IP contact center technology: eliminating the risks (Part XI); Achieving multisite efficiencies.

IP contact center technology provides multisite organizations with the opportunity to gain economies of scale in their technology investments by replacing dedicated site-specific systems at different locations with solutions that can be shared across all locations. These economies of scale are driven by:

* The elimination of dedicated technology license pools (for both individual sites and for individual communications technologies);

* Greatly reduced hardware requirements (by sharing resources across locations);

* The aggregation of telecom resources; and

* Dramatic reductions in IT management costs resulting from the consolidation of technology resources.

In a solution designed to support diverse sites, the solution's application logic and supporting software (Web servers, mail servers, application servers, etc.) can reside in one or more data centers. Sophisticated multisite solutions can also remotely control geographically distributed IP voice gateways at an unlimited number of locations.

With nearly all voice-over-IP solutions, individual sites can receive pure IP-based traffic without installing any local phone lines at individual locations. However, there are often compelling reasons for companies to deploy local telephone lines at different locations as entry points or exit points for the corporate voice network. With better voice-over-IP solutions, centralized application logic can actually control IP voice gateways around the world, enabling application logic to span across sites. This dramatically reduces IT costs (and the skills required) to support individual locations. In this model, phone lines are distributed across individual locations around the world yet remain under the control of centralized application logic. The benefit is that application server(s), database(s), Web server(s) and related infrastructure can be aggregated in one or more data centers to drive economies of scale.

The "Multitenant" Value Add

Multitenant IP contact center solutions extend these benefits for organizations with autonomous business units--where each requires their customized technology-driven business rules. The core benefit of the multitenant approach is that autonomous business units can be treated as separate "tenants" on shared infrastructure while protecting the data privacy of each individual business unit.

In previous columns, we've examined the issues of scalability, reliability and network security in a multisite context as well as how to assess a proposed solution's software architecture to ensure that these requirements will be met. (Feel free to e-mail us for soft copies.)

Companies with business units in different time zones should be sure to select solutions that offer time-zone sensitivity. Different business units operating in different time zones will each require their own real-time and historical reports to be presented in the context of their own time zones. (The same is true of date formats and monetary formats.)

Another key differentiator between multitenant solutions lies in the ability (or lack of ability) to provide local managers with comprehensive control over their own technology-driven business practices.

Why Local Control Matters

Since individual business unit managers are responsible for their own business results and are closest to the customers they serve, they will understandably prefer to maintain control over their own infrastructures to ensure responsiveness to the needs of their specific business unit. Most business unit managers will therefore be reluctant to rely on remote IT organizations to service their operations. Multitenant solutions that deliver fully segmented control over technology-driven business processes provide companies with the best of both worlds: the economies of scale resulting from shared systems without sacrificing the autonomy of individual business units.

The best solutions also provide business unit managers with real-time control over their own technology-driven business processes, providing them with the ability to adapt to changing business conditions in real-time. This can provide them with even greater control over their own "virtual" infrastructures than what they could expect from traditional, dedicated premise-based solutions.

Since the root of the value proposition lies in reducing costs by sharing common infrastructure across locations, the rest of this month's column will understandably focus specifically on multitenancy and the radically different architectural approaches that different vendors have taken to provide multitenant capabilities; and how these differences can impact on the success or failure of corporate objectives.

True Multitenancy Versus Primitive Partitioning 

There are two types of multitenant solutions. Both offer data segmentation. First, there is the class of primitive solutions where a single process (i.e., an ACD or routing function) is shared by all of the tenant companies (also called "in-tenant partitioning"). Second, there is "true" multitenancy where each tenant has dedicated application executables that are not shared--but those executables nevertheless rely on a shared hardware infrastructure, shared phone lines and shared software licenses. This latter approach offers compelling advantages:

* Dedicated resources are tenant-specific. This prevents other business units from disrupting the operations of other tenants.

* Tenant application software is segmented to eliminate the problem of "spaghetti code." In addition, ongoing systems integration (to accommodate the needs of each business unit) and all customization can be isolated. This means more reliability and predictability in operations. The key point is that you can accommodate the needs of an unlimited number of "tenants" without the risk of introducing instability for everyone.

* Control can be given to local business unit managers over their own "virtual" infrastructures without risking data security or stability for other tenants. This addresses the "too many cooks spoil the broth" problem that occurs when too many people are working on a single executable.

* Mirrored, hot back-up of individual dedicated resources on a tenant-by-tenant basis is also possible with solutions that offer software process segmentation. This is because they are designed as "network-based" executables that can run multiple instances in parallel across multiple servers. This means higher service levels for mission-critical applications.

Solutions that segment technology configuration information without software process segmentation (i.e., within a single executable or shared set of executables) offer a more primitive partitioning scheme. This approach is called "in-tenant partitioning" but it is often passed off as multitenant technology by providers who cannot segment software processes. You should be aware of the differences and how true multitenancy has a positive impact on efficiency, system stability and cost of ownership. When vendors talk about multitenancy, the litmus test for whether they're really referring to "in-tenant partitioning" is whether each tenant has its own set of software processes that run on common servers and leverage common gateways, licenses and phone lines. If there is only one ACD software executable (or one e-mail management or chat software executable, etc.), you're looking at a solution that offers in-tenant partitioning, not true multitenancy.

The Proper Role Of Partitioning

While primitive in-tenant partitioning is not a viable substitute for true multitenancy, there are circumstances where it can also add value to an individual business unit's corporate objectives.

In-tenant partitioning can be appropriate where segmentation offers a benefit within a business unit, as when a shared pool of agents must work on fully segmented campaigns. This is common in an outsourcer scenario where agents will work one call on behalf of a first client and the next call on behalf of another. Many companies, even single-site organizations, can benefit from software separation across campaigns. For example, a company might want separate points of control for managers handling order fulfillment, customer service and internal benefits projects.

With well-designed "in-tenant" partitioning, business units can:

* Leverage a common pool of agents (and their pre-defined profiles);

* Leverage content libraries across segmented system partitions;

* Fully segment private data between outsourcer clients (who may be their competitors!);

* Generate reports that aggregate data across partitions for strategic analysis; and

* Provide management personnel with a real-time view of organizational performance across business unit partitions.

These benefits have special value to outsourcers because they need to measure performance relative to each corporate client as well as on an overall basis for themselves.

In the best of all worlds, your multitenant solution should offer both in-tenant partitioning and multitenancy; i.e., in-tenant partitioning capabilities within each provisioned tenant. This approach recognizes the efficiency of leveraging different tools to achieve different objectives, with no sacrifices.

The Importance Of Due Diligence 

There are diverse definitions of multitenancy. Undertaking due diligence of any multitenant solution's software architecture is important because even the most architecturally unsound approaches can often survive cursory lab analysis--because lab demonstrations typically don't test against actual business case requirements at scale. Lab trials also don't typically test against a realistic number of tenants, agents, workgroups, tenant-specific integrations, overall load stress, and the realistic pace of ongoing demands for changes to technology-driven business processes by individual tenants. Simply put, they generally don't accurately represent the complexity of real-world needs at scale and changing business unit demands. As a result, low-scale lab trials can result in the approval of multitenant solutions that can become completely ineffective when run in a large-scale environment or across diverse business units. Understanding the software architecture of a proposed multitenant solution can reveal logical "holes in the story" that lab trials can often miss. Once such an understanding has been achieved, the up-front application of common sense can save your company a tremendous amount of time, money and institutional pain.

Eli Borodow is the CEO of Telephony@Work, the leading provider of adaptive, multitenant IP contact center technology for contact centers and service providers. He can be reached via e-mail at [e-mail].

Kevin Hayden is the Director of integrated Contact Center Solutions at TELUS Communications Inc., a tier-1 telecommunications carrier in Canada and the Canadian leader in hosted contact center services. He can be reached via e-mail at kevin.hayden@telus.

A Special Editorial Series Sponsored by Telephony@Work