The most elaborate technology will deliver precious little if not chosen and managed properly By During the 1996 election campaign, Bill Clinton had a sign hung up within his various campaign headquarters that read "It's the economy (stupid)" to remind himself during the campaign of what was the most important point to make. With the myriad issues, technologies and trends about which a CIO is regularly bombarded with information, he or she would do well to have a similar sign behind the door of his or her office reminding them of what is really important. Communications management is a multi-faceted position, much of which is concerned with technology. Management involves cost control, it involves selection of consultants and major systems, it involves project management, it involves infrastructure management. And more.
As everyone wants to reduce costs, that's a good place to start. In the pursuit of a reduced phone bill, most of corporate (and residential) Australia has sought to obtain the best possible deal for domestic long-distance and international calls. Selection of the most suitable service provider can, of course, reduce costs, and the periodic review of long-distance provider is a necessary element in the practice of cost minimisation. But there are others.
Almost any organisation that can achieve savings from a private network will already have one. But so do many that don't. The perception that private networks inherently save money is as widespread as it is incorrect. There are many organisations with tie-lines traversing cities and states that cost far more than the equivalent cost of the calls they carry. As a general rule, if the traffic is not sufficient to require at least half a dozen lines, it is far more cost-effective to not have tie-lines.
Traffic analysis and dimensioning is not rocket science, but it does require up-to-date accurate traffic statistics, and this requires a TIMS (Telephone Information and Management System) with an up-to-date configuration. Most medium-to-large organisations have a TIMS, but many of these do not have up-to-date configurations or tariff tables. Contrary to a widespread myth, once installed it does not manage itself.
Of course, most private networks do save money; but of those that do, many can be enhanced by adding such features as routing calls to bordering STD zones (for example, from Sydney to Geelong via the Melbourne office), using a VPN (Virtual Private Network) for sites not included and centralised attendant.
Moving on to non-call charges, the biggest savings can sometimes be made by auditing the carrier rental charges. Speaking at the ATUG 95 conference, consultant Paul Morris stated that "unless you have done a professional check of your phone bill recently, then you are probably being overcharged by Telecom (now Telstra). It is as simple as that". Such savings, achieved by eliminating unnecessary or incorrect charges listed within the service and equipment inventory, can be substantial. On an assignment a few years ago, the author saved one leading retailer $250,000 on its NSW operations.
Audits are just one example of savings that can be achieved by better management. Take personal calls which incur both call charges and staff time not working. Despite this cost, I have yet to identify any organisation, either government or a private enterprise, that has a formal written policy on personal telephone calls. Most have some sort of a pervasive understanding.
Then there's inventory management. Most communications departments have cupboards full of unused handsets and assorted paraphernalia. Typically, no one has any record of which ones work, which are still covered by warranty or, in the case of the peripherals, why they were bought in the first place. One property company even threw away hundreds of expensive analogue handsets because they incorrectly believed they wouldn't work with their new PABX.
Good inventory management is not difficult. One quasi-governmental organisation had each handset engraved with an asset number. This has allowed them to be tracked and, if applicable, repaired under warranty without cost and depreciated. This also helped to reduce pilferage which, if it was reduced by only 1 per cent a year, meant an annual saving of more than $7000 for the organisation.
Still on the issue of asset management is the maintenance of larger systems, such as PABXs, servers and the like. A piece in ATUG's August 2000 Newsbrief noted that despite increases in system reliability, most vendors charged maintenance fees in the order of 15 to 20 per cent of capital cost, per annum, arguing for the availability of alternatives, such as a small annual retainer plus a call-out fee. An annual fee does ease budget forecasting, but a lower cost is an objective worth pursuing. Most communication systems can be maintained by other vendors in competition to the selling vendor, and comparison of what they all offer is usually worthwhile.
Communications management is, of course, concerned with far more than just cost control. Management of communications involves the same specialities as one finds in a computing shop - development, operations, technical services and user liaison - the difference is that most communications groups do not recognise these specific functions or apply the necessary disciplines to their practice.
The culture of a communications group, especially one concerned with voice, is very different from that of a computing group and it is this difference why so many corporate attempts to marry the two have failed. Some key differences are:
Reliability is paramount to voice communications staff, but of less concern to computing staff, especially those working with PCs and servers;Voice communications staff thus have a strong preference for PABXs to be DC-powered from batteries with four or more hours of capacity, whereas computing staff are content with servers being powered by UPS, with just enough capacity for an orderly shutdown;Computing staff are accustomed to having operating system software regularly upgraded as part of a mainte-nance contract, whereas voice communications staff are not;Voice communications staff are accustomed to having hardware and software maintenance performed as a package, whereas computing staff are accustomed to separate hardware and software maintenance;Choice of operating system is often paramount to comput-ing staff, even for systems such as IVRs for which it is largely irrelevant, whereas voice communications staff are indifferent to the selection of operating system;Computing staff are generally more conscious of security than voice communications staff; andThe computing departments of leading organisations typically have a group that participates in ongoing user liaison and system development whereas voice communi-cations groups are largely operationally focused.
These differences can manifest in many ways, but particularly with the management of the implementation of key systems. Such projects involve the following five steps: user requirement identification; conceptual design; detailed design; implementation; ongoing operational management, accounting, and so on.
Whether the computing function is performed in-house or outsourced, such a development model is widely understood and practised for computing applications. This development model is equally as applicable to projects involving voice and/or data communications, although this is far less practised for these disciplines.
Within the domain of voice, a major project will usually involve the acquisition of a new system, such as a PABX or a call centre system. As such a project will involve the performance of each of the above five steps and the disciplines to perform the first four are often absent from a voice communications group, a consultant is often required. Whether or not a consultant is engaged (see "The Consultant Conundrum" box), the buying organisation will have contact with a number of potential vendors and other service providers. It is important to not pay much attention to vendors' claims of industry direction, as all vendors have a vested interest in creating the illusion that "everyone" is about to buy or implement the technology that they have just developed or acquired. In addition, ensure your chosen consultant is equally sceptical about such vendor claims.
A Tender Moment
Major systems are often selected by competitive tender. Tendering has a number of advantages.
Tendering requires the organisation's requirements to be determined and formally documented.
Tendering requires vendors to respond to every documented requirement.
Tendering minimises the impact of slick presentations by well-trained sales persons.
By issuing a request for tender (RFT), the organisation is signalling that its interest in the acquisition is serious and setting the agenda, rather than letting a vendor take control.
Most significantly, the process of writing, issuing and evaluating responses to a request for tender forces the organisation to adopt a disciplined approach to the entire acquisition process.
And it's widely practised. Representatives from Alcatel, Ericsson and NEC, the three leading vendors of PABXs in Australia, collectively estimated that two-thirds of large systems they sold to new accounts were acquired via a request for tender, whereas the figure for small systems was less than one-quarter.
Writing a request for tender document can be, when done for the first time, very time-consuming, but much less so if a template is used. Nevertheless, the investment in time will be worthwhile. The more comprehensive the RFT, the better the outcome. Some tips:
Do not bias a RFT towards any one vendor's products or services.
Do not let a vendor assist, as the result will surely be biased.
Involve all relevant business units, including communica-tions and computing.
Stipulate a response format, and define "complies".
Include a draft of the contract that would be used for the purchase.
Separate requirements into logical groups, such as system requirements, installation and maintenance.
Above all, define capabilities, not attributes.
Once the system has been selected, configured and implemented, it must be managed; and, whether or not a consultant has been engaged, the requirement for involvement of the organisation's staff must not be underestimated. Inventory management and maintenance contracts have been discussed in previous paragraphs, but there is also the issue of software version maintenance. Whether the system in question is a PABX, IVR or predictive dialler, there are advantages in having the software periodically updated, as is done with mainframe and PC operating systems. Ensure your vendor will come to the party.
The final chapter in the life of any asset is its eventual disposal. In contrast to computers, most PABXs and other communication systems typically have real commercial value when the organisation that first acquired it disposes of it; and even here, there can be complications.
I recently surveyed a number of PABX, IVR, CTI and predictive dialler vendors and found that while a narrow majority permitted resale without restriction, many require the purchaser of a second-hand system to pay a re-licence fee before using it. If you have a system you are planning to sell, check the original terms of purchase: if there is no stated exclusion, the system can be sold and reused without restriction. If planning to acquire a system, include unrestricted right of resale in the request for tender.
All of the above discussion assumes that an organisation will acquire, own and operate the infrastructure itself. Many outsource.
Former Australian Telemarketing Association CEO Richard Allen estimated that the level of outsourcing would approach 15 to 20 per cent whereas analyst Paul Budde stated: "At this stage 80 per cent of voice infrastructure gets outsourced." Of course, outsourcing occurs in many different formats, ranging from providing an off-site help desk to outright ownership of infrastructure.
Order from Chaos?
Outsourcing has its pros and cons, which CIO has previously covered. Nonetheless, a few points need to be made about outsourcing of voice. There are a number of providers of outsourcing services, only some of whom manage voice and only some of them do it well. However, more significantly, there appears to be a perception among some outsourcing clients that if their communications infrastructure is a mismanaged mess, without any traffic statistics or any documentation, an outsourcing company will wave a magic wand and create order out of chaos, complete with up-to-date infrastructure, properly dimensioned routes and accurate documentation.
Of course, this is exactly what an outsourcing company can and should do - but not for free. An outsourcing company that promises such an audit for no additional cost will have to pay for it out of its outsourcing charges and will be under pressure to do so on the cheap. If they quote separately, review what will be provided. If you're really serious, agree on an independent third party to perform a due diligence report. But if your infrastructure has been managed well, none of this should be necessary.
Assuming one manages the infrastructure well, one must not lose sight of why. Speaking at ATUG 93, George Colony of Forrester Research likened IT management to an iceberg, the bulk of which was underwater. That which was underwater - that is, infrastructure management - was invisible to the other business units, who only wanted to see the tip: that which served their respective business units. The job of IT management was to push the iceberg down.
A communication group that is managing its infrastructure well is thus in a position to focus on better servicing the other units of the business. One managerial development that is not all that new is to have internal account managers within IT who look after the needs of the other business units. It is a good step forward, but many such persons come from a background of systems development and understand computing, thus focusing on the computing needs, not communication needs, of their clients. The challenge facing many communications managers in such companies is to have the same interface to these business units, which requires either that these account managers understand communications, or that the communications group has its own account managers.
While account managers provide an input for other units' strategic requirements, there is also the day-to-day operation, and that interface is the help desk. An entire piece could be written on good help desk design. Suffice to say, a well-run communication group is fronted by an excellent help desk facility, which has some of the following attributes:
Its hours of coverage match that of all other business units - if one operates 7x24, so does the help desk;Its functions must including logging faults, performing minor moves and changes, and providing assistance to the caller on that call, not via a callback the next day;There must be one help desk for all IT functions;It must have a memorable extension number; andIt must be staffed by personnel who have a comprehensive knowledge of the organisation.
Finally, there is the contribution of the communications management group to the organisation's customers and other contacts. Call centres are typically managed by staff who follow recognised disciplines and practices to ensure that the service provided is within objectives. But outside the call centre, telephone behaviour is almost always devoid of any standards of good practice.
What reader has not called someone, heard a voice mail greeting asking the caller to leave a message and left a message, only to discover subsequently that the person they called will be away that entire week? Now who can say that they have taken steps to ensure that this will not occur within the organisation for whom they work? Good telephone practice outside of the call centre addresses voice mail greetings, call screening, the non-use of a PABX's do not disturb "feature" and how to answer the phone. In corporate Australia, telephone etiquette policies are conspicuous by their absence.
Any communications manager can, of course, prepare and issue a telephone etiquette policy, but who will take note? As with so many other aspects of communications management, the communications manager can do only so much without the support of top management. To manage effectively, the comms manager needs to report directly to the CIO who, in turn, needs to report to the MD or CEO. And for the CIO to really manage effectively, he or she needs a seat on the company's board - but that's another article.
Stephen Coates is a Sydney-based independent telecommunications consultant. He can be reached at firstname.lastname@example.orgThe Consultant ConundrumLove them or loathe them, or even joke about them, business today makes a significant and increasing use of consultants. On the face of it, consultants are usually hired for one of two reasons: to provide specialist expertise, be it in a specific specialty, or to provide additional resources during a major project.
Communications consultants typically come in three varieties:single-person companies and contractors;specialist companies whose focus is communications; andconsulting offshoots of larger companies whose bread-and-butter business can be anything from accounting to executive recruitment.
Individuals are potentially the least expensive, but are at risk of getting run over by the proverbial bus. Specialist companies will have a greater range of experience, but check the résumés of the staff who will actually be assigned to the project.
Then there are the accounting firms. The firm's partners may already have a good understanding of the client organisation, but the firm's staff may have been selected on criteria such as which school they attended or in which suburb they live rather than those more relevant to the field in which they are working. Accounting firms are often selected by non-IT management without consulting the managers of the areas in which they will consult, which is not conducive to a good working relationship between the two groups.
A true consultant sells only expertise, so don't be fooled by the sales engineers who prefer the term "consultant". A vendor cannot provide unbiased consulting. Using a vendor for consulting is like going to a car dealer and asking for a recommendation of which brand of car to buy.
Consultant selection must begin with the preparation of a scope of works, stipulating what the consultant is expected to provide and where the boundaries are. A badly written or absent scope can lead to a mismatch of expectations. And then there are those consulting firms that, once they have a foot in the door, attempt to take over the client's entire IT infrastructure. An open-ended scope can provide the opportunity.
An important but often overlooked aspect of consultant selection is that of historical bias. One well-known telecommunications consulting firm almost always recommends a certain brand of PABX, no matter who the client or what the application. Checking a consultant's track record for possible biases before engagement is essential.
Selection of a consultant may involve interviews or a brief tender, but it must be done fairly quickly. It is not uncommon for organisations to assess several consultants, spend months "thinking about it", then expect the chosen consultant to start at a moment's notice.
Selecting a consultant is, though, only the first stage of the process. Contrary to popular belief, a consultant cannot develop an understanding of the client's current environment and business objectives by telepathy. The client must allocate enough time to brief the consultant and must ensure that key people within the organisation also make themselves available as required for discussions. When a consultancy involves a few phases, each with its own report - as many do - ensure you allocate the time to read and review these reports. Do so promptly as acceptance of the phase one report is likely to be necessary for the consultant to begin the work of phase two.
Consultants are engaged, in essence, to answer very complex and detailed questions. While it is essential that a client properly define the question and provide all the information the consultant requires, it is not uncommon for a client to also "suggest" the answer, often on the basis of a casual conversation with a supposed expert, a newspaper article or unsubstantiated perceptions.
Preconceived ideas are not, of course, the sole prerogative of clients. In addition to biases toward products and suppliers, consultants are often biased towards strategies and supposed trends. Be wary of consultants who use terms like "the trend is . . . " and "the industry is moving towards . . . ". Following a supposed industry trend is rarely a sound basis for choosing a system or architecture that is significantly more expensive than that which best meets the business objectives.
Also be wary of the "all organisations need all technology" consultants and those that pepper their presentations with unnecessary jargon that seeks to dazzle rather than inform. Also be wary of consultants who purport to offer project management. Project management is, of course, a necessary component of complex projects, such as building construction. However, when a consultant seeks a substantial fee to manage the installation of a computer or telephone system - work that the respective suppliers are quite capable of carrying out without this additional party -- exactly what additional benefit the consultant will contribute must be carefully considered.
Finally, there is the question of fees. Most consultancies are paid either a fixed fee or daily rate. A fixed fee has the advantage that the cost is known up front, but will include a contingency. A daily rate, however, has the risk of being open-ended. One to avoid is the fee that is a percentage of total cost. While fine for architects, agreeing to a fee structure which is a percentage of total project cost provides an incentive for the consultant to maximise the project cost.
In summary, consultants are a valuable and necessary component of the business landscape. Don't lose sight of the fact that as the customer, you can stipulate on just what the consultant is to advise, but it is up to you to ensure that the consultant is sufficiently qualified and unbiased. However, also don't lose sight of the fact that the consultant is the expert - you are paying for that expertise, so make good use of it.
- S Coates
Gluing It Together
Screen pop, Internet integration, connecting a caller to the agent that caller spoke with when they last called. They all sound fantastic and are just some of the applications of CTI, or computer telephony integration.
CTI can do all this and more, but for all the pizzaz, it must not be forgotten that CTI is an integration of the computing and telephony applications, not the boxes; and that it is an enabling technology, not an end it itself. CTI should only ever be implemented if one or more of the applications to which it is to be put will achieve a demonstrable and measurable business benefit.
Moving from why to how, there are two basic CTI architectures: first party and third party. First-party CTI uses an interface between a single PC and a single phone line or extension to support functions on that one PC. While every home PC uses a first-party CTI application to access the Internet and send faxes, such applications have only limited applications in business. The real action uses third-party CTI.
Using an interface between an application usually resident on a server - but sometimes on a host and a PABX - third-party CTI interfaces the application used by many users and the telephony functionality of the PABX they are all using. Third-party CTI uses an agreed protocol, with the international standard CSTA (Computer Supported Telephony Application) being the most prevalent, although de facto standards TSAPI (Telephony Server Application Programming Interface) and TAPI (Telephony Application Program Interface) and some proprietary protocols are also used.
Another important distinction is that between what are referred to as CTI applications, and other systems that use CTI. Although a better classification would be "call centre/CTI applications", systems generally referred to as CTI applications performs such functions as screen pop and screen transfer with call transfer, and so on. Some CRM systems and other systems also offer this functionality and can correctly describe themselves as CTI-enabled; and some other systems such as predictive diallers and audio call recording systems may use a CTI link. However, avoid the vendors of IVR (interactive voice response) systems who use the term CTI for its perceived cache.
Although you'd never know it from some vendor presentations, after purchasing a CTI system, the application(s) must be designed, developed, implemented and, ultimately, maintained. Only if you have the budgets (both financial and staff time) for all of these activities, which often exceed the cost of buying the development system software, should such an acquisition be considered. - S Coates