End-User Services

Taos Professional Services Team

Introduction

Desktop Support is not just for desktops anymore. The realm of the workplace and what needs to be supported is expanding with the proliferation of PDA’s, laptops, thin clients, smart phones, wireless and Remote Access technologies, to name a few. And support demands are stretching to “any place, any time”. We have progressed from “Desktop Support” to “End-user Services,” with mobility being the new focal point.

Many of the EUS related technology projects that we’ve been seeing recently fall into one of these categories:

• Security: The number of security incidents is still increasing every year. Patch management and virus protection are big concerns and driving projects in EUS.
• Centralization: To make IT staff more efficient and reduce costs, many organizations are taking advantage of remote assist tools, group policies, single sign on, etc. to centralize EUS. The traditional model of support is quicklybecoming obsolete as technology evolves in these areas. It is no longer necessary to have full onsite support personnel in offices requiring fractional resources. Nor is it necessary to hire on-site, full-time employees to facilitate the demands of “desk-side” service, when an equal quality service, requiring fewer staff, can be delivered remotely. However, these changes are often more dependent on organizational changes than technology advances.
• Outsourcing: Outsourcing is also a growing trend. Many companies are interested in improving service and focusing on their core competencies. This has been the driver for many of the assessment projects that we have performed.
• Mobility: It is becoming more common for IT organizations to take on support for PDAs and smartphones. Also, changes in wireless and remote access technologies have prompted many new projects related to end-user mobility.
• Microsoft upgrades: As always, a large source of project work in EUS is related to keeping up with Microsoft Windows upgrades. Now is the best opportunity in years to get caught up.

Taos provides end-user services to a wide variety of clients ranging from a multi-billion dollar international manufacturing company to a start-up of less than one hundred end-users. We are in a unique position to see trends across many different IT organizations.

The scope of what is included in end-user services varies from one organization to the next. For the purpose of this article, we will be referring to many of the traditional responsibilities of desktop support, with a particular emphasis on the technologies that allow for mobility and centralized management. We will highlight just a few of the patterns that we have been seeing in end-user services.

EUS Technology Projects

Taos is involved with many projects that are related in some way to EUS. Projects include things such as patch management, ticketing systems, Citrix and terminal services, security, email, Active Directory, and VoIP.

These topics just represent a sampling of the technologies that are impacting EUS. They represent some of the key areas where organizations are spending their money in EUS in order to improve functionality and security – rather than just drive down costs. In addition to the technical trends, we have also been seeing organizational trends that are impacting EUS. Here are a few areas where we are seeing many companies spending their time in EUS.

• Windows XP Service Pack 2: There has been a lot of talk about difficulties upgrading to SP2. Most of the issues are now documented and many companies have successfully completed the upgrade. SP2 has many updates to help centralize management and improve security – such as the new Windows Firewall. However, it is not just a bundle of security patches; it has also added quite a bit of new functionality. Most of these functional changes are related to security, including: network security, memory protection, E-mail handling, web browsing security, and computer maintenance. These functional changes go beyond most Service Packs, which is why there have been more compatibility issues than usual.
  
  Because of these changes, it’s important to do thorough testing of new desktop builds before deployment, especially if you depend on having a somewhat open security model on internal networks. Many upcoming software updates will depend on these functional changes in SP2, therefore upgrading should happen sooner rather than later. So, if you have a policy of avoiding the bleeding edge of new Service Pack releases, the worst is past and now is probably a good time to get started on the migration.
• VPN Remote Access: VPN access still remains one of the biggest end-user support issues. Very often, client driver issues or confusing client configurations are the source of the problems. This is especially a problem when organizations are trying to support older client operating systems. End-user training can help with some of this, but as usual, to really reduce the number of support calls requires controlling the number of client operating systems and configurations.
  
  On the server side, renewed security concerns mean that there are still a number of companies trying to migrate to IPsec or migrate away from broken PPTP implementations. VPN concentrators can act as a firewall and support centralized authentication. Some systems will even allow you to quarantine a remote user until particular patches have been applied to their system. If viruses and worms are a high concern in your environment, you might also review the idea of only allowing virtual desktop access (e.g. Citrix or Terminal Services) into internal networks from remote users.
• PDAs and smartphones: One of the biggest trends in remote access is the addition of PDAs and smartphones to the list of devices that are being managed by IT departments. It used to be the case that if there was a PDA supported by IT, it was a Blackberry. But even more common was the case where employees had varying Palm or Windows PDA devices and each had to figure out their own solutions for syncing with corporate address books and calendars. Now companies are more commonly supporting Palm phones, pocketPC phones, and Blackberries, as well as tablets, laptops, and PCs.
  
  Blackberries require special server-side software and additional licensing, so don’t forget to factor that into the cost of the service. PocketPC phones can connect directly to an Exchange server at no additional cost, but be careful about security issues. Palm phones can sync with Exchange via a cradle or do over-the-air syncs via some service providers. Palm has some solutions on the horizon to help with these issues. Finally, many wireless PDAs can run VPN software and connect directly to your internal network. They can even run Remote Desktop, VNC, or ssh clients. However, because of the small screen size, these tools are generally not that useful for most end-users.
  
  This trend is only going to grow. There are a number of products lined up to hit the market and end-users are coming to expect more support for them. So you may want to review you PDA/phone solutions in the near future.
• Wireless: Even though wireless LANs have become fairly mainstream over the last few years, we have been seeing a number of new wireless projects. Companies are upgrading to new hardware to get new functionality and improved manageability and security. Newer hardware often provides 802.11b/a/g with roaming and easier centralized management for all access points. They also provide improved security with 802.11i (or WPA) with the ability to do centralized authentication through EAP.
  
  The traditional architecture for wireless networks involves isolating wireless clients behind a firewall or DMZ and requiring users to use a client VPN to connect to internal networks. Although there have been several interim solutions to fix the security holes in WEP, it has only been recently, with the wide-spread availability of WPA, that we have seen many companies allowing direct access between wireless networks and internal networks. Of course, the appropriate wireless architecture is dependent on the security requirements for each company.
• Future Technologies: Microsoft’s new OS, codenamed Longhorn, probably won’t be released until 2006 or 2007 (the server and WinFS will definitely launch after 2007). Changes in administration and the end-user experience will be significant; the OS is still based on the same core as Windows XP and 2003. The hardware requirements shouldn’t be much greater than for XP if you can live without the new 3D desktop features. So you don’t need to worry about Longhorn for your hardware purchases today. The big impact that Longhorn will have on EUS is in how you will be able to centrally administer large numbers of what Microsoft calls Smart and Rich Clients.
  

  Other future technologies that will impact EUS include Voice over IP and new wireless standards, including WiMAX and Voice over WiFi. VoIP has been gaining acceptance for awhile now and is starting to become more dependable. We are just now starting to see its adoption as a more mainstream technology. These technologies will also bring more voice functionality to the corporate desktop. This will lead us to the next step in online collaboration – video on the desktop. This is one of those technologies that we keep hearing about, but nothing ever happens, however, now it really is on the verge of becoming a reality.
  
  Another technology that has been getting renewed attention again is the thin client desktop. Sun has put new life into its Sun Ray product and has integrated VoIP and smart card authentication. They can support Microsoft Windows as well as the Java Desktop. Sun has also created other thin clients, such as a thin client laptop and a java phone that all tie into the same desktop infrastructure. There are also interesting hybrids, such as offerings from Clearcube and HP, where thin clients map onto thick clients located in the datacenter. Early indications are that it may be possible to get the worst of both worlds with these technologies, so investigate thoroughly. Historically, thin clients have only reached 2% of the market of desktop sales. Perhaps in these more cost conscious times the thin client will finally have its day. However, with the administrative features of Longhorn, and eventually, the Linux desktop, it may be hard for thin clients to make significant inroads on the desktop market.
  
  Similarly, a 2001 IDC survey found that Linux had a 2% share of the desktop market, while Windows had 94%. Linux has a reputation for being extremely difficult to set up and configure, requiring extensive support. We occasionally see Linux deployed on the desktop, but it is almost always that the end-users are engineers and partially administer their own systems. Primarily due to compatibility issues with Microsoft software, the near-term adoption rate of desktop Linux is unlikely to rise above a few percent. However, there are many efforts underway to change this, and with the backing of large government rollouts in Europe and China, Linux may be able to gain the critical mass it needs to insure that desktop tools move toward open standards. Linux will probably not be viable as a widespread desktop solution for a couple of years, and there are still many hurdles to overcome until then.

Centralizing Support

While technology advances continue to support IT efforts to move to greater centralization management of the EUS environment, the biggest change that has to happen is not about technology. The greatest challenge we’ve seen to our customers’ ability to be the most efficient and effective provider of EUS to their end-user community is in their ability to change the corporate culture. We’ll touch on ideas for how to introduce such change below.

First however, we wish to address a common question which comes up around the ratio of support personnel to systems or users. This is important to touch on as it is such a common question and one companies seem to use as a vital benchmark, and yet in many ways it is really not the right question to ask. The right question is “What is the most effective organizational structure for support given my end-user environment and the technological advancements in support available to me?”

• Desktop Support Ratio: We are often asked by customers about what we’ve seen for ratios of system administrators to number of systems or users. Unfortunately, in the area of desktop administration, we have to respond with the not-very-useful, “It depends.” Gartner did a study of 33 businesses and found the average ratio to be 87 to 1; however the low and high were 12:1 and 275:1, respectively. IDC did a study and found the corporate average to be between 50 and 75 to 1, but found schools averaged 500:1. Some outsource providers and some customers, through the use of thin clients, claim to be able to get the ratio up to thousands of end-users per administrator.
  

  These numbers should all be taken with a grain of salt because there is little effort to normalize what functions are performed by these desktop administrators vs. helpdesk or other administrators. Also, they fail to take into account the corporate culture of how IT is used: what level of support is provided to end-users, how customized their environments need to be, how self-sufficient end-users tend to be.

  When we assess the desktop support for an organization, we may look at a sample of the helpdesk tickets and conduct interviews with some of the desktop administrators. We first try to identify the scope of their job; are they responsible for hardware, PDAs, phones, printers, moves, labs, server-side troubleshooting? We look for IT needs unique to their environment and try to evaluate the corporate culture with respect to IT. We are primarily looking for ways that the environment could be made easier to maintain while still fitting the corporate culture; reducing number of images, locking down environments, automating tasks, etc. We also try to make sure that the current tickets are being serviced at the right support tier. It’s very common to find that more tickets could be resolved without escalation just by adding a little bit of training or automation. A combination of this information is what leads us to our determination of the appropriate ratio of support staff to systems or users for a given environment.
  
  

  Company Competing On:	Ratio
  innovation	25:1 – 50:1
  value	60:1 – 100:1
  little margin	125:1 – 200:1

• Organizational Changes for Centralized Support: Many companies have been moving toward more remote support and centralizing support resources. However, these projects are taking hold slowly. Why? Again, because the challenge is less about the rapid changes in technology and more about making changes within the organization. Organizational change is difficult and in this case requires two major shifts:
  • A fundamental change to the structure of the support organization itself to capitalize on the efficiencies offered by evolving support technologies, and
  • A shift in corporate culture toward greater end-user self-sufficiency and less reliance on the “house call” nature of traditional desk-side support models.

  We have found that a good way to get started is to separate out support that requires “hands-on”. For example, in our end-user services outsource with a leading manufacturing company, we have excluded warranty and non-warranty hardware repair from the support technicians daily responsibilities. Instead, we have dedicated full-time certified hardware technicians to resolve all hardware break / fix issues. Once a support technician determines if a ticket qualifies as a warranty or non-warranty repair, they transfer the ticket over to the hardware technician to resolve. The hardware technician then picks up the device, part, or laptop and brings it in for repair. We maintain a parts depot to limit turn around time.
  
  The result for hardware break/fix is:

  • Quicker and cheaper turnaround than it would be if our client was dealing directly with the hardware vendor.
  • Fewer shipping and logistics costs associated with sending parts to and from the hardware vendor.

  The goal was to improve the quality of hardware support. As a side effect, this set the stage for us to do more support work remotely by eliminating the hardware component from the support technician role. Now, up to 80% of the issues that the support technicians deal with are able to be solved remotely. This in turn allows for the same staff to support a larger number of end-users and this difference more than pays for the on-site hardware support service.
  
  Another step toward increased remote support is to help the end-user become less reliant on set individuals assigned to their department, building, or campus. This can be done through the use of a rotation model. If done well, this has a secondary benefit of being able to reduce the number of support personnel you need overall.
  
  There are several good discovery tools available these days to allow one to understand the movement of systems. These tools allow one to get a sense of how department, building, or campus support may be growing, shrinking or simply likely to hit a peak or valley in support needs. One can use this functionality to achieve a more dynamic allocation of support personnel. Again, the benefit being two-fold:

  • The support organization as a whole can run leaner, and
  • The end-user base becomes more accustomed to seeing a variety of support personnel and less tied to any one individual.

  As you move hands-on support increasingly toward a remote model, don’t lose sight of the value of working directly with the business unit or department in more strategic ways. You’ll want to stay close to them to help them look for pattern issues, changing needs, and bottlenecks which can be addressed by changes in support process or the technology they use. In other words, you need a Business Analyst/Customer Advocacy role close to the business groups you support. There should be a liaison between the support team and the end-users. Not only will this role reduce your support of the proliferation of redundant or ineffective software or tools in the workplace, it will also help to make the end-user community more effective and efficient and will increase your customer satisfaction level. Your end-users will feel heard, and attended to; they’ll quickly come to see that their concerns are consistently and proactively addressed. Again there will be a benefit in helping customers to not feel abandoned in their need for a “desk-side manner.”

• Getting Started on Organizational Change: The suggestions above are only the tip of the iceberg but give you a place to start. The greatest challenge to all of them is readjusting end-user expectations away from the “desk-side manner” of support they’ve become accustomed to. Taos has witnessed struggled attempts to do this time and time again. When a company fails, it is most often because of inadequate marketing to the end-user community, inadequate end-user training, or inadequate testing with end-users; all of which are within the support organization’s control.
  
  IT groups often don’t think to market their delivery model to the end-user. This is a critical mistake. However, the biggest killer of a remote support/end-user self-sufficiency model is when it doesn’t work. For example, if you implement a self-help website, be sure it’s intuitive, easy to navigate, useful, and available. Don’t make it overly cluttered. Start simple – only building more complexity as you prove end-user adoption and capability. Another important element of making remote support work is training your support team to deliver a different type of customer service - phone and email vs. desk-side. Only a small percentage of what we convey is through written or spoken language. If you take away the communication through body language, it becomes crucial for the written and spoken communication to be perfect. Invest in making sure you get this right.
  
  Each of the suggestions above is relatively easy to implement, but require the right planning and preparation to make sure they really are steps forward and not blocks to your ultimate goal.

Summary

In the future of EUS, we expect to see more centralization of services in order to drive up the administrator-to-user ratio while improving stability. Companies are looking for more standardization and are more willing to enforce it through locked down desktop configurations and the use of group policies.

At the same time, organizations are going to begin examining how to implement the many new technologies that are on the horizon. We expect many companies to take this to the next level and outsource their EUS environments – allowing them to focus on their core competencies while someone else stays on top of the technologies.

It seems that desktop support is making a fundamental shift and people are rethinking how end-users do their work. The desktop side of IT is about to get interesting again.