Power BI Makeover (Column Chart Edition)

Here on People-Friendly Power BI, I talk a LOT about how Power BI is horrible at visualizing data effectively (and by effectively, I mean in ways that are easy to understand so that decisions can be made, actions take, processes progressed, etc., etc.)

Sure, I’ve talked about how Power BI is completely awful on mobile devices (without a lot of work), how the entire visual system that Microsoft uses to develop is wrong, even why you shouldn’t (and probably don’t have to) use most external Power BI visuals (add-ons made by other developers).

This is all in the cause of helping people make Power BI visuals and reports that are effective and engaging. You know… people-friendly.

So, I definitely have opinions.

While I’ve started posting about better ways of making visuals in Power BI, like in the Essential and Advanced ways to make Bar Charts the best way in Power BI, and also alternatives to Bar Charts, I think it would also be useful if I took “official” visuals shown by Microsoft as examples of what Power BI can do and make them over.

Put your money where your mouth is, Joe! (okay, fine. challenge accepted.)

Why make them over? Because Microsoft’s examples are horrible. Capital H Horrible.

Let’s dive in and have a look at our first example and how it can be worlds better.

Here’s what we’re making over today:

This is a column chart shared by Microsoft in their “Sample Datasets” section of their Power BI site. They contracted with a company (and probably paid a ton of money) who call themselves “pioneers… shaping key features of Power BI…”

Now, you could quibble with me and say “Joe, this is just a SAMPLE”, but if you’ve seen a few Power BI dashboards, you KNOW that everyone (except me and a select few other people) designs charts just like this in Power BI.

How many things can you see wrong with this chart?

Almost everything is wrong with this chart. Let’s circle all the wrong things:

The biggest issue here is that it’s totally unclear what the point of this chart is. What is the message that is important to communicate to viewers?

We need to decipher what the point of creating this chart is. It looks like it’s showing “year over year” (YOY in data-speak) change over 12 month rolling periods across 3 regions.

So, it looks like they’re trying to show TWO things with this chart.

First, change over time (it looks like the change is more pronounced at the beginning and end of the chart, and minimal in the middle.

Second, they are trying to compare 3 different regions against one another.

We can use that “point” to determine how to visualize this data better.

It’s not hard to choose a chart type when you know the point / message / “raison d’etre”… and because there are TWO points to this chart (showing change over time AND comparison), that’s a clue that this data MAY be better communicated using TWO Charts.

So here are some options for telling these two data stories better, and this includes not only choosing the right chart type, but paying attention to your Title, your Legend, and your Axes… all important assistants that help your viewer, when they are built correctly.

Makeover A: Regular Line Chart

Makeover A: Regular Line Chart

Use a regular line chart when you want to show change over a time for just a few categories of data. 3 is about the maximum, as it gets too messy with more. Even these 3 are a bit messy.

Did you spot the color change? Ditch those green/red stoplight colors. Not only are they bad for people who are color blind, but because so many (bad) charts are made with red and green to denote bad and good respectively, a green line could be interpreted as good and a red one as bad for no valid reason whatsoever.

You’ll also notice we added a title that tells the viewer a story about the data. We also got rid of the legend and added Series Labels instead, embedding our legend at the ends of our lines.

We also fixed the horrible X-Axis Microsoft had with months labeled in reverse numerical order (see the “before” version above).

Makeover B: Small Multiple Lines

While the above Line Chart works just fine for showing the overall trends of all 3 lines as a group, it gets a little difficult to quickly figure out what’s happening in each individual line.

That’s when splitting a chart into small multiples is fantastic:

When you need your viewers to quickly see the nuances of each line, split them into Small Multiples. Quick and easy.

Makeover C: Small Multiple Diverging Columns

We can still have a column chart option, but let’s use the Small Multiple option again so the changes in each Region are easy to see AND we’re going to use a diverging color scheme to easily show when the Year-over-Year change is positive or negative. Users don’t have to look over at a Y-Axis to see if a value is positive or negative. The direction of the bars and their color does it for them.

I’ve labelled just the starting and ending columns here, but that’s customizable too, if you want to highlight just the highest or lowest bars, for instance.

3 Makeovers, all 3 way clearer (better) than Microsoft’s

Go scroll up and look at the “before” chart from Microsoft. How long does it take you to figure out what’s happening in the data over time AND in each Region? It takes a while, and honestly, most viewers will spend 10-15 seconds trying to figure it out before thinking “I have better things to do” and clicking away / turning the page.

ALL 3 of our makeovers communicate the message of our data quickly and easily.

The system that keeps Power BI visuals from being great (and how to beat it)

Have you heard the saying “If all you have is a hammer, everything looks like a nail”?

This is actually a cognitive bias concept that is about over-reliance on whatever tool is on hand.

It also reminds me of Peter, Paul and Mary’s “If I Had a Hammer”:

Or maybe Queen is more your speed?

In short, this saying came from a speech given in 1962 at a education research conference by a professor of philosophy, Abraham Kaplan. He was talking about how scientists had (have?) the habit of formulating problems in a way that solving them requires only the skills they already have.

“We tend to formulate our problems in such a way as to make it seem that the solutions to those problems demand precisely what we already happen to have at hand.”

Abraham Kaplan, 1964

Warren Buffet also referenced this saying 40 years ago when complaining about financial data analysis that was kinda useless:

…it’s simply that the data are there and academicians have worked hard to learn the mathematical skills needed to manipulate them. Once these skills are acquired, it seems sinful not to use them, even if the usage has no utility or negative utility….

Warren Buffett, 1984

So both of these old white guys are basically talking about what I’ll (and probably others) call “tool bias”. If someone (or a group of people) know a certain set of things, the tendency is to use only those things to solve problems. Not that they can’t learn new things, only that inertia and momentum plots them along a certain path, and it’s hard to deviate from that path.

A variation of this concept is *everywhere*. If something seems really hard to change (ie. how an org does things, how systemic racism is everywhere, even though most of us know it’s horrible, etc) it’s because things are systemically hardwired that way.

The system tries to keep you in a box.

What’s this have to do with data visualization and dashboards and Power BI?

I’m glad you asked.

If you’ve followed me for a while (and/or subscribed to this newsletter for a while) you know I’m constantly frustrated by the state of data dashboards. They are ugly, user-UNfriendly, and throw too much data at viewers at once. It’s like most of the people who design these things have NO experience or training in User Experience.

Here are some of the horrible examples that came up when I googled “show me great Power BI dashboards”. NONE of these are great, and while this is just a few.. ALL the results were horrible.

Almost every dashboard out there has TOO MUCH data, hitting viewers all at once, put into horrible looking charts that are hard to digest.

Why, for the love of everything good in the universe, WHY?

I’ve been wondering WHY there are such bad dashboard / chart designs out there, and specifically with Power BI (although this problem is not software-specific). Why do default Power BI charts look awful? Why are they designed by Microsoft the way they are? Why are formatting options so awful and limited? Why do so many dashboards CRAM a ton of info into every page and every chart?

Why does making a user-friendly dashboard take a TON of knowledge and time?

It’s because of a system. This system is called the International Business Communication System, or “IBCS” and if you spend any time in the Power BI sphere, you start seeing everyone talking about it. It’s what the people behind Power BI’s charts use when they are adding features to Power BI and making new versions of chart types.

The IBCS is why formatting options are SO limiting in Power BI… because the IBCS says “this is they way charts should be made” and Power BI is developed with that narrow (and incorrect) ruleset in mind. The IBCS is being used by Power BI as their ONLY tool to guide visual development and updates.

That’s a HUGE problem. We can’t format charts as much as we want to to get them as user-friendly as they should be (but, hey, it’s also why I’m known for hacking my way around these limitations) because the ICBS says “charts need to look this ONE way”.

You may have already guessed it, but the “International Business Communication System” is HORRIBLE at the only important concept in their acronym: Communication.

Blame Capitalism

This system was NOT designed so people could easily understand charts. This system was designed so the business community could have some kind of standardization… so everyone would HAVE to look at the same boring, confusing, awful charts.

The IBCS website says things like (and has videos about) “Imagine if every music composer documented their music with *different* notation! Imagine if every architect used their own *individual* system for creating building blueprints. We wouldn’t be able to have musicians play music that wasn’t their own, or we wouldn’t be able to build buildings… we at IBCS bring this same standardized system to data visualization.” (I’m paraphrasing here, but not much.)

Now all this SOUNDS good, but it’s completely missing the point.

A musician writing down their music the same way as other musicians or an architect drawing up a blueprint in a standardized way is NOT the same as data visualization. These analogies apply to capturing and documenting data, so that everyone knows what the data means and how to interpret it, but data visualization is the creative phase of the data life cycle.

Data visualization is when a chart or dashboard designer takes the data and then takes information about the context it’s being used in and applies it to the visual. What does a chart or dashboard audience need from the data? What goals are trying to be achieved by that audience? This is DIFFERENT for every organization, every dataset, every audience.

If we applied the logic of IBCS to music and architecture, every version of a song would sound exactly the same. Every building would be built exactly the same. No creativity. No consideration of context.

Context is EVERYTHING. Everything fails without context.

Blame Capitalism some more

It’s the “Business” in the IBCS acronym that is causing this. The desire to have everything fit into nice little boxes so it can be measured and commodified and made efficient.

While the IBCS documents and templates say things like “keep chart clutter to a minimum” and “label data points” (both good recommendations), in practice and in their visual templates, their chart “examples” are the stuff of nightmares (and Power BI is adding features to charts to pursue these IBCS standards).

Here’s an example that gets “praise” (NOT from me) in Power BI forums and posts:

This is probably the most uncluttered “IBCS” chart I’ve seen, and it’s still freaking horrible.

Here’s a list of issues from this column chart from a 2 second glance:

  • The title sucks
  • The legend is barely a legend. There’s 4 colors in this chart (+ white) and the legend explains 2 of them, with mysterious acronyms.
  • There are 3 data points in each column, so you’ve tripled the mental work your viewers have to do
  • Red and green, for positive and negative, are the only colors really recommended in the IBCS, which is not accessible at all for anyone with red-green color blindness.
  • There’s not explanation of what green and red mean (which, if you HAVE to use red/green, is essential for your colorblind viewers along with directly labelling a color with what it means.
  • The red/green bars don’t even have standardized placement. Sometimes on the left of columns, sometimes on the right?

I WISH I could say this was the worst example, but it’s not. Here are 3 of their “templates” they offer to “guide” data visualization professionals, all from the Templates section of their website:

The IBCS standards advocate for putting MORE data points into visuals (Power BI has been adding features to do this for the last few months) in the interest of “showing more data” and at the same time obviously disregarding every User Experience recommendation out there (Basically, don’t drown your users in data. Give them a sample, and let them choose to see more)

Know thy Enemy

This is what you’re up against if you want to design user-friendly Power BI reports that are easy, digestible, and effective. The system wants NONE of these things. The system wants reports made in a standardized and ugly way.

This is where we beat the system.

Knowing that these IBCS standards (and Power BI by extension) is NOT being designed for people (but for business) is our advantage.

While most Power BI designers are going to lean on the IBCS “standards” as the only tool in their toolbox (see our discussion about hammers above) and say “oh, IBCS says to design it this way…”, those of us who are paying attention know that it’s PEOPLE at businesses (but also hospitals, and school boards, and foundations, and non-profits, etc.) that USE the reports.

If your report is designed to “communicate” to a business, it’s gonna fail and go unused.

If your report is designed to communicate to PEOPLE, those people are gonna love it.

Know the system you’re up against, so you can actively work around it.

Power BI has a Jurassic Park problem.

Years and years ago (like, pre-podcast era) I heard a radio interview with a Russian who had escaped to the West just before the fall of communism. The only part of the interview I remember is that he tried to go grocery shopping once he got settled into his new life in America and couldn’t.

One of the items he went to buy was a jar of jam. In contrast to his old life, where finding jam at the grocery store was sometimes a rare event, he found a huge assortment of jam at his new grocery store. Different flavors. Different sizes. Different qualities.

He was simply overwhelmed by choice. It was TOO MUCH. After decades of having to just use whatever jam (sometimes) would up in a grocery store, he was hit over the head with basically every jam ever.

He ended up being paralyzed by choice, leaving his cart in the jam aisle, and going home, not buying anything as he was too overwhelmed by the experience.

Data dashboards have this same problem, and it hurts them. It paralyzes the people who try to use them to do their jobs.

I’m willing to bet that the last dashboard you looked at was busy. Lots of charts and numbers and filter options. There’s also a good chance it was designed with a lot of different purposes in mind.

That’s the beauty of dashboards, right? You can throw a TON of data into them and show tons of visuals for different people for different purposes.

That’s what (most) Power BI developers do. Power BI can do it, so let’s do it, right?


This is the Jurassic Park problem.

Just because you CAN hit your viewers with a ton of data, you SHOULDN’T.

Just because you may have worked with data for years/decades, it does NOT mean you’re automatically allowed to confuse people with overly complicated charts and dashboards.

Let’s use an real-world example to demonstrate this. This is just ONE example.. there are a billion out there in Power BI / Dashboard-land.

I recently came across a post from a tech dude (honestly there are so many tech dudes that think because they know how to make a complicated graph it makes them a data visualization “expert” everyone should listen to…) who wanted to take a tweet by Elon Musk (who, let’s be frank, is a horrible person) that had a table and make it a “well designed dashboard.”

Here’s the tweet:

It’s a table with a lot of data. Different regions, different metrics, total users vs mobile users. It’s kind of hard to figure out what’s going on here easily.

Here’s a (their words) “well designed dashboard” using this data that I witnessed other tech dudes fall over themselves praising:

So, a couple bar charts, still a TON of numbers… some heat map -ish colors in that grid of numbers (good job Canada re using twitter less!) … it’s still hard to figure out what to pay attention to. What is most important here? Who knows?

This “dashboard” is horribly designed. It’s awful, and completely misses the point.

This is the Jurassic Park problem.

Many (not all, but many) data visualization people have powerful tools at their disposal. They can take a table full of numbers (like Musk’s) and visualize it in a different way. They have the tools, and the data, so their instinct is to puke all that data over a dashboard and firehose it at viewers.

Again, just because you CAN, it doesn’t necessarily mean that you SHOULD.

I can hear you saying “Yeah, but Joe… what would be better than this? Is there a better way to visualize that table”. Hell yeah there is!

Let’s look at the original tweet/table again, but this time let’s look at what the world’s biggest grifter (or is it Tr–p?) is trying to communicate as his main message.

This image has an empty alt attribute; its file name is image.png

Did you catch it? What’s his main takeaway here?

That’s it. The main takeaway / message / insight is that usage is up 3.5%. (well, it’s actually 3.6% with rounding, but that’s beyond Musk, maybe?)

All the rest of the data is extra. It allows people to see usage by region or by total vs. mobile if they want.

Do we need to firehose all this data at people on a dashboard? NO, we don’t.

An actually good well-designed dashboard using this data would ONLY show that takeaway message (it doesn’t even have to be in a table/chart! It could just be a sentence on the screen! It’s ONE number!) and THEN if people wanted to know more about it, then they could interact with the dashboard (it’s what dashboard software is designed to do) to get more information.

Enough talk Travers, right? Let’s give this a try.

First, communicate the key insight. If it’s ONE number, you don’t need a freakin’ chart or table crammed with numbers.

This is all you need:

Do you see how much more user-friendly this is? It communicates the key insight and ONLY the key insight, but it also gives your audience the option of seeing more data.

Clicking that little “click here” link can make a chart appear (on the same page, in another page, wherever) that gives them some more info:

You could show Total usage and also Mobile usage, too.

Or you could add another layer of detail that would show up in a tooltip IF your audience wants it (yeah, we’re talking multiple levels of “details on demand” here)!

It’s SO easy to design dashboards that are actually good and pay attention to what key insights and messages need to be communicated, but the Power BI world and the dashboard world in general (but not everyone!) just crams too much on there, without paying attention to what users actually need.

Next time you get a table of data and asked to visualize it, think about the dinosaurs at Jurassic Park making lives miserable (and short). That’s what busy dashboards and reports do to your audience.

They may look impressive at first glance, just like genetically re-engineered God Lizards, but you’re gonna regret it. God, are you gonna regret it.

4 Alternatives to Bar Charts in Power BI

Don’t you hate how whenever you buy something online, you see algorithm ads for days/weeks/months afterwards for that same product?

It’s like that big AI Skynet hivemind living in Jeff Bezo’s basement, while very smart, isn’t necessary intelligent. You have bought, say, a chandelier, and now the Amazon/Google monster thinks:

“they bought one chandelier, and haven’t returned it, or left a bad review, therefore they must want MORE chandeliers. Bombard them with chandelier ads! Money money capitalism singularity sell to the humans”.

It has LOST the thread of our chandelier story. We don’t need another chandelier once we’ve bought one (usually, I don’t know your lighting needs situation). What we need to buy is maybe special-sized lightbulbs for that chandelier, or we need a guide to figure out how to install it. Or we need some kind of special duster to clean it.

We need a different tool. We bought the tool that holds light sources. Now we may need the light sources, or an installation guide tool, or a cleaning tool.

This is why it’s SO important, when you are building something, to actually know the needs of the person/people that will be using it.

This is why we need charts other than regular ol’ Bar/Column Charts. While easy and useful, sometimes something else is better.

In the last two editions of People-Friendly Power BI, we learned 5 Essential methods for making great Bar Charts in Power BI (things you should do *every* time you make a bar chart) and 5 Advanced methods to make Bar Charts amazing.

We’re ending our Bar Chart series (for now) with 4 Alternatives to Bar Charts. A couple of these are variations of bar/column charts (they still use bars) but have important differences, and a couple are completely different chart types.

Alternative 1: Overlapping Bar Charts

True, we’re still in the Bar Chart part of the data visualization world with an Overlapping Bar Chart, but it serves a whole different purpose than a regular bar chart.

While a regular bar chart allows us to compare categories of something really easily, an overlapping bar chart lets us do this PLUS show how a subset of data relates to the whole.

This chart type is VERY alternative in the Power BI world because it’s NOT a chart type that comes with Power BI. It has to be hacked using Power BI’s formatting options. It’s not hard once you know how. I try not to use external visuals and all their issues.

If you’d like to learn about how to make an Overlapping Bar Chart, get yourself on the VIP List for the Evergreen Data Visualization Academy. I create the Power BI tutorials there and we have one on how to make Overlapping Bars.

Alternative 2: Waterfall Charts

Waterfall charts are a different form of bar/column chart. Instead of each column starting at the 0 point on the y-axis, each bar either rises (or falls) from where the previous bar ends. It’s fantastic way to show how different elements or factor contribute to an increase (or decrease) of a metric.

Waterfall charts come as a “stock” option in Power BI. I customized it by formatting the heck out of it, adding dynamic colored value labels and of course a great descriptive title.

Alternative 3: Lollipops!

Sometimes a regular bar chart IS the right chart for your users and the insight you are trying to communicate from your data, but bar chart fatigue is a real thing.

Change things up with a Lollipop Chart, which is still excellent for comparing categories of something (just like a bar or column chart), but is a little easier on the eyes AND emphasizes the value more visually. A dot that represents the number gives more emphasize to that number.

Lollipops don’t come stock with Power BI, and there are some custom visuals you can add to make them (all with limited formatting), but why go out and get a visual when a Lollipop Chart is totally doable (and easy) to do in Power BI?

Vertical Lollipops (columns) are totally doable and easy. I’ve alllllmost figured out Horizontal Dumbbells (without resorting to adding in an external visual). Almost.

If you’d like to learn about how to make a Lollipop Chart, get yourself on the VIP List for the Evergreen Data Visualization Academy. I create the Power BI tutorials there and we have one on how to make Lollipop Chart.

Alternative 4: Dumbbell Dot Plots

Sometimes, people try and use Bar Charts to show the difference between two metrics across categories, and they always end up with those charts with a ton of bars that are impossible to decipher with a lot of work. It’s a mess. A Mess.

If your message is about the difference, then just show the difference, and emphasize that difference. Dumbbell dot plot charts are great at doing this. They focus the visual on just the data points and the differences between them.

This is another chart type that Power BI doesn’t come with, and while there are external visuals that will make these, they can be made with an easy Power BI hack.

If you’d like to learn about how to make a Dumbbell Dot Plot, get yourself on the VIP List for the Evergreen Data Visualization Academy. I create the Power BI tutorials there and we have one on how to make a Dumbbell Dot Plot.

Now you have 4 options to change things up and visualize bar/column chart data in different ways to tell the stories in the data better. That’s what it’s all about.

A Power BI Holiday Music Dashboard!

Happy holidays!

I’ve been playing with Power BI and Spotify music player embeds the last little bit to see if I could get music embeds to show up in an attractive way (and have the color formatting of everything change for every song.

Here’s what I’ve come up with (and will likely expand and enhance this next year).
I hope you enjoy clicking and listening to different songs.

Holiday Music 2023

See you in 2024!



Five ADVANCED ways to make your Bar Charts amazing in Power BI

We’re going to talk more about bar charts everyone. Contain your excitement.

Okay… okay… let that excitement OUT! We’re talking about BAR CHARTS, after all!

Last time on People-Friendly Power BI, we talked about the five essential ways to make your bar charts the best in Power BI.

Namely, great titles (dynamic ones!), better labels and taking out chart clutter are all relatively easy to do in Power BI if you know where to click, and you WANT to do ALL of these things to make your bar charts easier and better for your users.

In this post though, we’re leveling up a TON. We’re going to take our simple (yet effective) bar charts and add features that not only make our charts EASIER for our viewers and give them MORE data, but makes people (like your boss) sit up and say “I had no idea a Bar Chart could DO that in Power BI!”

First let’s remind ourselves about where we left off in our last Bar Chart post with a visual.

A beautiful bar chart, right? This should be the bare minimum for all your bar charts. Easy labelling, no chart clutter, and a title that tells a story.

Now let’s explore multiple ways to take a bar chart like this to the next level.

1. MORE Dynamic text and calculations

We finished off our last post talking about creating awesome dynamic titles for our bar charts that CHANGE and tell a story as data changes. Well, Dynamic titles can tell simple stories but dynamic text can tell complicated stories. With a bit more work, you can perform ANY calculation in your data, and communicate it effectively in a title (or anywhere else).

Here’s an example. Let’s say that we not only want to highlight what province has the highest percentage of food security, but HOW MUCH HIGHER that percentage is that the value in the province with the LOWEST value?

It’s doable! We can add a subtitle with that extra insight.

So, this chart title is now not only telling MORE of a story to viewers, but it’s doing work for them. It’s figuring out the difference between the highest and lowest values instantly and communicating it FOR your viewers, so they don’t have to!

Let’s say our data changes. No problem!

Remember, what story and insights your viewers need can only be known by TALKING to them. Go talk to them and create dynamic titles that will knock all their socks off.

We can literally put ANYTHING we want in our titles, and have it change as our data changes.

2. Benchmarks!

We can also add in Benchmark lines to easily show what bars in our bar chart are exceeding (or not) a certain value. Managers LOVE benchmark lines as they can instantly tell if that benchmark value is being met.

In this example, we have a benchmark line at the average percentage of food insecurity:

While we can’t format benchmark lines very extensively in Power BI (at least not yet), they can still be useful in providing some context for your viewers.

Conditional Colors

We can also color our bars based on conditions. This bar chart colors the bars of provinces that exceed the average differently that those below the average.

It’s a great easy way to really draw attention to the data you want your viewers to focus on.

4. Hack your labels (and more)

Power BI is designed so that we as designers have to do things in a particular way. It’s not designed for “out-of-the-box” thinking, but that doesn’t meant that we can’t hack Power BI to do things it’s not designed to do.

With a bit of time with the program and creative thinking, you can do all sorts of cool things.

Maybe your boss wants show your n-values or sample sizes right there on the bar chart in addition to the percentages. It’s not a built-in option in Power BI, but that doesn’t mean we can’t hack it:

Or maybe we want to get rid of our Y-axis completely and put our Province names right IN our bars. There’s a hack for that:

How about popping our Y-axis labels OVER our bars? We have the hacking technology! 🙂

Essentially, even though Microsoft puts a ton of limitations on how we can build and format charts, we can still hack our charts into creative variations.

5. MORE Data (but only when needed)

You never want to hit your viewers with too much data. Too much data at once makes charts and dashboards LESS usable, as your viewers will spend a ton of time figuring out (or just looking for) the data and insights that are relevant to them. We want to give them a high level look at the data and then allow them to delve deeper into the details IF they want to.

Give them that choice.

One way to give them that choice is by adding customizing the tooltips that show up when a viewer hovers their mouse over a bar. They have access to the data if they need it, but it’s not splayed all over the dashboard by default.

In this example, the detailed data shows food insecurity levels in each province over time (and you can incorporate dynamic text into these as well).

As amazing and awesome that tooltips can be, one important thing to remember is that screenreaders do not read tooltips, so they are not the most accessible feature in Power BI.

An alternative could be what’s called a Drill Through filter. With a Drill Through filter, someone can click on a bar and get access to more data about that bar on a new custom filtered page that screenreaders CAN read.

In the below examples, some clickable text shows up when a bar (any bar!) is clicked on, allowing your viewers to access more data about that province (or whatever your bars are about).

Now you have 5 different advanced ways to make your bar charts more useful for your viewers in Power BI.

Even though you now have great examples of how to make bar charts awesome (and essential fixes you need to make to the Power BI defaults), you may also way to move away from bar charts sometimes.

Next time on People-Friendly Power BI, we’ll talk about alternatives to bar charts that we can make in Power BI.

Five ESSENTIAL ways to make your Bar Charts the best in Power BI.

Who fell asleep when they saw that this edition of People-Friendly Power BI was about Bar Charts? You can admit it, I won’t think less of you.

Bar Charts are the unsung heroes of the data visualization world. When you ask anyone what their fave type of chart is, “Bar Chart!” is never the answer.

However, we all use them, don’t we? Chances are, if you’ve visualized data for an annual report, a powerpoint, or a dashboard, you’ve most definitely made a bar chart.

This is because they are EASY to understand. Any audience can understand them. The bigger the bar, the higher the number it represents. Humans are *really* good at comparing the length of bars and it doesn’t matter if the human is a theoretical physis Nobel prize winner or a kid looking at a bar chart about halloween candy.

Bar charts, as easy as they are, aren’t perfect when we make one in Power BI.

This edition of People-Friendly Power BI is all about the ESSENTIAL changes you should make to Bar Charts when you make them in Power BI. Power BI makes garbage bar charts by default… you gotta tweak those bar babies into something your Power BI dashboard users are going to love.

Let’s look at the kind of default bar chart Power BI gives us when we add data into the bar chart visual.

This chart is about food insecurity in Canada’s provinces.

It’s recognizable as a bar chart (yay!) but it’s not as good as it could be, so here are 5 essential modifications to make to really make your bar charts shine.

1. A good descriptive title

Tell your users what your bar chart is about. This is the most important part of any chart (yup, even more important that those bars). You want to have a descriptive title that describes what your chart is about and ideally what the main key takeaway is.

This way, even if your chart is hard to understand (but let’s make sure it isn’t!) your viewers will still get the point.

Delete the awful default title Power BI gave the chart and write out a better one that describes your data. Make it bigger and more visible. People need to see it. Always left align your title and keep it at the top. Our brains and eyes in the western world are programmed to start reading content at the upper-left.

2. Add value labels and enhance your Y-axis

By default, Power BI gives us an x-axis with values, and faint gridlines that give our viewers a little bit of help estimating how large each of our bars is.

While this may work for some bar charts, namely those where you just want to show that some categories of data are larger than others (with no one caring about exact values), most of the time, if you’re comparing categories of data (which is what bar and column charts are amazing at showing), viewers will also want to know the exact size of the bars so they can easily figure out the differences between them.

Add value labels, and since these are percentage amounts, format them as percentages. Also increase the font on your Y-axis so those data categories can be seen easily.

3. Take out chart clutter

There are a few things on this chart that don’t add anything to the message we’re trying to communicate. In addition getting rid of our axes titles, we can also remove the x-axis itself. We don’t need the x-axis labels anymore, as our bars are labelled directly.

4. Give your Y-Axis more space (if needed)

Our bar chart is looking pretty darn good, except for our Y-Axis. We can’t see the full province name for Newfoundland and Labrador. Power BI just cuts it off with a “…”

If this is happening with your bar charts, go into the Y-Axis options and look for a Max Area Width slider. You can tell Power BI to devote as much as 50% of the width of your chart to your Y-Axis labels. Then long items like Newfoundland and Labrador will fit (and god, it looks so much better).

5. A Dynamic Title

Our last technique seems advanced, but really isn’t. It’s also something that every chart and dashboard needs but so few have. Dynamic titles are titles that *change* as your data changes (and we’re talking dashboards, so there’s a good chance the data will always be changing).

Dynamic titles are world changing for your viewers. If you’ve done your job and know what your users need to know from your dashboard (always the first step in dashboard creation), you can create really simple DAX measures that let you highlight key insights with titles. Titles that change when data changes.

If Alberta has the highest rate of food insecurity, the title can call that out:

If it’s New Brunswick, the title can say that.

Now we know 5 essential ways to make your Power BI bar charts not only better, but *easier* for your Power BI report users.

Next time, we’ll learn at least 5 *advanced* ways to make bar charts better by taking them to the next level.

Yes, you CAN tell narrative data stories with dashboards!

Hey there!

When I was young(er) I was very petrified of public speaking.

You may have this fear too, and can relate to the overwhelming anxiety that used to hit me before I was due to speak (even just in front of a small school class). The annual public speaking project we had to do every year is something I dreaded. My stomach would feel sick. I’d sweat and have chills at the same time. I felt like I was going to pass out. This was even BEFORE I stood up to speak.

The usual tips of “just pretend the audience isn’t there” or “look at the back of the room” or “find a friend in the audience and speak to them” or even “imagine the audience naked” (hot, but not useful) didn’t work, at all.

You know what worked, and worked better than I could even imagine? A paradigm shift of how I thought about the audience.

I randomly read a tip somewhere (I wish I could remember where) that since public speaking is *such* a common fear, most of an audience is secretly (or not so secretly) in awe of a speaker who can stand up and present.

That ONE TIP shifted my whole mental approach to public speaking.

Instead of thinking about how nervous and anxious *I* was, I realized that I could think about how the *audience* was feeling, and draw strength (emotional and figurative) from them. There was a GREAT chance that every time I stood up in front of a group, a lot of them were thinking “Wow… I wish *I* could do that…”

The larger the group, the better. In larger groups, even more people would be thinking that, and I knowing that gave me confidence. I used that confident energy that THEY were giving me.

I fed off their energy, like an energy vampire:

Yes, that is actually me dressed up last Halloween as “Energy Vampire” Colin Robinson from “What We Do In The Shadows”…

That ONE realization TOTALLY changed the public speaking game for me. It wasn’t overnight, but each time I spoke in front of an audience I got a little more comfortable and sure of myself.

I had been approaching public speaking in a certain way (the way most people do) because that’s just how it was done. I was focusing on how *I* was feeling instead of what my *audience* was feeling.

This type of paradigm shift is what’s needed in the data dashboard world. Most data dashboard professionals approach dashboards in a certain way and with certain restrictions and preconceived notions about how you can (or can not) communicate data.

There are two kinds of dashboards

There are generally two kinds of dashboards, exploratory and narrative (sometimes called explanatory). Some will say there are more (such as operational, strategic, and analytical), but it’s just semantics. The audience may change with these, but not the overall functionality.

Exploratory dashboards are ones where data is explored, and are most often used by data analysts to suss out stories in the data. It’s like panning for gold in a river. You can spend all day sifting that (data) sand in a river bed and only get a couple nuggets of (insightful) gold.

Narrative dashboards are the ones where identified stories about the data are put on display. It’s like taking your nuggets of gold and polishing them up and putting them in a display case for people to ooh and ahh over.

In the data dashboard world, unfortunately, the consensus seems to be that dashboards are only appropriate for exploratory needs. This belief is what has resulted in dashboards with a ton of different metrics all showing up at once with many filters and slicers for users to struggle through.

Like these, which are highlighted on Microsoft’s “What is a Data Dashboard” webpage:

These are prime examples of exploratory dashboards, and let’s be frank, 99% of all dashboards are exploratory. If your dashboard only reveals data stories when your users click around and filter and slice data certain ways, it’s not narrative at all. The narrative is there, but you’re making THEM find that narrative. You are making THEM pan for gold nuggets. They do NOT have time for that.

The solution is narrative dashboards, or dashboards that KNOW what users are looking for and GIVES it to them. Whether that’s one short story or an epic novel with chapters and character development arcs and a 2nd act dilemma and a 3rd act climactic resolution.

The data dashboard world still operates with the assumption that dashboards can NOT tell stories, because it’s impossible to know what ALL your users may want to know AND that dashboard data is dynamic, so it’s impossible to tell a story with it.

Both of these assertions are false and lazy.

Your audience is more important than your data.

If you don’t know your users, that needs to be addressed. The first thing you need to know when designing a good effective dashboard product is the audience. NOT the data.

Let’s repeat that.

The first thing any dashboard developer should know is the AUDIENCE. Everything revolves around knowing what they NEED.

If the data doesn’t support those needs, maybe the dashboard shouldn’t be built at all.

Knowing your audience requires talking to them, or someone who knows them really well. What questions do THEY care about that the data can answer? How well do they know the data being funneled into the dashboard? What do they need to DO with the insights (stories) a dashboard can tell them?

Knowing these things allows us to build a dashboard that answers questions quickly. No hunting for nuggets of gold. No one has time to sit by a river sifting that sand.

If someone can’t open up a dashboard and find the answer they are looking for in a few seconds (or a complicated insight in less than a minute), it’s NOT a good dashboard.


My own personal goal when I build a dashboard is to get a viewer in and out of the tool as quickly as possible. If they are spending time searching for the gold nuggets they need, I’ve failed. They have their regular work to do, and they need quick insights to make a decision or move a process along. They can’t be sitting by a river all day looking for that gold.

Talking to dashboard users is not a normal practice. A quick google or search on Linkedin about what one needs to be a great data analyst or a great dashboard developer almost NEVER includes talking to users. These lists include learning SQL, Python, javascript, M Language, DAX, etc., etc., etc. It’s all data and code. It’s never about people and what they need.

Knowing what the people using your dashboards need from them is the first step. It’s the ONLY first step.

If you’re thinking that a dashboard can get unweildy and overly complicated if there are a lot of user groups that all need different things from a dashboard, you’re totally right.

If you’re trying to make a dashboard that answers the pressing questions of multiple disparate user groups, you are doomed to failure. Trying to please all of these groups will mean making compromises and you’ll end up with a dashboard that maybe everyone can use, but no one will be happy with.

At the very least, if you MUST have a dashboard with different users, create different pages for different groups… but ideally, you’ll want different dashboards for different user groups. They can all use the same data, but it’ll be displayed differently depending on the user group.

This user-focused approach ensures that dashboards and charts and every data reporting tool exists to answer user needs. This is even MORE important when we consider (and we should ALWAYS consider) different cultural, racial, and identity contexts. WE don’t know our users’ experiences. Only THEY know them. We have to LEARN from them.

It’s EASY to tell data stories with dynamic data.

On to the 2nd “problem” of narrative dashboards being impossible to build because the data feeding into them is dynamic and one can’t tell stories if the insights are changing.

Seriously, I just read a LinkedIn post from a data communication “expert” the other day that said one couldn’t tell stories with dashboards “because sometimes a metric may be increasing, and the next day it may be decreasing.”

This is, frankly, bulls–t.

Have these people used a computer program in the last decade? Applications built for data visualization (Power BI, Tableau, R, and more) and even those where data viz is icing on top (Excel, for instance) ALL have ways to look at what data is doing and tell a story with it, even when it’s different every day (or hour, or second).

We can literally get a dashboard to do calculations (simple or very complex) and spit out a narrative title/sentence/paragraph that accounts for what the data looks like AND can generate an entirely new title/sentence/paragraph the very next day/hour/minute. It’s not just narrative text either, but visual colors can change, graphics can change, everything can change and adapt to what the data is doing.

It’s not even hard. When I teach half-day workshops to people entirely brand new to Power BI and we have 10 minutes to fill at the end, I teach them the basics of dynamic story-telling..

In 10 minutes. To people absolutely brand new to the software.

It’s not only extremely possible to build narratives and story-telling into dashboards, but if dashboards DON’T do this (which most don’t), they probably aren’t used as much as they should be.

In summary, don’t let anyone tell you that a dashboard can’t be a narrative tool to communicate stories about data.

If someone is saying this, they just haven’t thought about data and communicating it with an end-user in mind enough.

They need a user-oriented paradigm shift.

They need to STOP thinking about THEIR experience with a dashboard and SHIFT to thinking about the experience of their USERS. A whole new world opens up with that paradigm shift.


TraversData.com | LinkedIn | Instagram | Threads

Put the People in your data first.

Hey everyone,

We have a bit of a departure for this issue of People-Friendly Power BI. While we always talk about how to make Power BI easier to use for our dashboard and report audiences (the people!), this issue is about the people in our data.

Those of you (hi!) who I’ve worked with, or have taught Power BI to (or even been following this newsletter for a bit) know that this newsletter has the name it does because for me, the most difficult and stupid part of working with data is how BADLY it’s communicated to people. It is my #1 pet peeve.

You’ve put all this money and time into collecting data, cleaning data, analyzing data, and then you get to the communicating / reporting stage and …. you do this part badly and all that money and time on the previous steps is basically wasted.

Either data is visualized badly (ignoring the needs of the audience) or it’s communicated by someone who is too techy (sorry techies!) and while they have the “hard” data skills, their people skills are not stellar, and communication breaks down.

Frequently, the issue is BOTH. Bad visuals AND bad communication.

We Gotta Figure It Out GIF - We Gotta Figure It Out Communicate GIFs

I named this newsletter People-Friendly Power BI because I make Power BI friendly to you, who may need to start using it (or continue using it) to develop reports but ALSO for your audience so they’ll LOVE your reports.

People are important to me. They are the most important part of data. They make decisions with data. Use data to move along a process or project. They ARE the data.

Data is people.

(I’m refraining from putting a gif here of Charlton Heston yelling “It’s People!!! It’s made from people!!!”) You’re welcome.

So, what do I mean by this “data is people” statement? Of course it’s people, you say. Data is about people.

We SAY this. Data people SAY this. But do we pay attention to it? Way too often, data is analyzed and displayed WITHOUT thinking about the people and the lives that make up that data. Way too often, a data analyst is trying to make a good chart, or trying to find a pattern or story in the data. It becomes just numbers.

Let’s use an example.

Here’s two charts made with the same data. One is a bar chart, one is a beeswarm (some people call it a jitter plot, but that’s less fun).

Left: Bar chart showing pay disparities between 4 groups of restaurant workers. Right: Jitter plot showing the same data.

The bar chart is aggregating everything together and visualizing the average for each group. Nevermind that a bar chart isn’t the best way to show an average, lots of people use bars for averages… look at the Groups… and imagine if instead of Group A,B,C,D we had races listed instead. Some conclusions, generalizations, and stereotypes would jump out immediately.

The beeswarm is the exact same data, but the individual data points are shown, and they show that the “story” isn’t so simple as the one being told (or being assumed by the audience) by the bars. The averages are really NOT the story. The spread and variance are the story.

(Charts by Eli Holder, via Dr. Stephanie Evergeen. Read more about how visualizing data in a society with hundreds of years of built up systemic racism can make things worse here.)

The data is people.

Let’s take this one step farther, and it’s a step we should all take constantly. It’s a step that anyone working with data should think about. The data isn’t numbers (or words). Any data worth working with (in my opinion) is capturing the lived experiences of people. Their births, their lives, their deaths.

I recently worked with Mamow Ahyamowen, which is an epidemiology alliance of First Nations governed health service organizations based in northern Ontario. Their goal is to provide health information to communities that they can use to work toward health equity.

Before I even started working with them, I attended a virtual session where they presented some health data (specifically mortality data), but before launching into all the numbers of deaths and co-morbidities and how their mortality data compares with province-wide data, they paused.

They paused to reflect on the nature of the data, who the data was about, and what it meant to all of them. When I asked them about it, staff shared that this practice was inspired by teachings from Elders and Knowledge Keepers who have been involved in their work.

More organizations (and the “data world” as a whole) need to do this. Data is not just data. It’s lived experiences, ancestors and history.

Mamow Ahyamowen’s health data is perhaps the perfect example of this. Indigenous communities in Canada (and all over North America, and the world) have lived through centuries of land theft, oppression, and systemic racism. Families and communities were ripped apart throughout the 19th and 20th century and thousands of children were abused and killed in government-run and church-run residential schools. It doesn’t end there; colonialism has changed forms but continues to affect the lives of Indigenous people to this day.

All this affects the health of communities in a big way.

I trained a lot of great people at Mamow Ahyamowen in getting started with Power BI, and these were training sessions that used their data. Data about their families and ancestors.

So, the first thing we did at the start of every session was the same thing they did during the webinar, like all their meetings where data plays a part. We paused and honored the people and lives captured by this data.

We also built a slide like this into the start of the presentation:

This wasn’t lip service. We shared links and phone numbers to mental health and crisis support lines. Because exploring data that tells a story about how your community and family has been abused for generations is traumatic.

I really wish more organizations and data people would think about things like this.

We, the people who collect, analyze and communicate data, cannot in good conscious ignore the stories in the data. We can not present it “without bias” (which is impossible), no matter how much we try. The bias is implicit and systemic.

But we can expose the bias, know why it exists and counteract it, alongside the racism systems that perpetuate it. We can also think mindfully about the lives and communities the data speaks to. Context matters. Data is useless without context.

So this is what I’m doing.. because I think I can use my privilege and skills to help in some meaningful way.

What are you doing?

More about Mamow Ahyamowen:

Building on the success of the mortality analysis discussed above, Mamow Ahyamowen is working on three new projects that will explore chronic conditions, mental health and addictions, and injuries. You can stay up to date with their work by signing up for their newsletter.

More about Maureen Gustafson

(Maureen helped me write and lent her thoughtful editorial eyes to this post):
Maureen Gustafson is a member of Couchiching First Nation with mixed Ojibwe and settler roots. She grounds herself first and foremost in her relationships as an auntie, sister, daughter, cousin, and friend. Maureen holds a Master of Public Health and is privileged to serve Mamow Ahyamowen as a Knowledge Translation & Exchange Specialist.

How to design in Power BI (and anything else) like a freakin’ rockstar.

Hey you. How’s it going?

This is my mom, Marguerite (or Margie, as everyone knew her).

Mom passed away a few years ago, and I still miss her every day, and she factors hugely into today’s post about design.

Mom was an x-ray and ultrasound technician. For over 40 years… from the late 1950s to when she retired in 2005. She worked with a huge range of technology over a lifetime of giving x-rays and ultrasounds to tens of thousands of people.

Adults, kid, famous hockey players.

She was such an expert in her field that by her 40s, she could spot things some radiologists (like, doctors trained to read x-rays) missed. Young doctors would ask her to check over x-rays sometimes, just in case.

So, she knew EVERYTHING about x-rays and ultrasounds, but also managing an office (because she’d usually have to do that too) and putting people at ease (both kids and adults) because they had to be calm and motionless for the scans to be clear.

Also, mom never could figure out home computers. Deciphering them was like learning a new language for her, and by the time there was one in their house, she was at the point in life where she didn’t need to.

She’d often say “I’ve somehow survived without using a home computer so far… I’ll keep doing what I’m doing”.

Not to say she didn’t try, but at some point the effort exerted wasn’t worth the benefits to her. She didn’t need email.. she’d phone us when she wanted to talk.

I miss those calls so much.

I’m sure you know someone like my Mom. Someone who is an absolute ROCKSTAR at what they’ve dedicated years and years to. Someone who can tell you EVERYTHING about that subject, whether it’s x-rays, or event production, or e-commerce, or app development, or evaluation survey creation, or whatever.

I’m a rockstar with visualizing data with Excel and Power BI. It’s not bragging if it’s true, right?

How does this relate to design in Power BI? People-friendly Power BI? Or in anything else. Maybe you’ve already figured it out.

When you’re designing something (anything), there’s a 99.99% chance you’re gonna be sharing it. You try and design something awesome, because if who you share it with really likes it, they’re gonna share it with others.

It could be your boss sharing it with their boss, and their boss sharing it with your Board of Directors.

Or it’s something (a graphic, a chart, a dashboard, whatever!) you put out on social media with the hopes of it going viral (we’ll ignore the dumpster fire that social media is for now).

The point here is that you are making something for an audience. It’s not going to be only YOU looking at what you’re designing. You are designing for others, and they do not have the same skills that you do.

When designing with Power BI, YOU are a data person. You probably know that data inside and out. You dream about that data (or have nightmares about it).

But, no one else does.

Yet, most Power BI dashboards are designed like their audience are data people. Most of them cram a ton of data onto every page, with no explanation about what the data means, or if a chart is showing something good or bad, or even worse, there’s a giant table of numbers with no explanation why any of them may be important.

Power BI developers design for themselves. If they can understand a dashboard, then it’s awesome… everyone else should “get” it too.

But most Power BI developers are data people. Like, hardcore data people. They know the difference between “On Prem” and “Azure” and why a company should use a Data Lake versus a Data Warehouse (or vice versa). They are AWESOME at what they do… making data work.

They are not good at communicating the messages and information that data to others (Sorry data people. you know I’m right). If you have a data person who is the life of the party, belting out karaoke tunes, having regular conversations with regular people, you have a unicorn on your hands (and that unicorn will probably be promoted soon, as they can provide a bridge between technology and people).

Unicorns are super rare. That’s why they’re called unicorns.

You NEED to understand and be able to talk to people to design great Power BI reports that people will USE regularly, if not LOVE.

That’s why, whenever I design a report for any audience that isn’t data people (for a CEO, or a Board, or a department full of people who’s job is NOT data, or something that will be on a website, or shared on social media), I think of my mom.

Would mom understand this dashboard? Could she look at it and immediately know what it was saying? What the key insights were? We’re talking seconds here. Would she know everything she needed to know within seconds?

This is YOUR audience when you design. My mom (metaphorically).

Your audience is people who are excellent at what they do, which is almost certainly not data. People who need to use your dashboard (or chart, or whatever) and get something from it immediately.

They don’t have the time (or the desire) to click around and filter things different ways and waste a bunch of time looking for a number.

You HAVE to remember that.

I’m not saying I’m perfect at this. When I get feedback about a dashboard, it sometimes is like “Joe, I love this… but I couldn’t quite figure out …” and as soon as I hear that I listen VERY closely to what they couldn’t figure out. I don’t explain to them where to find the information they need, I redesign and GIVE them the information, so they don’t have to go looking for it.

It’s what Mom would’ve loved.

We’re not designing for data people. We’re designing for normal people.