A Data Science Central Community
Good Morning and Welcome to this edition of the Morning Analytic Coffee Blog.
Today, we talk about the understanding of project parameters, and being eager to please. One of the best and worst things for the analyst is project parameters: Written clearly, with a good understanding, project parameters can really help provide the structure for the results needed. Notice that I didn’t say “results desired or wanted”. Needed results are not necessarily favorable and they aren’t always what the decision makers want. The analyst is not paid to be popular, he or she is paid to provide “disinterested advice”. Poorly understood, or with achievements that cannot be done reasonably in the time-geography context, lead to poor results.
So, here’s the conundrum for the analyst, or the analyst-as-project-manager: If the staff is inadequate or the time for completion unrealistic, you have two choices: push through and do the best you can, reporting on mistakes or errors after the fact, or you can wait until resources are properly marshaled for the event.
My tendency is to do first, apologize later. It can be a very bad one, at that. I’m not patient, and I don’t always wait well. However, poor resources and misunderstandings from the employer or requester of the data can make for a troubled relationship. It is vital that if time is the ultimate arbiter of success, realistic expectations must be established. It is the analyst’s first duty to tell the requester of the analysis to say that limited time makes for less than accurate results.
Another issue is lack of staff appropriate to a task. Staff have 8-9 working hours a day, then they go home. Gathering a panel of broad-based time data may require so many hours. It may require a manager and sub-managers. If you can’t cover the time, gather more staff, or extend the project. Establish project parameters based on a 24 hour clock, and make sure that accounting for travel time and commutation is included. Each person should have a role, each role should have a schedule, and each schedule should be realistic. Write it out first. If your schedule doesn’t look doable, it probably isn’t.
The reverse problem can come from having too much time: an open-ended, no results project with few deadlines. Planning agencies are filled with studies that have no concrete results and no impacts. They are suggestive of “if this” or “if that” or “when funding is available”. A good analyst will tell the requester to ask for specific results, and just as importantly, interim deliverables. Say the best result is to build a new bridge, but there’s no funding for a new bridge. Are there alternatives? What about impacts of different alternatives, or changes in modes? If no money ever comes for the new bridge, what services or business might have to relocate?
Parameters, both of time and staff, matter to the analyst. A good analyst stresses the possible flaws, and checks back with those who need the data if something is seriously wrong. Personal experience says pushing through without re-setting the project, if necessary, can result in serious deficiencies. Being too eager to please, even if you think you can, is not a good thing. I have to remind myself of this. I might leave this blog post up, to remind me.
In any event, the analyst is bound by a set of parameters. Don’t just blame the parameters or the data requester if they’re unrealistic. Drop back, build a plan to make it right, and then present the plan. Be the solution, or at least, propose it. Not all projects succeed, but planning against failure can give you a reliable post-mortem where you can track the mistakes and make them right.