Step 1: Scan, skim, and color-code each column for highs and lows The end goal of the spreadsheet analysis is to distill your learnings into a form, brief, or presentation that explains the rationale behind your recommendations. To do that, you need to do what is called systems thinking to organize information and processes. (See the interview with Milana Sobol in Chapter 10 for more on systems thinking.) So, let’s begin doing that by looking at the heap of raw data in your spreadsheet. Scanning and skimming the data After inputting all the data into the spreadsheet, it’s probably a good idea to reacquaint yourself with all rows (competitors) and columns (attributes) before analyzing them. To do that, I use two speed-reading techniques: skimming and scanning. Skimming means moving your eyes rapidly over text to get the basic meaning of it. Scanning is rapidly covering a lot of material while searching for a specific thing. I skim and scan a lot during a data analysis but not to be sloppy or to cut corners. Instead, I want to quickly discern how simple or complicated the task at hand is. Is my spreadsheet 5 rows by 5 columns or 12 rows by 24 columns with lots of missing data? Estimate the density and completeness of the content you are about to analyze so that you’ll know how long it will take. This matters because you probably have a fixed amount of time to do this task, and you don’t want to waste precious project hours going down a rabbit hole on just one row of analysis. For instance, if you have 20 competitors to analyze and 20 hours in which to complete this task, you have 1 hour to analyze each competitor. Time blocking for both research and analysis is essential because you want to have a balanced perspective — no blind spots. Also notice if something looks incomplete or missing. Did you or whoever it was who did the research overlook an obvious competitor that really needs to be considered? Is the column with the monthly traffic or apps downloaded blank? That attribute could be quite important to know, and it’s a colossal distraction to have to stop an analysis to switch gears back into research mode. Measuring raw data points A data point is a discrete unit of information. Any single fact or observation is a data point. In our analysis, data points can help us illuminate whether something is a failure or success. There are two kinds of data points to keep an eye out for in our columns: quantitative data and qualitative data. Quantitative data is numbers and statistics. How much traffic did a site get? How many transactions happened on the site? How many SKUs are on the site? Numbers can be metrics, transactions, and/or a finite set of options. Unlike