At some point, it became difficult to discuss marketing without mentioning data. Even beyond performance marketing, which tracks results daily, there is a growing number of attempts to measure performance in fields such as content marketing, influencer marketing, and offline marketing—areas previously considered difficult to quantify. Consequently, marketers are also finding it increasingly necessary to prepare reports that include data (performance reports, proposals, etc.).
According to recent research from UC Berkeley , people tend to accept visualized data as objective truth . While we are aware of the powerful impact of data visualization, many of you are likely still confused about whether your current approach is correct, aren't you?
For B2B marketers who are struggling, Elifunt has prepared four essential elements that marketers must know when visualizing data, answers to frequently asked questions, and tips for reporting.
1. Data Visualization: Let's at least follow these rules…!
Does data visualization seem difficult? Just following these steps will get you the basics down! In fact, these are such fundamental things that we probably all know them. However, let's first look at four essential elements that are easy to forget once you actually start drawing charts.
4 Essential Elements for Drawing Charts
① Title that summarizes the chart content
First of all, there is the title. It is surprisingly easy to omit the title when drawing a chart. You might think that everyone understands the content even without a title. However, without a title, the person viewing the data gets confused about what it is trying to convey.
② Axis indicators that explain the data contained in the chart
Next is the axes. We need to indicate the horizontal and vertical axes so that we can identify what data the chart we are drawing contains.
③ A legend that distinguishes various types of data
If you want to display two or more items in a graph, you must also include a legend. A legend explains which item is represented by a specific color or symbol. Without this legend, interpreting the graph may become impossible.
④ Unit display that accurately indicates data values
Finally, unit indications are essential. Charts tell a lot through numbers, but the numbers we intend to convey are not always the same. Sometimes it is time, sometimes the number of people, and sometimes a ratio. Without unit indications, the information you wish to convey could be distorted.
2. Data Visualization FAQs
Now that we have looked at the basics of data visualization, let's delve into more detail by answering frequently asked questions.
Q. Can't you show everything with just one chart?
If you try to show too much data, you might fail to properly convey the truly important message.
This is a common mistake we often make when performing data visualization. It is easy to feel the urge to cram all our data into a single chart. While having everything in one chart makes comparisons easier and allows you to view the data at a glance without needing to create multiple charts, trying to display too much data in a single chart often results in a situation where you end up with nothing to show.
To prevent this problem, the clear purpose of what you intend to showchart. For example, you might want to show the trend of increasing organic traffic or allow for a quick comparison of traffic differences by channel.
Q. Bar graphs, line graphs, pie charts… When should I use which graph?
There is a good chart to display for each piece of data.
Even if you have clearly defined your purpose, it becomes difficult to convey your message if you fail to select the appropriate chart type. When writing reports, it is easy to unconsciously choose charts that are aesthetically pleasing or familiar. Did you know that there is a chart suitable for each data type?
① Line graph when showing changes over time series
Line graphs are suitable for expressing flow because the line extends from the beginning to the end of the chart. Conversely, if you use a line graph to display different item values, it is easy to create the illusion of a continuous flow.
② When showing proportions, pie chart
Pie charts are suitable when you want to emphasize the proportion of a value within the whole rather than the individual values themselves. If you want to display each value and compare them accurately, it is better to choose other forms like bar graphs instead of pie charts.
③ When showing category structure, bar graph
If your data consists of multiple categories and you want to display that composition together, try using a stacked bar graph.
④ Scatter plots are used to show the correlation between variables
Scatter plots are suitable for checking the correlation between multiple variables. Adding a trend line allows you to intuitively visualize the relationship between two variables.
⑤ 3D charts when showing the relationship between three variables
In other words, this means that if you intend to display two variables instead of three, you should not use 3D charts. As the name suggests, a 3D chart represents three dimensions, meaning there are three variables. If you leave one dimension empty and set it to an arbitrary value, it can confuse the viewer.
💡 The simplest is the best visualization!
Using fancy graphs isn't everything
There are many types of graphs, such as violet plots, box plots, and box-and-whisker plots, that are visually and audibly appealing and can display a large amount of information at once. However, if the reader of the report does not understand the graph, it is often worse than not using it at all. This is because the explanation of the graph could end up being longer than the content of the report itself.
Don't forget that the purpose of data visualization is to convey data more easily and with more impact!
Tables are also an excellent method of data visualization
Sometimes, a simple table is better than 100 charts. For example, this is the case when you have a lot of data to display or when you want to compare multiple figures between items at once.
In particular, if you can see the relative size of specific items at a glance using features like Google Sheets' 'Conditional Formatting' or the Sparkline function, there are many times when you won't envy ordinary charts.
Q. How should colors be chosen in data visualization?
Since color is also an important element in representing data, it must be approached according to principles.
Color is also one of the areas where mistakes are easy to make in data visualization. As you draw charts, the urge to choose colors that look pretty to you often creeps up. Even those who stayed focused until the chart type often crumble completely when it comes to color.
To avoid common mistakes, remembering just the following three points is enough!
① Select colors consistently
If you have selected blue as the main color, please maintain it as the default color throughout the report. Changing the color instinctively gives the impression that something has changed. If there are multiple charts comparing category data, please consistently represent the same category data in the same color. Once you recognize that a particular category is red, you can easily distinguish it from then on without having to look closely.
② Use gradient effects only where absolutely necessary
It is comforting to look at a graph beautifully unfolded with a gradient of similar colors, rather than one painted haphazardly with various hues. However, the situations in which gradients can be used are very limited. For example, this applies when there is an order or hierarchy among data values, such as 0, 1, 2, or 3. When there is no order among the values, using a gradient actually makes it easier to misinterpret the data. You have also likely seen rainbow gradient (!) graphs; you really need to be careful with these. Nothing in a chart should be used carelessly. Since every color carries meaning, they can easily lead to misunderstandings.
Representing unrelated items with a gradient, as in the graph on the left, can cause confusion.
Values that are unrelated to each other, such as 'Traffic Channels,' should be represented in different colors, as in the graph on the right.
③ Avoid counter-intuitive color choices
Finally, it is best to avoid choosing colors that go against intuition. Colors that people intuitively categorize include positive/neutral/negative and feminine/masculine. Even if you provide an explanation, viewers may still be confused, so it is advisable to refrain from such classifications unless there is a specific reason not to.
Q. What should I do if I draw a chart but the growth trend isn't clearly visible?
You must approach this correctly, as blindly cropping or scaling the graph can distort the data.
It is a truly agonizing moment when you have finished drawing the graph, but the message isn't being emphasized as much as you'd like. A common mistake in this situation is adjusting the vertical axis to distort the values. You must never touch the axis scale, even if the values are hard to see. It can cause a significant misconception for the viewers.
I recently checked the subscriber trends of the YouTube channel run by Elifunt. I noticed that the number of subscribers had increased by more than three times. Just as I was rejoicing, thinking, "Wow...!", I looked at the vertical axis next to it and felt like I had been hit on the head. It turned out that the vertical axis scale doesn't start from 0. You must be extremely careful when adjusting the axis scale in this way, as it can cause confusion for viewers.
① To clarify that only a portion of the vertical axis is displayed
If you really want to emphasize something, you can design it by drawing a graph starting from 0 and adding highlights.
② Applying a logarithmic scale to the vertical axis
A more mathematical approach involves applying a logarithmic scale. This is frequently used to represent data that initially rises slightly but increases explosively over time.
3. Tips for Reporting for High-Performing Marketers
I think we've covered all the basics of data visualization. However, there are other things to consider in data reports besides visualization. Finally, I'd like to talk about some tips to add that extra touch to your reports.
Every number needs an explanation
When writing a report, simply presenting numbers is not enough. Every number requires an explanation. There must be an interpretation of whether the number is good or bad, and if good, just how good it is.
If you find it difficult to add your own interpretation of the data, try starting by comparing it to last week, last month, or last year. You don't need to go as far as comparing yourself to the industry average; if you are moving forward day by day by comparing yourself to yesterday, you are doing well enough.
Let's write with the reader in mind
It is important to write the content by understanding who the reader is and their level of comprehension.
If you are reporting KPIs to an internal marketing team leader, you can probably omit some explanations regarding the metrics. However, for an external proposal, you will need to provide additional detailed explanations on how the metrics were defined, where they were sourced, the timeframe used, and the criteria used for calculation.
A conclusion is needed
Just because it's a data report doesn't mean it's any different from a regular report. If you simply hand over the current data status, the team leader will probably ask again, "So, what are you trying to say?"
If you have analyzed the current situation using data, it must include the lessons learned from it and suggestions for what to do going forward.
Get more insights for B2B marketers with Elifunt Newsletter!
So, we've talked about data visualization today. I hope this has been of some help to B2B marketers who are struggling to write effective data reports!
If you want to receive more insights for B2B marketers, subscribe to Elifunt newsletter!
