Stacked line charts are a great and yet simple tool. Here is why. We often run into a situation where we need to analyze data with different units of measure. Think about a classic but yet simple situation: Vital company data such as revenue, margin % and expenses is used to obtain insights about the past and current performance . One could dismiss this as an easy task and simply review a standard table. But raw data is really tough to analyze. Detecting trends and patterns quickly is almost impossible. Especially with regular data sets that span multiple organizational units
The other option would be to stick the data into a traditional line chart. But this won’t work in many cases for two obvious reasons:
The units of measure are different (Revenue ($), Margin (%), Headcount (#), Volume (#), etc..)
The units of measure have large differences (example: Revenue is measured in millions, travel cost in thousands)
Both cases result in a pretty much useless chart. You can see a fine example right below:
For data sets containing just two different units of measure, we could alternatively consider a dual axis graph. But I personally find them distracting and many casual users get confused. This is where stacked line charts come in handy.
The power of stacked line charts
Stacked line charts are basically a bunch of line charts that we stack. Why is that useful? Well, take a look:
The stacked line charts allows us to easily identify and compare the trends and patterns in our data. Using this stack is fairly easy. We just have to keep in mind that the units of measure or the scale is different in each one of the line charts. But that should be obvious.
Generating these stacked line charts is really easy with personal analytics tools like Cognos Insight. Spreadsheets typically required us to generate various different charts and to align them manually.
If you haven’t use them before, get started today! Stacked line charts are very powerful, yet easy to use.
The other day I reviewed a dashboard. It looked great. But there was a chart on the bottom that just did not make any sense. It was way too long and stretched out. As a result, it was very difficult to use it appropriately. And that reminded me: We have to watch out for the chart aspect ratio.
The basic idea
Wikipedia defines the aspect ratio as follows: “The aspect ratio of an image describes the proportional relationship between its width and its height.” It’s as simple as that. We get confronted with the aspect ration when we purchase a TV or computer monitor or when we work with photographs. Does the aspect ratio matter? Oh, yeah it does! Take a look at the two photographs below. The first one uses the common HD 16:9 ratio. I cropped the second one down to a square format (1:1). Do you see the difference in the overall impression of the photo?
The aspect ratio does matter for charts as well. We have to watch out for that when we create reports and dashboards or when we perform ad-hoc analysis. Not every chart aspect ratio works equally well. Take a look at the two examples below. Both of these charts have problems:
The first chart is definitely too flat – it is very difficult to analyze it. The second one is probably a bit too dense. The peaks are extremely pronounced and it would be easy to come to wrong conclusions.
A better approach
What is the idea aspect ratio then? Hard to say. It is typically a good idea to use a ratio that is wider than it is tall (2:1 or something like that). But it depends on what you want to show. From my point of view, it makes sense to experiment a little bit. I have noticed that some visualization experts have issues advice but I have found it to be very academic and hard to implement. To stick with the example from above, I did re-size the graph a bit and finally settled on this chart aspect ratio:
Your dashboards & reports
Pay attention to the chart aspect ratio. Only because there is some space left in a dashboard does not mean we can or should stick a certain graph in there. The chart aspect ratio does matter quite a bit as we have just seen in these simple examples. Also, try experimenting with different chart aspect ratios when you perform analysis. Resizing charts with personal analytics tools such as Cognos Insight is really simple.
This has been an extremely busy but exciting week. It seems like the whole world is full of energy. Here are a few things you might want to be aware of.
CFO.com Webinar Forecasting
If you are interested in forecasting, make sure to register for the upcoming CFO.com webinar ‘Forecasting in turbulent times‘. Together with Tom Willman (Principle, The Hackett Group), I will discuss trends and best practices for improving your forecasting processes. The webinar is scheduled for Thursday, March 15th.
Cognos Insight & TM1 10.1 launch
Yesterday was the official launch event for Cognos Insight and TM1 10.1. I was blown away by how many people participated. As a track host, I was especially excited to see so many questions coming through. In case you missed it, you can still watch most of the sessions on demand. I highly recommend the keynote. Robby Meyers from DirecTV gave a fantastic demo of Cognos Insight. Make sure to watch that one. It’s great to see how a successful company like DirecTV leverages Cognos Insight.
There is a great new website and community entirely dedicated to Cognos Insight. Make sure to check it out. The new site provides you with a bunch of great stuff: sample Insight models, tutorials, discussion forums etc.. You can also download a revised version of the famous IBM Cognos Blueprints. Yes, they have been redesigned to work in Cognos Insight. Make sure to also upload your files and share your experiences!
Updated iPad app
There is an updated version of the Cognos iPad app. You can downloaded it directly from the iTunes store. The latest version has a slightly different look and feel. It also feels snappier. There are also a bunch of other enhancements under the hood. And there is also additional demo content in there. The upgrade takes about a minute. And….can you imagine how awesome all your Cognos report will look on the new resolutionary iPad?
Harriet & Christoph – the story continues
Want to see me as a bobble head? Some of you may have watched the Cognos Insight demo at the IBM BA Forum in October 2011. My colleague Harriet Fryman and I demonstrated how the business and IT can get along using Cognos Insight. Our creative team took that story and has created a series of hilarious bobble head movies. The latest edition was released last night. In the prior video, Harriet put Sleep-eeze into my coffee. Time to get even! The other parts are also available on You Tube.
Exciting news! The latest member of the IBM Cognos family of business analytics solutions IBM Cognos Insight is here. This solution will provide business users with analytical freedom while allowing IT to maintain proper control. Some of you have might have already seen a demo. The purpose of this post is to give you a really quick overview of Cognos Insight. Please keep in mind, though, that this post won’t cover all the exciting things you can do with this new solution. Check back for follow-up posts later this month.
Cognos Insight sits on the desktop
Cognos Insight is a desktop tool that allows you to do a lot of things: data exploration, analysis, what-if scenarios, planning, forecasting, dashboarding, prototyping, etc.. You download it and install it on your Windows machine. Having the software on your desktop provides you with the advantage of being able to work in a disconnected and connected mode while leveraging the full power of your machine. Speaking of power and speed – Cognos Insight runs in-memory. The product is based on the highly successful IBM Cognos TM1 engine. When you first open it up, you will see the desktop that invites you to create a new workspace or to open up existing applications.
As a photographer I was really excited to see three traditional alphorn players. The early results looked good on the camera monitor (left photo). At that point I was tempted to pack up and celebrate with my friends. But I resisted and began to experiment with different viewpoints. The final shot ended up as my personal favorite (photograph on the right). Same scene, different perspective. Changing viewpoints paid off.
Visualization of data is one of the hottest topics these days. No matter where I go, people are taking a huge interest in it. Infographics are floating the Internet, for example. Companies are looking to refine their dashboards with better visuals. This was also apparent at the Gartner BI Summit earlier this week.
Yau does a fine job with engaging the reader in the first part of the book. He explains a number of important fundamentals of visualization. This includes a process that he suggests people should follow:
Get your data
Ask a question (what do you want to know about it?)
Choose your visualization tools
Explore the data (look for trends, patterns, differences, etc.)
Tell the story and design the visual
There is a lot of relevant information for business analytics professionals in this section. I particularly like that Yau urges his readers to clearly figure out what story they want to tell by visualizing data. This is often forgotten in the design of a dashboard (e.g. do I use a line-chart to show the trend, or do I use a bar chart to show the variances?)
“Approach visualization as if you were telling a story. What kind of story are you trying to tell? Is it a report, or is it a novel? Do you want to convince people that action is necessary?” Nathan Yau
The other chapters
The remaining chapters of the book contain valuable content as well. The author covers topics such as handling data and picking tools for building charts. Several chapters are dedicated towards describing how to best visualize certain problems (e.g. patterns, proportions, spatial relationships, etc.). Each section provides plenty of examples and some good ideas. I enjoyed working through this. But I do have to say that the content isn’t nearly as deep as let’s say Stephen Few’s material.
A good book for BI professionals?
So far so good. There is just one thing that you should know: Many chapters are also full of technical instructions that teach you how to build graphs and charts in the open source package R along with Adobe Illustrator. There is a lot of code in the book. Technical folks might enjoy this. But it is not my cup of tea and most BI professionals will hopefully build their charts using the corporate BI platform. To be honest, I went ahead and skipped those pages.
Nathan Yau’s book Visualize this! is definitely a good book. I learned a few things here and there and took ample notes. It is also entertaining. However, one has to understand that this is not necessarily a book dedicated towards BI professionals. Rather, this is a book for people who are looking to build infographics and other standalone visualizations. Nevertheless, you can tell that Nathan Yau is passionate about it and he inspired me to hone my skills. If you are looking for a deeper and more business oriented read, I would rather recommend the books by Stephen Few and Edward Tufte.
Part-to-whole analysis is a common task in business. Let’s say we want to analyze how much different product groups contribute towards total revenue. Or we want to analyze our cost across different cost element groups. One way to do this visually is to leverage waterfall or pareto charts. Another popular option is to use stacked bar charts. Stacked bar charts are just a special type of bar chart. Instead of spreading the different categories out across the x or y axis, we stack them. But are they really useful? I have mixed feelings about them.
Single Stacked Bar Charts
Below is an example of a stacked bar chart. This provides an overview of the cost structure for a certain fiscal quarter. You can see that each stack in the chart represents a specific cost element group. The entire stack indicates the total cost.
Is this a good chart? Sort of. Notice how much effort is involved in reading the graph. Comparing the individual stacks requires effort (look at Commissions and Travel, for example). I also find it hard to read the specific values of each stack (how much was spent on advertising?). On the positive side, this chart allows me to quickly identify the total cost. And I obtain a somewhat solid overview of how the money was spent. However, most business analytics platforms like Cognos 10 allow you to hover over a chart section to see the individual values. That makes the stacked bar chart above somewhat more useful.
Another way to display this – and I prefer this option – is to use a regular bar chart. Take a look:
Notice that the comparison of the different cost categories is a lot easier. You can quickly read the individual values and the comparison between the cost elements is easy as well. But I am missing the total of my cost. We would either have to calculate that or include the information in a different manner. The stacked bar chart therefore does not really impress me in this type of setting.
So, should we toss those stacked bar charts then? Not necessarily. Take a look the next example. The analysis is now extended to the entire fiscal year. There are multiple instances of the stacked bar chart.
Notice how different this looks. You can quickly see that total cost have increased in the last quarter (after decreasing slightly). I am also able to see how the different cost elements have changed throughout the fiscal year (look at advertising, for example). The stacked bar chart is now much more useful. I personally like this. But what about the traditional bar chart in this situation? Let’s take a look:
The graph invites you to compare the cost composition quarter by quarter. The comparison between different quarters is also not difficult. The only problem with this version is that the overall cost are difficult to assess. Both versions have their strengths and weaknesses.
Stacked Bar Charts – Summary
Stacked bar charts are certainly not bad. But as the examples above show, they are stronger in a multi-instance setting. But even then, you need to be careful: stacked bar graphs tend to look strange when you have negative values (give it a try!). The single stack is not that strong as compared to the traditional bar chart. Both offer different insights. And let’s not forget about waterfall and pareto charts as well.
From an analysis point of view, I would probably want to switch between the different charts. IBM Cognos 10 provide users with the ability to change chart types on the fly. That makes the analysis of data very interactive.
Have you added stacked bar charts to your toolbox? If yes, make sure to use them in the right circumstances.
P.S.: I will take a look at stacked area charts in February.
Last week I argued that a detailed variance report is not very helpful before and during the forecasting and budgeting process. That post continues to be one of the most popular ones recently. But why not take the basic ideas a few steps forward and create a dedicated forecasting dashboard? A dashboard allows us to view the critical information that we need to get our job done (i.e. create the forecast or the budget) in a single place. Conducting forecast analysis with this dashboard becomes easy and is less time-consuming than analyzing hundreds of variances in a spreadsheet.
A COGNOS 10 DASHBOARD
My colleague Paul took the ideas from the last post and he created an awesome forecasting dashboard in Cognos 10. Take a look (click on the image to enlarge):
This forecasting dashboard is geared towards a revenue forecast. The widget in the upper left corner provides a quick overview of year-to-date product sales. You might notice the use of micro-charts: the sparklines display the sales trend for each region. The accompanying bullet charts show the current status against plan (YTD).
The other widgets provide a balanced mix of historical data (revenue, deal-size, expense ratio) and leading indicators (Win/ Loss Ratio, Customer Satisfaction). But there is also other important forward-looking information. Take a look at the lower left corner: We can view upcoming marketing events along with the anticipated number of participants and the expected sales pipeline. That is helpful for assessing future sales.
EFFECTIVE FORECAST ANALYSIS
This forecasting dashboard can help prepare for the actual forecasting process. It provides a better picture of the business than any detailed variance report can. And think about the time savings as well. The latter requires a lot of effort to be consumed. The dashboard on the other hand is efficient and effective. Last but not least, the dashboard can be utilized on a daily basis.
So, that is a forecasting dashboard built with Cognos 10. I love the look and feel. It is simple, clean and easy to interact with.
P.S.: The type of information to be included in such a dashboard obviously varies by company and industry.
Much has been written about developing better forecasting and budgeting templates or improving the overall process. But to my surprise there is hardly any focus on the role of analysis. I have seen many organizations where managers ‘survive’ the forecasting and budgeting cycle without ever spending time performing meaningful analysis of their data. They simply focus on getting the numbers in to satisfy finance and senior management.
This is a wasted opportunity. People should use that occasion to gain insights about their business. Lack thereof is likely to result in forecasts and budgets that are not meaningful. Some of you might say: ‘Wait a second! Managers do obtain some reports.’ True. They get the classic variance report with a ton of detail. But working with this is time-consuming and it is extremely difficult to identify critical trends and to see the big picture.
BETTER FORECASTING WITH ANALYSIS
Using a Business Analytics platform like IBM Cognos 10, you can make is easier for managers to gain critical insights. Here are a few ideas that you might find useful. Let’s look at the example of a sales manager for a European division of a global company. This manager has to forecast revenue and associated expense.
1. GO VISUAL
First of all, toss those detailed variance reports. Line of Business managers will most likely not obtain any information from them. Human beings do much better processing visual information. You can find a lot of information about this topic on this blog. So, try to swap out those hundreds of data points with a few meaningful charts. Your teams will be thankful.
2. CONSIDER EXTERNAL DATA
The variance report does not really tell us anything about our business potential. We could therefore consider looking at external data such as market trends. More and more of my clients do that. It helps them with assessing their overall position and it also helps them set realistic but ambitious targets. The example below shows that market growth in Europe is a bit limited compared to North America and Asia.
3. STUDY HISTORY
History is not necessarily a predictor of the future. But we should not ignore it. We might be able to identify seasonality and to detect general trends. Pick the critical measures. Line charts are usually a great choice to display this type of data. The example below shows that revenue is cyclical and that the general trend is positive:
4. CHANGE YOUR PERSPECTIVE
One of the nice things about modern Business Analytics tools like Cognos 10 is that we can view data from multiple different angles. Use that capability to your advantage! Try to explore different perspectives. Look at the example above. Now, compare this to the view below. Same data. Just a simple change in Cognos 10:
Our biggest months used to be in summer time. But that has shifted towards year-end. Same data – different perspective. Explore!
5. BENCHMARK YOURSELF
It makes sense to learn from others as well. We could do some internal benchmarking as well. In our example, we could look at deal sizes (looks like Europe’s deals are growing nicely and they are above company average):
Ok. That sounds good. But does the deal size come at a cost? Once again, let’s do some internal benchmarking and look at the ratio of expenses and the associated revenue. It looks like Europe is slightly higher which might explain the higher deal size.
That information is valuable. It also leads us to think further and to ask some critical questions (Does it make sense to review our spending? Does the higher spending lead to bigger deals?). We should obviously not stop right here.
6. LOOK AT LEADING INDICATORS
What about other non-financial data as well? For revenue budgeting, I might also want to look at a leading indicator like customer satisfaction. And I might also want to look at our track record of winning deals (win-loss-ratio). Take a look:
This is a simple example. The manager is now equipped with a few key insights:
Market growth is low
Our revenue trend is still positive
Buying patterns have shifted
Our strategy of investing in selling activities has increased the deal size
Customer satisfaction is increasing which could lead to higher sales
These are valuable insights. And it did not take much time to obtain them. The old variance report would not have provided that insight and it would have consumed a lot of time.
Try to incorporate a few of those ideas in your forecasting and budgeting processes. Doing this with spreadsheets is obviously difficult and probably explains why so many organizations are stuck with the traditional approach. Business Analytics software like IBM Cognos 10 makes it a lot easier to do that.
If you follow this blog you might remember that Mike Duncan from the small business consulting firm Bizzeness posted two guest entries back in August. Mike shared some interesting views about Dashboarding and the selection of proper KPIs. Those were amongst the most popular entries this summer.
Today I was finally able to return the favor. Check out the guest post on the Bizzeness blog. The article is about one of my favorite topics: Data Visualization. While you are on that site, make sure to take a good look at their blog. There is great content!