Check that two runs produced the same Excel reports

Set up the two runs

Here's the question that got Workbook Compare Pro started. You change something in your statistics software, then run the same survey again. Did any of the numbers change? Let's try it. We've got three Excel reports from the version we trust in a folder called Baseline. The updated software wrote its copies to a separate folder called Rerun.

Compare two files

If you're following along, download the practice files below the video and unzip them. Open Workbook Compare Pro and select Two files. On the left, click Choose baseline file and open the Response profile workbook in Baseline. On the right, click Choose new file and open the Response profile in Rerun. Same name, two different folders.

Leave the comparison options at their defaults for this example. Click Compare. Under Results, the Response profile row says Same. Look one column over. Zero differences. If someone only asked about the response profile, you've got your answer. But we also sent out a set of banner tables and a report on base counts. Let's check those, too.

Compare two folders

Go back to the top and select Two folders. Click Choose baseline folder and select Baseline itself. Then click Choose new folder and select Rerun. The app pairs reports by filename and their location inside each folder. Since our exports kept the same names, all three pairs line up. Run the comparison.

Now the summary says two Same and one Different. In the list, Response profile and Base counts both match. Banner tables has one difference. On a bigger job, click the Different filter just above the list to see which reports need a look. Here that leaves a single row. Select Banner tables, then click View side by side.

Read the changed cell

Both panes are showing the same address, D eight. Look at the weekly shoppers count for the West region. In the baseline it's forty-six. In the new output it's sixty-four. Looks like the four and the six traded seats. That's eighteen more people in a report we expected to match. Now we know which count to check.

Check the next export

I'd go back to the survey data and check who qualified as a weekly shopper in the West. Once you've worked out what happened, export a fresh set of reports to a third folder. Keep the original baseline on the left, choose that new folder on the right, and compare again. Now you can check the fix across the whole delivery.