Here's a scenario to beat the mid-winter malaise: Ask your boss for a raise, drawing their attention to the fact that you spent the entire week after Christmas shuffling disks and filesystems. If you think you'll get a pay increase simply by asking, you have an immensely pliable manager or, more likely, you're suffering from delirium brought on by prefetching too many goodies from the Valentine's Day cache.
Whether your manager is weak and says "yes" to everything or a benevolent dictator, it's up to you to maximize your options and performance by managing your manager.
No matter where your manager falls on the temperament scale, you'll find him or her playing the roles of adjudicator, accountant, advisor, confidant, leader, drill sergeant, and friend. The degree to which they succeed in these capacities colors your relationship, and often the degree to which you get along with management. Even the most trying employee-manager relationships can be improved if you make yourself more manageable.
Start by recognizing that you are but one person reporting to your manager, who hears many of the same stories and problems from others, and can easily tire of hearing the litany of requests. How do you stand out in the crowd? Technical exchanges that occur within your sphere are most beneficial when the counterparties are on your ground. Similarly, know your manager's framework in terms of ethics, vision, command skills, integrity, and stylistic nuances. Be prepared to operate in your boss' comfort zone, not your own.
Let's walk through some good employee practices, using an example: You are charged with routing traffic between two Novell LANs in Chicago and New York. The router between the cities is already quite busy and doesn't support IPX/SPX packets. Your choices are to upgrade the router software and buy additional bandwidth between the cities, buy a new router that will handle IPX/SPX and do compression to make better use of the existing line, or convert the Novell LAN to PC/NFS and leave the router alone. PC/NFS and the TCP/IP path look most attractive because you know the protocols and can manage the results more efficiently, but that decision would create political unrest in the Windy City.
Aggregate and abstract. Reduce your issue list to a few common problems. Summarize them in one or two sentences, rather than droning on with relevant background information. This informational phase may be a recap of your task list or an entree to seeking advice or support for resolving a conflict like the one in our example. The best managers apply their strength to remove roadblocks for you. Identify the impediments easily and early so your manager can formulate a plan, too.
Use the right level of detail. If your manager is not technical, don't attempt to explain IP encapsulation to them. Background information is useful to clarify how you reached a decision or to corroborate a position you've taken, but it's of no value to someone who doesn't care about the detail.
Establish your independence in the face of a micro-manager. The manager who makes you the executor of their low-level decisions removes both fun and motivation from the workplace. Escaping the grasp of a full-service manager requires building trust, which we'll come back to later.
You get one shot. Avoid long lists of needs and desires, and don't drone on about the same problem in consecutive meetings. Make your priority clear. If you continually press for half-a-dozen items, each one gets proportionately less attention. Simplify the problem to one single issue, and keep the communication channels simple and direct.
You own your problems. Come to your manager with solutions, not problems. Don't ask your manager to choose a vendor -- instead, have a choice made backed up with your experiences dealing with the vendors, total implementation cost, time to complete the installation or conversion, and user impact.
If you want advice, come with a preference and be prepared to justify your gut feeling. Don't expect your manager to make decisions for you.
Roadblock-moving managers are doing resource allocation in real-time, or are doing global prioritization of their direct reports' task lists. If you can read into the resource-allocation constraints, use this information to shape your solution. When money is tight, for example, go for the lowest cost router solution even if it is not the most elegant from a technical perspective. Be sure to include both direct and indirect costs in your justification; if you want to go with vendor A because you have found it easy to get information from them, the savings in your time may counterbalance a higher price tag. Converting all the Chicago-based PCs to TCP/IP and PC/NFS might cost twice as much as the router, but if it saves you hours of debugging, then the total cost may be the same either way.
An interesting exercise is to ask a budding manager to rank their employees in the order in which they would be terminated. You are seeking, in effect, a matching of business needs to skill sets. The rankings reflect those prioritization criteria the manager holds most dearly: labor cost, tenure, or the value of specialized skills, to name a few. Does your manager sort you on cost savings or on productivity losses? If you think you contribute unique skills to the group, make sure your manager understands the value so you fare well in the next imaginary fiscal crisis.
Black, white, and plaid
Management growth exercises contain an element of black humor. The decisions made by your manager are not always pleasant; spending money on your router problem may preclude a discretionary purchase that will send one of your co-workers nonlinear. Share the pain and fun of making resource-allocation decisions, showing respect for the high-order constraints. The more you participate, the more you gain the trust of your manager. The micro-manager often suffers from a lack of trust in their charges; rather than delegate any duty they'd rather make every decision.
Trust is the key element in most relationships, because it gives you the autonomy to make decisions and gain implicit support for them, rather than fight for explicit consensus on each point. Gain your manager's trust, and the relationship grows exponentially as you see more of your organization's inner workings, which entrusts you to make decisions on a grander scale. How do you get started? Show good judgment when making decisions, show an appreciation for over-constrained problems, and find your own solutions before seeking help.
Gimme some money
Let's focus on money issues, which seem to be the most common, heated manager-employee discussion points. Asking for a raise is in the upper echelon of difficulty, due to a combination of our Puritan ethics that frown on money discussions and the simple fear that you'll be told you're worth less than you think. Don't start by bringing up vague money issues, complaining that you're underpaid and overworked every week. Instead, build an information-rich case and plan for some alternative actions. You need data points and an indication of the relative strength of each item.
Collect salary information; find out what your peers make at similar companies in the same geography. Don't compare yourself to system administrators on Wall Street if you don't expect to have a mouse thrown at your head or be screamed at as millions of dollars pour down the drain during each hour of system downtime. Know what the local demand picture is for your skills and experience. If there's a shortage of Web Masters and distributed-security experts, your market value increases.
In conjunction with the figures, you need to assess your leverage. What are you worth to your company; how much do they not want to lose you? You can truncate any leverage you have by turning a negotiation into a threat. Resist the temptation to threaten unless your "else" clause is so attractive that you're prepared to leave. Many of these ideas were drawn from an outstanding introduction to practical negotiating by mergers and acquisitions master James Freund (Smart Negotiating: How To Make Good Deals In The Real World, Simon & Schuster, 1992).
If you look at system administration as a linear career path, you need to determine how far your experience will carry you on that line. Nontechnical professionals grow their income by growing their business, an option that does not apply to system administrators. Resolving money issues requires assessing your own needs for compensation, quality of life, and value of experience.
What happens when you strike a chord with management but hear that their budgets are frozen and there's simply no more money for salaries? Get creative and tie together the cost models, value-add propositions, and efficiency measurements we've discussed over the past six months. Be prepared for this stock response, and come back with an ironclad proposal to find the money and make the salary slice of the budget pie bigger. See the sidebar Pie-in-the-sky charts for some ideas and examples. Like any other issue you bring to management, predetermine your ideal outcome and resource allocations.
Even if you find the savings elsewhere in the company, there's no guarantee you'll see a raise as a direct result. The indirect benefit is to prove that you can think "big picture" with the less technical folks, marking yourself for a career direction off of the generic sysadmin axis. These pie-slicing exercises are the material of which new CIO and CTOs are grown. Today's generation of chief officers came from single-vendor worlds. The new generation, starting now and continuing over the next decade, will come from the Unix world. Companies that adopt technology most successfully will draw on their own resources to map business problems into the technical offering, identifying upsides and risk exposures. If there's anywhere a corporate ascendancy can be played out today, it's from the electronic mailroom to the technical boardroom. Start the journey today with a few steps in the direction of your manager.
About the author
Hal Stern is an area technology manager with Sun Microsystems. Many of the ideas and insights this month have been gleaned from his manager, Reynold Jabbour. Blame Reynold if you get into trouble, or write Hal at email@example.com.
You can buy Hal Stern's Managing NFS and NIS at Amazon.com Books.
(A list of Hal Stern's Sysadmin columns in SunWorld Online.)
Advanced negotiators move through seemingly impassable ground by expanding the pool of items or money that is being divided between the parties. Instead of slicing the same pie different ways, they get a bigger pie.
Your job: Modify the global budget allocation, taking money out of one pie-chart slice and adding it to your salary budget. If you can save $10,000 on network costs or on chargebacks from another department, that could easily translate into a $5,000 increase in the salary pool. There won't be a dollar-for-dollar transfer of wealth due to costs like benefits and payroll taxes. A $5,000 increment in salaries could easily cost the company $9,000 in expenses, known as the "fully burdened" cost in accounting lingo.
Where do you look? The best long-term cost savings come from an examination of processes and workflows. Is it possible to cache (or replicate) data at a remote site and conserve WAN bandwidth? Take a careful look at how you receive and distribute mail, and what you are picking up off the Internet via WWW and ftp. If everyone is downloading the latest copy of sqHotMetal, then it's time to create an internal archive, index, and Web page. When you pay for your Internet link by connect time, data duplication is wasted money.
The key is to think "out of the box": Rework your current procedures from the ground up without the former gating assumptions about resource costs.
New technologies or productivity tools can increase your costs or greatly increase your efficiency. Radical changes in your data center procedures are the best example of re-engineering -- the emphasis is on engineering first.
If you have problems with this magazine, contact firstname.lastname@example.org
Last updated: 1 February 1995