When outsourcing is slow-sourcing

Engineering and IT Insight: You cannot improve what you don’t measure, so at the very least it is important that every outsourced support organization includes a measurable feedback policy for each issue. Also consider these three steps.

By Dennis Brandl August 6, 2012

Most large and many medium-sized companies are outsourcing standard IT tasks to centralized support groups. The tasks include infrastructure elements such as server builds, desktop builds, database configuration and support, firewall configuration and support, user account management, router installation and configuration, anti-virus support, and password security services. Outsourced support groups are also used for application-specific support such as email, messaging, office tools, R&D tools, engineering tools, and ERP support. Even medium-sized companies may end up with six or more independent outsourced support organizations.

The upside of outsourced support is the 24×7 availability of expert support. With outsourced support there are experts who have the knowledge and the access rights to quickly diagnose and solve your problems. Other advantages of centralized outsourcing are the consistency of server and desktop builds and the consistency of application tool installation. With a well-run centralized support organization the majority of tasks, such as adding application access through a firewall, creating a standard desktop system, or adding a database account, can be handled in minutes.

Unfortunately, there are also downsides to centralized support, especially when there are multiple independent outsourcing organizations. When problems occur that are not typical or not previously seen, then finding the right support organization and the right people in the organization can be a “Voyage of Discovery” rather than one phone call. For example, an application interface problem that requires a configuration of a firewall port change and  would take less than 30 minutes of actual work, might take weeks of discovery, negotiation, and escalation.

Outsourced support organizations also work problems according to a priority system. In most manufacturing companies, production systems have the highest priority, business critical systems next, and R&D support usually is one of the lowest priorities. This can be a problem when installing and building not-yet-in-production systems. When completed, these systems will be production systems, but with a low priority, any delayed requests for support can add weeks of lost project time and slipped schedules.

Companies with multiple outsourced support organizations usually operate them in a globally choreographed manner. Each support organization knows the tasks that it must perform and does them without requiring centralized coordination. They are like dancers on a stage, each doing his or her own job so that the overall work appears seamless and natural. The choreographed approach works fine until someone trips, then there is a period of chaos until everyone gets back into their predefined roles. Companies using the choreographed outsourced model will not handle out-of-normal problems very well.

You have the problem of outsourcing turning to slow-sourcing when tasks which used to take minutes now take days or weeks, or when you need to discover who you need to talk to when you encounter a problem that doesn’t fit into any of the specific support organizations’ charter. If your outsourcing is now slow-sourcing, then there are several steps you can take.

First, try to minimize the number of independent support organizations. Unfortunately, this is often a CIO-level decision and is hard to justify only for manufacturing system support.

A second alternative is to define a master support organization to take over whenever a problem is unresolved after a specific period of time and give the master organization the rights to use experts from any support organization to resolve the problem. This may be an organization in IT that is focused just on production and manufacturing support, so that it may not require a CIO-level decision.

A third alternative is to create an ombudsman position within the company’s IT organization that can escalate support problems and call in resources to get them fixed. Ombudsman positions are useful if the ombudsman has sufficient authority to call in resources, but only if the ombudsman calls them in as a last resort. The worst solution is to create an ombudsman position that focuses on explaining away support deficiencies rather than fixing them.

An application interface problem that requires a configuration of a firewall port change (less than 30 minutes of actual work) might take weeks of discovery, negotiation, and escalation.

As control engineers, we know that you cannot improve what you don’t measure, so at the very least it is important that every outsourced support organization includes a measurable feedback policy for each issue. The feedback policy should include a measure of the lost time and impact of the problem, including the time required to find the right support organization. If only the actual “fix” time is measured, as is typical in support organizations, then the real cost of the problem is hidden. For example, when you have a team of ten people waiting because the database support team has not responded to a request for a database account, then the lost time must include all 10 people.

Outsourcing of IT support tasks can be a big benefit to a company and provide better service. But if your outsourcing is not meeting your support needs and is slowing down complex support, then address the problem early before it becomes a major project problem, and all of your deployment schedules start to slip.

– Dennis Brandl is president of BR&L Consulting in Cary, N.C., www.brlconsulting.com. His firm focuses on manufacturing IT. Contact him at dbrandl@brlconsulting.com. Edited by Mark T. Hoske, content manager CFE Media, Control Engineering.