Tuesday, July 23, 2019

"Little Data"

"Big Data" is the meme du jour, but most projects run on "little data", the sort of data that fits into the constraints of spreadsheets like Excel. It's everyday stuff that drives estimates, scorecards, dashboards, task assignments, and all manner of project analytics.

Let's Excel
So, assuming you using Excel as a spreadsheet for doing actual calculations and data entry, you will find that you need to do some analytics and data analysis from time to time.

"Little Data" spreadsheet tools
Filters are one of the most useful "little data" tools in spreadsheets. Filter is the Excel name for the process and tool, but which the database people -- familiar with SQL for row by column -- would call a query. 

Pivot tables are another spreadsheet data query tool, but they are not the discussion today.

"Comments" -- those text annotations to a cell -- are not analysis tools, but they can be very useful regarding the context of the data. It's amazing how many people ignore adding comments to a spreadsheet.

How to filter
And so, how to do a filter in Excel, something practical for the project manager? There are YouTube clips galore on the subject, but here's a neat, step by step, illustrated process that goes from the simple to the advanced.

Just what the PMO need to get into the data business

Some other rules
Beyond what you will read in the linked article, there are a few data rules that will make life simpler
  • Every column should have a header or title that is unique, even if just column1, column2, etc
  • Only one data value in a cell. Thus, first and last names should be in separate columns; so also city and state. But maybe also captain and ship's names. This is called "normalizing" the data
  • Keep the static data in separate row/column areas from the data that changes. So, if a ship sails to San Francisco on a certain date, the ship's description goes in one area; the city description in another; but the city/date/ship is dynamic and belongs in a third area.
  • Don't put 'word processing' paragraphs or labels in the middle of the data. In other words, maintain the integrity of row by column
  • It's good to have at least one column that is guaranteed to be uniquely valued in each cell, like a row ID
  • If you can avoid using "spaces" in the data, that's good. It makes the query more sure. So, "column1" instead of "column 1"

Buy them at any online book retailer!