases. But real effectiveness requires more data.
The technology that sifts through subscriber behavior will be comparing individual behavior to aggregate trends, and this can require data that go back months or even years. It is a rare operator that maintains this much historical data at all, much less in a form that is easily accessible in real time by the applications that would be making ad targeting recommendations. Yet, taken along with the data requirements for the applications discussed here and others yet to come, the argument for building a scalable, open and secure subscriber behavior data warehouse becomes more compelling. Beyond point solutions Taken together, this brief look at future requirements for a back office architecture that will support increased revenue, improved delivery costs and decreased subscriber churn suggests that point solutions to the specific problems of each application space will be more expensive and less effective than a more cohesive and open back office architecture.
Designing and implementing such an architecture, though, can seem daunting. Discarding existing technology investments is impractical. Budgets are easier to justify for technology that promises near-term cost or revenue improvements than for major revamps with return on investment (ROI) quarters or years off. Finally, it is sometimes hard to find the most promising technologies available, because vendors often hide their technology in tactical products designed to solve some near-term problem even if the long-term benefits their technologies provide are even more substantial.
With operators buying and vendors consequently selling point products, and absent any real industry standards on back office components, how might one work with vendors to build out an open and flexible back office architecture? Simply stated, operators should lead the way.
Without the help of those operators’ IT thought leaders, vendors trying to position themselves in useful ways will continue to struggle. By parsing out the known requirements that will drive the back office over the next decade, sorting them into well-defined problem areas and beginning to describe what the interactions between those components will need to be, operators can lead the most innovative technology providers to package their products in ways that satisfy those needs. Reference model To accomplish the goal of working not just on the problem of the day but also for the long term, operators should consider designing their own back office reference model.
Here is a four-part suggestion for designing a model and giving vendors specific criteria they can use to best position their products and technology:
First, divide back-office requirements into well-defined components. Second, identify how other components and applications can access them. Third, define the data sources each will have available. Fourth, define the interfaces between components in a technology-independent way.
Of course, defining an architecture and implementing it are different things, especially when budgets are tight and benefits not immediately obvious.
One way to proceed is to take the well-defined, near-term problems, for which budget can be justified, and spend a little more time interpreting these as well-defined components of an emerging reference model. Insist that the vendors who supply those components open up their capabilities for reuse in the future. This can cost more and take longer in the short term than implementing a point solution. But it will typically only take one more project with overlapping requirements to come out ahead. Getting started The best thing about this incremental approach is that it can be applied to projects already being planned. Whether the highest priority for this year is reducing support call times, rolling out third-party Web services or increasing ad revenue, there probably are requirements for current projects that apply to future initiatives. It is worth the incremental time spent assessing projects using a wider lens and defining problems in terms of open and reusable components in order to start designing and implementing a reference architecture.
How to get started? The general process should be familiar to most operators. First, work with vendors through requests for information (RFIs) and other ways to determine the technology best suited to your needs as well as to vet out your architecture. Once it becomes clear what what kind of technologies are available, start to do a more in-depth evaluation through requests for proposals (RFPs) or vendor bakeoffs. Next, move to a detailed design phase; a development phase using some combination of internal staff, vendors and integrators; a system integration phase; then test and deployment phases.
By taking a small step back from existing projects and considering ways that the ongoing work can benefit longer- term strategic goals, an operator’s IT and OSS organization can lead the way for the benefit of its own business as well as for the vitality and innovation of a robust vendor community.
It is not simply a matter of being more efficient, saving costs or making management easier. Without a more robust back office, the stated goals of the cable industry simply cannot be achieved.
It is time to think of back-office investments as revenue-enabling. It is key to an "any device, any content, any time" delivery approach. Developing a comprehensive back office architecture is not a chore, but an opportunity for the industry to to position itself to increase revenue per user in ways that even subscribers will welcome because of the capabilities they get in return.
Edwin Smith is director, pureIntegration. Reach him at [email protected].
