Archive for April, 2009|Monthly archive page

10 worst Microsoft Excel practices

tableI have met a lot of people who know they ‘know’ Excel, even sometimes stating they are experts of some sort. Well, they were not, and I could usually spot it because they make one of the following ‘mistakes’. Of course they get the job done, but in the long run they usually got themselves into trouble with it. I used to be such a person myself, until I became a full time Excel and VBA trainer.

So here it goes, starting with the worst:

1. Using cross tabs

Cross tabs (contingency tables, like e.g. years on the left column and months on the top row) are nice for a visual report, but the data cannot be used in any other way. If you’ve read my recent article about the TableToListConverter, you know why. Excel is an excellent analyzer of data, as long as this data is organized in rows (lists), much like a database.

2. Using cell references

Don’t you hate those formulas like =C1*B6-H3+F5?. They’re horrible to work with. Wouldn’t it be easier to see =Rate*Hours-Discount+ExtraCost?
Well, that’s possible. Just click on the C1 in the top left corner and type the name ‘Rate’, then press Enter. From then on, C1 has an alias of ‘Rate’, and you can use it in formulas.

3. Using too many worksheets

Too many people use too many worksheet. As above in #1, they create Worksheets named ‘jan’, ‘feb’, ‘mar’, ‘apr’ and so on. Later on it will become very difficult to get totals or filter this data. Forget about multiple worksheets and put everything in one sheet, as long as the data belongs together of course. Update: too little worksheets is bad as well, if you think you need to put everything including charts and pivot tables on one sheet, think again.

4. Using color to indicate meta data

Using a color to indicate something is great if you are the only person using the Excel file. That is hardly ever the case. You would have to explain: ‘well, red means this row is to be deleted, yellow means it’s not checked, and green means it is checked’. That’s nice, thank you, but the sheet looks butt ugly, and if I print this on black and white all that information is suddenly lost.

The correct way to indicate this information is to add another column or more columns next to your data. Give them descriptive heading values, like ‘deleted, checked’. Notice yellow and green can simply be combined by putting yes or no in the ‘checked’ column.

And did you know about the conditional formatting feature? Based on a value or even a formula you can color or format a cell. But, make sure to only do that for formatting that doesn’t relate to valuable information you should have visible in your sheet. It’s a good feature though for e.g. zebra-stripes (odd rows colored different from even rows).

5. Copy & Paste

Using copy and paste is fine, as long as you don’t copy when you don’t need to. A simple cell reference will point to value without needing a copy. So if you have a price in a C1 in one worksheet, and need it somewhere else, use =C1, or of course =Rate (see #2 in this list). Another way, which even works for many cells, is go to ‘Edit > Paste Special > Paste Link’.

6. Bad formatting

Excel is good at formatting data automatically, if you insert it in the ‘right’ way. E.g. try typing ‘1-1’ and instead of ‘0’ you will get ‘1-jan’ depending on your date format (calculations need to start with ‘=’). So the value is automatically converted to a date value. If you would type ‘January 1st, 2009’, most likely Excel won’t recognize this and leave it as a text value. You cannot calculate with text values, but you can with Date values. So if A1 has ‘1-jan’, and B1 has =A1+1 then B1 will become ‘2-jan’.

7. Too complex formulas

Of course, some formulas actually are quite complex, there’s no doubt about it. But there’s no need to make them look complex. Split your formulas in multiple parts, and use named ranges. So instead of =Rate * Hours - Disount + ExtraCost you can use =Rate * Hours in the Amount field and =Amount - Discount + ExtraCost in the Total field.

8. Empty rows and columns

Empty rows are inserted to ‘make a sheet look nice’. However, by splitting e.g. January data and February data this way, Excel will assume you are dealing with two lists, not with one. The totals underneath using autosum will therefor work only on one of these sets. If you wish to have a bit more space, just drag the row height handle, or use Format > Row > Height to set a specific height.

9. Formatting for print

If you enlarge a font to make it look bigger on paper, you are making a mistake. Do it using File > Page Setup > Adjust to

You can adjust to any size without changing the font. The same goes for displaying on the screen in fact. Use View > Zoom, or use the percentage dropdown box in the formatting toolbar.

10.  Not using Pivot Tables.

I intentionally name it this way, since there are so many bad ways of analyzing data, getting totals, count, average and more on a set of data. But there’s only one way to do it extremely efficient, powerful and it brings lots more uses with it. Also, it keeps your data separate from your analysis. It’s called a pivot table. Make sure to have a list of data rows with headings, and then choose Data > Pivot Table. Now just click Finish right away instead of following the wizard. You can now drag in column heading names and be presented with a table with totals.

There’s so much you can do with a pivot table I am going to leave it to Microsoft to explain.

A very practical use of Pivot Tables is e.g. to get only unique values for a certain row. Since these are automatically grouped in a pivot table, you can simply copy and paste the list of unique values from there.