Why E-Mail On The Mainframe Still Makes Sense

By Tim R. Bardwell

Peter Drucker of the Claremont Graduate School of Business once said: "The only feasible solution to our economic problem is to raise productivity, and the only way to accomplish this is by raising the productivity of knowledge work." Today, in this last decade of the 20th century, E-mail is one technology being turned to in an attempt to keep up with the pace of business change.

Arthur D. Little, the management consulting firm, recently conducted an exhaustive survey of 328,000 sites in the United States, focused on companies with annual revenues exceeding $100 million. The survey revealed 83 percent growth in the number of business sites using E-mail between 1991 and 1993, while the total number of E-mail users grew 17 percent during the same period. Fifty-five percent of sites had implemented E-mail in some fashion and about 20 percent of remaining sites planned to evaluate or implement E-mail within two years. Surprisingly, 90 percent of sites with 1000 or more employees already use some form of E-mail--a fairly high level of saturation.

In an evolution much like the telephone more than half a century ago, many companies are coming to view E-mail as a necessary utility. However, even if statistics are accurate in portraying a virtual stampede toward E-mail, any business would be well-advised to evaluate E-mail for Return On Investment (ROI) within the context of its own particular business structure and market conditions.

How can you tell if E-mail--and, in particular, E-mail on the mainframe--is a good investment for your company? Answering several key questions will help you make an informed decision.

Why E-mail at all?

In some companies, management asks the question: "Why do we need E-mail when we already do messaging using FAX or voice-mail?" That question can be answered by summarizing the benefits of E-mail.

• Better decision-making. E-mail facilitates contacting other co-workers to present and discuss issues, and make timely decisions. It provides easy access to key people (managers, executives, specialists) for advice or consultation and, in contrast to voice-mail, can present the data needed for a business decision.
• Shortened cycle times. Whenever people can exchange information (and the supporting data) without being in physical proximity, this can translate into faster design, error correction, trouble-shooting or quicker responses to market conditions.
• Faster competitive feedback. Some companies bring tailored newsfeeds from public value-added networks into their corporate E-mail system for distribution to pre-arranged mailing lists (e.g., key executives, marketing, regional managers). This is another way E-mail becomes a highway for information distribution.
• Broadcasting time-critical information to all employees. Bulletin board subsystems provided by some E-mail products make it possible to broadcast critical information to groups of employees. Users can be notified whenever new information is posted on a board to which they have access.

End-users usually believe E-mail is a worthwhile investment because it makes their jobs easier. However, many E-mail benefits are "soft"--though readily apparent to end users, they are not always easy to quantify. Proving a hard-dollar ROI is more difficult because of hidden or interrelated cost factors.

Can E-mail be proven to be a good investment? E-mail is often justifiable in hard-dollar terms and, in fact, can be justified more easily on the mainframe than anywhere else. Person-to-person messaging may only represent 20 to 30 percent of potential cost savings. Applications built on top of E-mail can be shown to shorten business processing times, improve responsiveness or cut administrative costs.

Why Mainframe E-mail?

If you are already convinced E-mail is an important technology to implement, the next question you face is a platform decision. Given the downsizing phenomenon and the accelerating growth of LAN E-mail products, why put E-mail on the mainframe? There are several good reasons.

• Ubiquity. For many organizations, the mainframe is still the hub of operations and data management, providing the closest form of universal access whether the user has a terminal or PC. In addition, the messaging infrastructure or highway (a wide-area SNA network) is already in place.
• Scalability. The mainframe is still the lowest-cost approach to providing computing to large numbers of users. LAN E-mail pricing is almost always user-based; mainframe E-mail is priced by site or CPU and thus supports a larger number of users for less.
• Economy of scale. Centralized management, administration and technical trouble-shooting are already in place at mainframe sites. This means people-costs of implementing E-mail on the mainframe are lower than LAN E-mail costs, fully loaded. A leading E-mail consulting report, The Ferris E-mail Analyzer, has estimated the cost of running a PC network at about $2500 per PC per year and the cost of adding E-mail to a PC network at about $400 per mailbox per year. That is before adding the cost of technical support and buying and administering applications running on the network.
• Sunk cost. Since mainframes in most companies have already been funded to support other corporate mission-critical applications, up-front platform cost is not as critical a factor as it is on the LAN, which usually requires considerable new investment. With LAN E-mail, there are requirements to purchase PC servers dedicated solely to E-mail service, upgrade existing PC workstations with more memory, add further communications infrastructure (Ethernet, token ring, etc.) and hire or train personnel to integrate and manage LAN network resources.
• Integration. A mature mainframe E-mail system offers not only messaging but fairly complete integration of bulletin boards, filing cabinets, telephone directories, calendaring and scheduling, and forms routing. It has the advantage of close coupling with mainframe systems facilities, allowing the user, for example, to view and manipulate partitioned data sets. In contrast, many LAN E-mail systems require third-party add-on products to provide these additional capabilities, exposing you to further integration and support issues.

While a good case can be made for running enterprise-wide applications on the mainframe, it cannot be denied that there are also many advantages to LAN-based E-mail, such as a GUI and better attachment processing. What some companies have discovered to their painful surprise is that the cost-advantage of LAN E-mail at the small workgroup level rapidly disappears in proportion to the growth in size of the messaging user base.

When deciding platform and architectural strategies, E-mail can become a political issue resulting in a prolonged battle and consequent delay in achieving the productivity gains you seek. For the forseeable future, a more pragmatic strategy for many organizations often consists of "doing a little of both."

In defining an E-mail platform strategy and cost-justification, there are four questions that should be asked. The answers will indicate which solution (LAN, mainframe or both) you should use as well as how you might cost-justify your E-mail strategy.

Question #1: Is the E-mail need "groupwide" or "corporatewide"?

If the need is limited to a local workgroup, then a LAN E-mail system might be more cost-effective and easier to implement. However, if messaging needs will extend across the corporate organization, then you should take a closer look at the logistical and support advantages of a host-based messaging system that already has the infrastructure in place to support an enterprise-wide reach.

Question #2: Would phone traffic be reduced using E-mail as an alternate technology?

Most corporatewide E-mail implementations exhibit a flattening or decline in phone traffic as a result of being able to move messages over the computer network. Do you have company personnel operating in a geographically dispersed manner across several time zones or across international boundaries? E-mail will allow communication to take place without having the other person on the phone in "realtime," reducing the demand for telephone connect time.

TABLE 2: CORPORATE E-MAIL STUDY
Length of each call:	3 min./call
# Calls per employee:	5 per week
# Employees:	574 employees
Phone time saved:	8610 minutes
Avg. cost of call per minute:	14 cents per call
>Total weekly phone savings:	$1,025 per week
Total yearly phone savings:	$62,681 per year
Source: Ricoh Corp.

By looking at the routine communications that occur frequently in a manager's or secretary's day, you can forecast potential phone-call displacement. For example, one large office-product manufacturer measured the length and frequency of phone calls to and from its regional offices. As shown in Table 2, their E-mail pilot revealed they would be able to reduce phone calls to the field and realize substantial savings as a result.

Phone-call displacement has been documented at a number of other sites where corporatewide E-mail was implemented over a WAN.

Question #3: Are there periodic requirements to send documents between offices?

The office-products company previously referenced also discovered that document distribution (of dealer listings, major account plan summaries, sales-lead distribution and technical documentation) could be handled in a paperless fashion via E-mail. These capabilities added up to an additional $9652 savings per year. When it was discovered their E-mail system also provided intelligent forms-routing, the company began to think about how it could process expense reports in a paperless fashion. This leads to the fourth and most important issue--how to exploit E-mail by developing "mail-enabled applications."

Question #4: What paper processes could be automated?

Many business workers spend a considerable amount of time filling in forms and routing them from person to person or department to department. Most of these processes can be automated by using an E-mail package providing intelligent forms-routing. This can be accomplished on the mainframe with little additional effort and is a good way to reduce programming backlog by allowing nonprogrammers to design and implement their own automated forms-routing procedures.

Forms-Automation Examples

Security change requests and timecard reporting are particularly good candidates for forms automation because they are common to almost every company with a mainframe.

Every I/S department has to handle security change requests. Mainframe users are granted authorized access to the host after filling out a profile form that is signed by a line manager and approved by the corporate security administrator.

In a large organization with several thousand users, the administrative work of adding, deleting and updating user profiles can be quite heavy. An electronic version of this form, accessible only by the security administrator, can be set up by the E-mail administrator. Then, each new user working with a temporary ID is E-mailed the security form and fills it out on-line; it is automatically routed to that user's manager for approval, then back to the security administrator's inbox for further validation. In this way, an administrative process taking several days can be compressed into several hours or less.

Timecard reporting is another process that can be automated using intelligent forms-routing. An aerospace company is required by Department of Defense regulations to keep detailed and accurate records of labor, allocated by contract and subcontract. In the past, this was handled through a manual timecard system, but the process was slow and error-ridden.

Using the forms-processing language provided by its mainframe E-mail system, this company was able to create an on-line timecard "front-end" to their CICS applications. Employees select their timecards from an on-line departmental cabinet and update them on a realtime basis throughout the day.

The intelligent form calculates all elapsed-time intervals and wage amounts so the risk of user error is almost completely eliminated. At the end of the day, each employee invokes the APPROVE function, causing the form to be launched to the department supervisory level for approval.

Once supervisory approval is obtained, the form converts into a standard E-mail message and is sent to an E-mail inbox that is defined to a CICS transient data queue. The timecard form application saves the company time and money by shortening the processing/approval cycle, reducing human error by providing a layer of computerized validation and checking, and reducing paper consumption.

Forms Routing Leverages E-mail ROI

A forms-based process will need to be analyzed and broken down into its component steps to isolate potential time savings. In one example, a company had 221 employees who manually filled out and processed timesheets.To justify moving to an automated forms-processing system built on top of E-mail, a time/motion analysis was conducted.

The company found that forms-routing automation provided an annual ROI of 135 percent and that additional forms would boost the ROI correspondingly. According to the Ferris E-mail Analyzer, forms automation is a low-cost, low-risk approach that has yielded savings of 1 to 2 percent of a company's revenues. Such savings can have the effect of boosting a company's profitability by as much as 10 percent.

By applying E-mail technology to specific application areas (such as document distribution, telephone call displacement and forms-routing improvements), a hard-dollar justification can usually be made for E-mail.

Companies wanting to stay competitive into the 21st century need to embed E-mail technology and applications into the framework of many business processes. By implementing messaging for as few as 100 mainframe users, companines can obtain a payback period of less than 24 months. An added boost comes from implementing a forms-processing system on top of E-mail. It is here that a company can squeeze thousands of additional dollars from operating costs and boost ROI significantly. For companies that seize the opportunity, mainframe E-mail can help make a significant contribution to the bottom line.

Perhaps the answer to the question, "Why put E-mail on the mainframe?" was best stated in Sir Edmund Hillary's famous response to the question of why climb Everest--"Because it is there." Because the mainframe is already "there" in support of other applications, it is already a sunk cost in evaluating the savings and productivity gains generated by mainframe E-mail and forms-routing automation.

E-mail on the mainframe is generally a sound business decision; in fact, given the central presence of the mainframe in nearly all large organizations, it is surprising it has not been done more.

---

Article originally published in Enterprise Systems Journal, September 1994.

About the Author: Tim R. Bardwell has spent 18 years in data processing in various marketing and management roles. For the past several years, he has worked with E-mail and mail gateway software at Computer Application Services, Inc. (CASI) as well as at Soft·Switch.