One of the simplest, clearest cases for VOIP Ive ever heard was made in the retail environment and had to do with distributing the customer-service workload among physically separated personnel. Its a case study Ive written up many times, featuring many different vendors. It goes like this:
A chain of stores/restaurants/beauty salons is spread across part of a state. Each has its own two analog phone lines, or perhaps a small, six-line key system; each has its own listing in its own local Yellow Pages, its own voice-mail service or answering machine, some customer-facing employees and some back-room personnel.
In the “before” picture, we see a potential customer calling the store that happens to be closest. We see an overwhelmed counterman too busy serving customers to answer the phone. The caller gets a busy signal or an offer to leave voice mail, hangs up and dials the next store/restaurant/salon listed in the Yellow Pages. Her business is lost.
In the “after VOIP” picture, we see the caller dialing that stores same published number. This time, if no one picks up on the third ring, a greeting is played that asks the caller if they want sales, directions or hours of operation, and please press 1, 2 or 3.
The last two choices offload a good percentage of calls all by themselves. The third (sales) choice rings every phone in the chain. (It also could be set up to ring the next available line.)
Whoever picks up can see which line the call is coming in on, and therefore which store has been called; customers never need to know whether branch A or branch B is helping them.
In the “after VOIP” picture, each branch store is hooked up with IP phones to DSL Internet connections. These phones act just like typical digital PBX extensions, but they exchange their signals and their voice streams over DSL to the IP PBX at the central branch.
These manufacturers typically have prior, or perhaps concurrent, product lines in traditional digital phone systems. They try to preserve all of the features of these legacy systems when they revise them for IP transport.
Such features may even include intercom, so that staff can back-channel through the phones to visually confirm that a part is in stock across town, during the customer call. Ive also written variations on this theme that involve bar-code scanning wireless phones, which verify inventory on the same session as an IP call.
This application scenario goes by the name of informal ACD (automatic call distributor). It means that as long as anyone–in any location–is available to pick up the phone, that caller will not receive a busy signal.
Incoming calls will automatically be sent to the next store if theyre not answered. Those personnel in less-busy branches will pick up the load for those who are swamped.
Transferring calls between branches, similarly, will be as easy as transferring calls between extensions–easier, in fact, since new IP phones tend to take the mystery out of these tasks with LCD prompts.
So, if a Spanish-speaking caller calls, she can be routed to the one employee–in a Latino neighborhood branch, lets say–who can help her.
Automatic call distribution, provisioned through the traditional telco network, has been too expensive to be practical for these little chains. Across data links, its self-provisioned and indifferent to physical location: Since the phones are IP endpoints, they can be picked up and plopped down anywhere an employee might go, just as your laptop can register its IP address to an MSN Messenger server and let you start instant-messaging from wherever you connect.
Its not only about pooling manpower; its about pooling phone lines. Each branch still keeps a PSTN lifeline for just-in-case and for E911 compliance, but it gets to dump any others.
The IP PBX-equipped branch gets additional PSTN lines and gateways out to them to take in the extra incoming call load–perhaps bumping up to fractional T-1–but it gets better volume pricing in the process and pays fewer associated mystery charges. Another savings: Whatever the chain used to spend on interoffice calls goes away.
The smallest scale of this story that Ive ever written concerns a chain of four replacement-parts stores around Dallas and stars the ESI Estech system, for which installation ran about $14,000 in late 2000. But its a concept that the bigger vendors–as well as IP Centrex startups–have taken up on a grander scale.
Kanata, Ontario-based Mitel Networks, in the larger but still midsize space, has recent customer win announcements with CompUSA and Auchan, a chain of mega-markets in France.
For these megachains, both the IP links–over a managed WAN, not mere DSL–and the session capabilities are much fancier, multimedia and far-ranging. But the benefits are similar, first of which is pooling your human resources for the maximum customer-service load.