In order to do this you must first create your chart in the normal way, with a typical legend. Suppose your information reflects increases or decreases of some KPI, such as profits. Assuming your profit figures are in the column A3 to A10, you can set up a formula such as this…  =IF(A10>A3, “up arrow symbol”,”down arrow symbol”) & “Profits by”&TEXT(A10/A1-1,”0%”). You can repeat this process for each of your data series.

You must now add text boxes at the bottom of your chart where your descriptions will be placed. Click on ‘Insert’, followed by Textbox, after selecting the chart. Make sure that you format the text boxes so that they have the same colors as their corresponding data series’ on the chart. Then you can select the first text box, and in the formula bar you can add the formula that you have already set up to provide the correct legend for this data series. This step should then be repeated with the other text boxes.

]]> http://www.exceldashboard.org/how-to-add-descriptive-titles-to-chart-legends/feed 0 http://www.exceldashboard.org/format-and-group-small-slices-in-pie-charts http://www.exceldashboard.org/format-and-group-small-slices-in-pie-charts#comments Thu, 10 Mar 2011 00:20:22 +0000 dash_editor http://www.exceldashboard.org/?p=436 Although pie charts are quite common and almost universally understood, they are often misused. While they are suitable for data sets with a small number of data points, many people often try to use them for data series with multiple points. This means that they end up with a chart that has so many segments it is almost illegible and generally unintelligible. While it is best to avoid pie charts if your data set is above a certain size, there may be occasions when you feel obliged to use one for certain reasons. In cases such as these, it is best to make use of a pie chart where any small segments are grouped together as one unit.

When creating a chart like this you should take the standard approach of selecting your data and then click on ‘Insert’, followed by ‘Chart’. Then choose a ‘Pie of Pie’ chart. This will give you a large pie chart, with a smaller pie chart providing a ‘blow-up’ view of the smallest segments. Once you have created your chart you can then right click on it and select ‘Format Series’. Choose ‘Series Options’, and then Split Series By Percentage Value. You can then choose which segments will be shown in the secondary pie chart by setting a percentage limit. Everything below this limit will be transferred to the smaller chart.

One of the lesser known options that is available with Excel is the ability to add colored labels to your charts. This is a very simple procedure which produce a nice effect. Firstly you will have to apply data labels to your chart, and then format them. Then you right click on the data cell in order to open up the formatting options. From here you will have to go to the ‘numbers’ tab. This tab is generally used as a means to specify how you wish to have the digits in a particular cell displayed. For example it allows you to choose a date layout, a fraction format or a scientific numbering layout. However it is also possible for you to use this formatting menu to choose colors for your data label.

In order to do this you must choose the ‘custom’ option from the menu and then enter a specific colour code. For example, if you want to colour all of your negative values in red, but all of your positive data labels in green, you should input [green]+0%;[red]-0%. This is a very straightforward and logical code that can easily be adjusted to suit your requirements. Once this has been done you can then see that the data labels of your chart have been color coded.

It is very easy to create a panel chart using Excel. While some tutorials may recommend a complicated approach, involving the creation of a single chart which contains a number of smaller charts within it, this actually requires a lot of formulae and other set-up time. A much simpler solution is to simply generate each of the small charts individually and then arrange them in a panel formation on your worksheet.

When creating your smaller charts the standard approach in Excel is used. Simply choose the type of chart you require and fill in the relevant data. As you are creating a panel chart, each of the other smaller charts will mirror the structure and style of your initial chart. Therefore it is a good idea to create one chart, format it as required, and then copy it to create the additional charts. A quick way of copying a chart is to click on it and then press CTRL + D. Of course each new chart you generate will need new source data, which can easily be adjusted in the usual way.

A very important aspect of panel charts is that they allow for good comparison between different elements of a data set. However this is only effective if the scale of all of the charts are identical. If the axes are set to different limits, visual comparison is automatically affected. Therefore it is essential that all axes be set to the same maximum and minimum limits for all of the charts. Once all your charts have been designed, it is a simple matter to line them up in the layout you require.

In previous posts I have discussed how distracting graphics and crazy colour schemes can make even the most informative dashboard appear cluttered and unintelligible, but that is not to say that some simple formatting techniques cannot be employed. There is actually a very simply technique to softening the angular appearance of these charts, which won’t obstruct the data or make them any less efficient.

If you visualise a very basic dashboard with a clear white background and some charts superimposed on it, you will mostly likely find this image cold, unattractive and spartan. The same applies to a slightly lesser degree to a dashboard with a black background, or indeed any given colour. What you essentially have is a blank page containing charts. However using Excel’s built-in drawing features you can give a dashboard of this kind a much more integrated and organic appearance, with each of the elements fitting together nicely.

What many people do not realise is that they can draw a shape on their dashboard and superimpose a chart on top of it. For example a coloured square can be used to provide a nice background for a single chart, or a combination of shapes can be used to create a simple but elegant backdrop to one a more related charts. Of course the chart must be created first, and its colour scheme chosen so that it is fully legible when placed on its backdrop. Then a shape can be selected from the ‘Insert’ menu in Excel, and formatted as per requirements. Multiple shapes can be added using drag and drop. Once everything is finished the chart can be dragged into place on top of the backing shape.

When creating an animated chart, it is important to include certain features. It should be possible to pause the chart at any point, so as to analyse its evolution at every step. It is also important that you animation run smoothly. A flash chart that develops in fits and bursts and has gaps in its data, will prove detrimental to your dashboard. Make sure that your chart operates smoothly and that it has a polished and professional appearance, otherwise it will look unprofessional and poorly planned.

To many people, the concept of an animated dashboard sounds chaotic, distracting and slightly gimicky. The vast majority of us want functional dashboards that give us the data we need with the minimum amount of hassle, and in the clearest manner possible. When considered in this way, an animated dashboard sounds exactly the opposite of what we might want.

There are a number of caveats when it comes to animation on dashboards. If you create a flash chart, it should be for a valid and functional reason. All too often users are seduced by the fun features of animation, and stick such charts in wherever they can without any thought for function. Animation for animation’s sake is basically a waste of time. A movie showing your chart appearing out of nowhere and building itself in segments is not really informative. Your audience will simply be waiting for the movie to stop so that they can analyse the completed chart. Like annoying unskippable cut scenes in games, for the first one or two occasions it will be a novelty and then it will start to get frustrating.

Unfortunately dashboards are not designed to be entertaining. If you do choose to use animation in your dashboards, make sure that it adds to the impact of your chart, instead of being a gimicky sideshow.

Put simply, your dashboard makes a statement of some kind. It gives an up-to-date assessment of a certain issue. But like any statement you make,  you should be prepared to back it up. This is what your charts and data sets are for. Imagine your overall statement is that your sales have grown overall but that certain market developments have reduced your numbers amongst a certain demographic. There are at least three charts right there.

You will need an overall sales trend chart. After this you will need at least one chart that will demonstrate the various market factors, probably in comparison to the same period of the previous year. If you want to show that sales amongst a certain demographic have decreased you will also be expected to provide a chart illustrating this. A good way to think about your dashboard is that your are making a point, but you will need evidence. This evidence should be clear, indisputable and relevant, and of course is given to you by your carefully selected charts.

Of course we all know that charts are a key element of dashboards. What type you use, what colours, your layout, your choice of data, all of these aspects have been discussed in detail. However the overall concept of the dashboard itself is much more difficult to explain and define. Before you take any steps towards setting up your dashboard you will need to ask yourself what it’s all about.

What is it for, and what are you trying to say? The overall message needs to be clear in your own mind, before you can put it down in graphical form. For example, if you were to summarize your dashboard in words, what would you say? If it is a performance dashboard, you might find yourself giving a general report on how your company or department is doing overall. Once you can put it into words, it is automatically easier to put it into charts.

A gauge chart is capable of providing at most four items of information. Firstly, its needle tells you a particular value. This is typically the current value of some key data set. The colours on the gauge then provide you with three other pieces of information – which values are defined as the danger areas, which are considered to be average, and what value most be exceeded to achieve optimum performance. While these are all valid and important data points, they can easily be displayed using the background colours of a line chart and a single data point.

You can view a gauge chart as a vertical cross section of the very last point of a line chart, where the background has been coloured red, orange and green. The red area is of course the negative region of the chart, the orange is average and the green is good. Clearly, the finally data point of the line chart gives you an indicator of where this particular KPI lies, and also it’s precise value. Crucially however, the remainder of the line chart provides you with additional information that can dramatically affect the interpretation of this data point. If the trend line has dropped suddenly towards this final value this gives a certain impression, as opposed to a trend line that has fluctuated in or around that same final value over a given period of time. Clearly, this data is important for correct analysis of KPIs and to help determine strategies going forward. There are many add-ons available for download which allow you to generate gauge charts in Excel. However it is quite obvious that these charts are largely ineffective and inefficient.