When it comes to self-provisioning, the industry needs more than just a front-end illusion. Back-end issues must be tackled before customers are allowed to fully manage their accounts online. Otherwise, a company opens itself up to a nightmare where customers lose trust in the company’s ability to automatically process an order from the time it is placed until the service is received and billed for. “If it’s not done right the first time,” says Mark Palmer, vice president of operations and service quality at the Canadian Cannect Communications. “it’s such a bad experience, they [the customers] won’t ever want to try it again when it’s working.”
It will take some time, industry experts say, for widespread adoption, but flow-through self-provisioning is coming. “Pretty much everybody has this in their [business] plans,” says Steven Bamberger, director of marketing at Cygent. “We rarely talk to anyone where they say, ‘Hmmm, I don’t think I’m going to advocate budget for that, I didn’t allocate a budget for e-business this year, and I’m not going to do it next year either.’ I just don’t hear that.”
Many providers want to include this functionality on their online self-care; yet many obstacles, with both the business and network systems, must be overcome before companies will give the customer control over such vital elements of their operations.
HOW IT WORKS
At Sigma Systems Group, once the systems have captured an order, the order is marked as ready to be provisioned and goes to the provisioning system. Once that system responds that the order has been provisioned, the order capturing system is notified that the service has been provisioned, and a request is written to the billing system to start billing.
An example is Sigma’s work in the cable space. Using the company’s Web commerce product, a customer can buy a cable modem, bring it home, log on to the Internet and register for service online to have instant access to an e-mail account and Web space. The provider can activate the service automatically over the Internet. “If a user can self-activate their cable modems, there is no need to send a technician and no need to have a customer service representative deal with clients. It’s a tremendous savings,” says Robert Bratulic, director of product management at Sigma Systems Group.
At Telution, order management, including self-provisioning, is broken into three parts, the first of which is the Web front end. Here, customers can view charges, sign up, order products or bundles of products, manage them, change configuration values, set up preferences and lodge trouble tickets.
The middle third involves the business processes, including an application model at order management, customer care and billing. Here, messages are sent back to the appropriate application to show the customer the invoices, or a customer can be walked through the ordering process.
The back third is really a provisioning workflow piece that touches all of the network elements or the partners that need to be involved in actually turning up one of those services. It is where service bundles are decomposed and work requests are sent out throughout the organization to initiate each division to complete its task. This third part ties all the pieces together at the end, so when a product is activated, the system automatically starts collecting usage records to bill for services.
FIRST, IN PHASES
Today, most providers have automated only the top layers of the Telecommunications Management Network (TMN) stack, while the bottom layers are totally manual, according to Brad Ketch, cofounder and senior vice president of HyperEdge. The Telemenagement Forum’s TMN model conceptualizes four layers from the top down—business management, service management, network management and element management—and how they are interrelated and work together. “If you don’t have an automated path through every layer of that stack, you don’t have automated provisioning,” he asserts.
According to Bratulic, when most companies make the transition to Web-enabled self-care, what they often prefer is to ease into it by having Web-enabled screens, with manual processes behind that. Clerks will print the orders from the Web at the end of the day and then manually process them. Providers “are usually wary of a new customer coming along and touching their network right away,” Bratulic says. “It’s a way of testing the system as well.” This way, a company can see whether the graphical user interfaces work properly on the Web, then have the orders dumped into a file, have human beings provision them, work out the kinks and then, when the company is comfortable, move toward fully automated provisioning. A down side to this approach is that a customer has to wait for service.
Cannect echoes a step-by-step philosophy in its business strategy. “We’re looking at starting off modestly with one of our simpler products that would be able to allow customers to initiate order provisioning over the Web,” says Bill Clendinning, information systems manager in Cannect’s Operations Department.
While Cannect is currently working on the trouble ticket and care side, where a customer can open, view and manage a trouble ticket online, the company has not yet started customer ordering via the Web. It is currently working with Sigma and Nortel to develop this functionality.
Many believe it will be easier to add or change services in a self-provisioning set-up versus allowing a customer to sign up for new services from scratch. “If they have four lines and they want to add a fifth line, or they have a 5-megabit circuit that they want to make a 10-megabit circuit, we already know the configuration,” says Cannect’s Palmer. “It’s simple to take a check box that says, ‘I want one more of these and I want it here.’ There are fewer nuances and intricacies in that environment, as opposed to a new customer coming to us and saying, ‘Hey, I’m a brand new customer. Here is what I want.’ ”
THE NETWORK LAYER: DRILLING DOWN
Tackling the challenges at the physical layer will not only increase the speed of service, but also dramatically lower the cost of provisioning. The number of call center personnel could be reduced, says Jim Alsman, analyst and Frost and Sullivan. Also, the accuracy of orders could dramatically increase if a provider’s systems offered flow-through provisioning.
Because self-provisioning breaks down at the physical layer today, companies must address this head-on to make sure the process is successful. Even handling some processes manually, many experts say, defeats one of the main points of self-provisioning, which is to have completely automated flow-through capability.
Ketch explains UNELink, HyperEdge’s automated loop management system that addresses the physical connection. HyperEdge is partnering with an OSS company to fit in Web browser functionality. This would enable an end user to log on, submit an address and be taken through a qualification. Eventually this process would access the loop management system and tell it to take the physical line and switch it over to a test head that would apply a tone—various signals to determine the speed of service a line can support. The end user would be asked to wait a minute for the results, and then the Web interface would tell the user what the available options are—what line speeds can be supported at what price. The user would select the desired service, and that choice would go through a billing routine. When that is all set up, the OSS would find the loop management system and instruct it to take the customer’s line and switch it over to the appropriate DSL port on the network element that’s sitting there. This reference model design would do that instantly and return a message to the Web browser telling the customer he or she is now on the service.
Ketch says the physical layer is an essential piece for self-provisioning. If a company has to dispatch a technician to a collocation cage to turn on service, for example, the point of self-provisioning is lost, because full automation is where the true benefits lie.
THE CHALLENGES
Because the industry goal for self-provisioning is to fully automate the process, any mistakes made along the way can have a dramatic impact on the billing system. Several significant obstacles will have to be overcome.
Customer Know-How
Most billing mistakes happen at the point of order entry. Today, this happens even with trained CSRs guiding the process to establish accounts. “If we don’t understand the nuances of what [customers] have and how they use it,” Palmer says, “they are going to end up with something that doesn’t work the way they want it to.”
A lot of the complexity in ordering online boils down to how well a provider can simplify its Web interface. “It can’t be rocket science for a consumer to do this.” Palmer insists. He believes there must be an adequate level of instruction, as well as checks in the system to ensure that a service is provisioned and billed for correctly.
“You don’t want to go to a lot of trouble to build up this Web-enabled front end and then have the user get lost in technical terms and features they don’t quite understand,” Sigma’s Bratulic notes.
Self-provisioning developers must think about the functionality of their Web interface. It must be efficient, load quickly and prevent the user from getting frustrated because the system cannot process an order based on user input.
Internal and External Systems
Both internal and external systems are a tremendous challenge when a provider tries to implement self-provisioning. Systems within a company must be able to speak with one another to ensure the entire workflow is accomplished and that a service is provisioned.
Cannect’s Clendinning says, “Usually the systems are so distributed you have to work on them every time something happens. You have to build bridges to 14 separate systems, in some cases.”
But being able to talk within the walls of a business is not enough. One provider’s systems must also be able to talk to those of another provider.
When a provider must change a customer’s local phone service from Bell Canada to Cannect, Palmer explains, the challenging part is getting the request to flow through a variety of different systems from different vendors without human interactions. “That’s a ways down the road, I would suspect, before that would happen,” he predicts.
Hurdles for ILECs and CLECs
ILECs will have the most difficult time getting systems—both internal and external—to communicate with one another. Every time they add a new product, they typically add a new provisioning and billing component, which has made it very difficult to get all of those things to work together, let alone talk to other providers’ systems.
Many agree that CLECs will lead the race to initiate self-provisioning. Palmer believes it will take the ILECs time before they head full-force with this capability. “They typically watch things like this happen for a period of years before they are able to turn the steamship around, going down the river with the current,” he says. The ILECs are typically so independent and distributed that it will be difficult for them to sew together their systems in a seamless fashion, making self-provisioning a far greater effort.
Ketch at HyperEdge believes CLECs will not be able to offer complete self-provisioning on their own. “They must have cooperation from the RBOCs, or else they won’t get access to the copper line that’s sitting out at the end-user’s site,” he notes.
The legacy copper vendor must be able to switch lines over to the CLEC quickly, so the CLEC can provision the order.
In addition, if a company uses a single DSL provider and tomorrow it wants to use someone else, but maintain the same Web front end and the same middle application layer, the provisioning workflows will change when the new partner comes on board.
Mediation Is Key
Because communication among internal systems is so vital, mediation vendors will play a key role in enabling self-provisioning between the network elements and the provider’s OSS systems. “It is incumbent upon the mediation companies to provide the hooks into the existing OSS that the network element needs,” Ketch says. “The telcos have these big mainframe OSSs, especially the RBOCs. Network element vendors don’t always have hooks into those.”
Some of these hooks include TL1, X.25 and other high-level protocols. The mediation software provides the formatting function to convert output of a network element, such as SNMP, into the data format of the OSS, such as TL1.
Kent Steffen, president and chief executive officer at Telution, believes mediation is essential because front and back offices can no longer operate separately. They need to work together to give customers real-time access to usage and other account information. Just as processes to provision a service need to flow through a provider’s systems, the information collected off the network needs to flow back through to the customer.
BILLING ISSUES
Beyond the challenges of designing user-friendly interfaces and coordinating the disparate systems that come into play, self-provisioning poses several issues for the accuracy and reliability of the billing process.
Bundling Nightmare
Bundles are a very complex art and science. Creating them, therefore, would be a tricky task, especially if a provider enables customers to create their own bundles online. To avoid this, many experts agree, companies must offer a predetermined package of services.
One complexity is that many business rules are involved in a bundle. Unknowingly, customers could try to self-order services without having the proper existing services in place to support another application. Long distance, for example, cannot operate without a basic line. Business rules must be in place to prevent such conflicts, or else customers could accidentally be billed for services they are not even qualified to receive.
Another big bundling issue that must be addressed is time-to-revenue. “Providers need to be able to manage the different pieces of the bundled order together so that the pieces of that order that can, flow—and the pieces that can’t, get the manual attention—so you’re not holding up the whole order for one piece,” says Cygent’s Bamberger. Providers, therefore, must have a way to manage provisioning processes together or separately, even when they require manual intervention.
“If I have 10 services that are on the same order and only one of them is being held up, I want to be able to bill for the first nine, but I want to be able to manage each of those separately,” Bamberger says.
He says providers must think of services as both independent offers as well as bundled services. “If I have services A, B, C, D and E, and bundle one is comprised of services ABC and bundle two is comprised of services CDE, as long as I have the ability to manage those as both independent offers and bundled services, the fact that C is in both bundles is really irrelevant,” he says. “But it is a challenge to make sure that your catalog supports that kind of structure, where a service can either be a stand-alone offer or part of one or more bundled solutions.”
Bamberger says that the catalog must enable a company to define how a bundle fits together and not treat every bundle as a fixed set of options. The infrastructure for complex bundles must go beyond defining every bundle as an immutable set of services, and all back-office systems must understand their piece of the bundle.
Down Time
Self-provisioning brings with it the customer expectation of being able to order or check on the status of service 7 days a week, 24 hours a day. Yet companies must back up their systems. What happens if a system is down or is being backed up?
Reza Fikree, senior architect at Sigma, explains how his company’s product handles this. “We use our transaction servers to do the self-provisioning. These servers maintain fail-over and load-balancing capabilities. When one server is being backed up, it is actually replicated to more than one server; therefore, the other server can take it and take care of it.” In Sigma’s case, customers can choose what to do during backups. Fikree says some clients prefer not to allow users any functions such as self-provisioning during backup. Others have no problem with that and use data replication engines to allow self-provisioning during backups.
Fikree emphasizes that communication between provisioning and billing systems is critical. He cites three-way calling as an example. When the user requests the service, the order is captured and goes through a provisioning system; once a provisioning system has provisioned it, a number of procedures are executed. The first is to find out what billing system the order will interface into. The billing system requires certain parameters, such as an account number, the start date for the service, how to bill the service and whether it is part of a billing plan, as well as any associated discounts or coupons.
A company may, for example, offer the first month of a service free if a customer uses its three-way calling. The service is provided for a month, but the first month has no billing. Thereafter, the customer is billed at a certain rate. But what rate would a provider apply to this billing? What plan does this customer belong to? Is there a bundle associated with this? All of these parameters will affect the billing system.
Steffen at Telution says providers must determine which system owns the product catalog. Just as the order entry system must capture information associated with products, a billing system has its own way of thinking about products, typically centered on pricing plans, discounting and rating. Providers must work to keep these systems in sync.
Giving Customers Control
Some fear enabling customers to sign on to a service or make changes to existing services, given how customers will access their business information. “If a provider can partition the network or allow the enterprise access to only circuits that they control, then that’s one thing. But if they give them access to, say, a network element that may have other enterprise customers on it, I can’t say that you’d want to give them gratis access to that,” Frost and Sullivan analyst Alsman says.
Providers face the challenge of having to complete flow-through provisioning with such precision that they do not disappoint customers. As Steffen points out, if you take away the CSRs and make the total customer experience dependent on a Web front end, any hiccups from back-office systems could give a negative impression of the company. “You don’t have that person in the middle that’s running interference for you,” he says. Still others fear giving the customer an easy method to also disconnect from a service.
Yet all agree on the importance of empowering customers to manage their own accounts. While self-provisioning could put a friendly face on order management, it is more than just an online façade. Back-office issues must be dredged out before companies and customers will sign on to self-provisioning en masse.
[Pull Quotes]
“You don’t want to go to a lot of trouble to build up this Web-enabled front end and then have the user get lost in technical terms and features they don’t quite understand.”
—Robert Bratulic, Sigma Systems Group
“If you don’t have an automated path through every layer of that stack, you don’t have automated provisioning.”
—Brad Ketch, HyperEdge
Email this article
Add a comment
Printer version
Order reprints
RSS Feed
Bookmark article