How to Build a Brain-Friendly Bar Chart in Domo

by | Jun 29, 2017

The life of preparing and presenting data isn’t easy.

You pour a lot of hard work into design­ing data visu­al­iza­tions that enlight­en your stake­hold­ers and inform decisions.

And yet, some­times your stake­hold­ers still don’t seem to “get it”. I can tell you, I’ve been there.

It took me years of data visu­al­iza­tion train­ing to under­stand that there were two key rea­sons my dig­i­tal ana­lyt­ics data charts weren’t get­ting under­stood and act­ed upon:

  1. I wasn’t choos­ing the most opti­mal chart type for my data story.
  2. I wasn’t ren­der­ing the chart in the most opti­mal way for comprehension.

That’s why I find my work at Search Dis­cov­ery so impor­tant and ful­fill­ing. One of the most reward­ing things I do is help our busi­ness intel­li­gence team work with clients to apply a neu­ro­science-based data design best prac­tices using Domo, an enter­prise data visu­al­iza­tion plat­form that is tak­ing the busi­ness world by storm. Search Dis­cov­ery is a Domo Pre­mier Part­ner with 150 suc­cess­ful imple­men­ta­tions and counting.

The project’s mission? 

To ensure that the visu­als we cre­ate in Domo for our clients are opti­mized for accu­rate inter­pre­ta­tion and poised to effec­tive­ly inform busi­ness deci­sions. Work­ing with many of SDI’s tal­ent­ed dig­i­tal ana­lysts has shown me that there is a strong need for guid­ance in this are­na, and the two chal­lenges I faced years ago are eas­i­ly sur­mount­ed with a time-test­ed approach.

Domo

So allow me to skip you ahead a few grades, because I firm­ly believe that a strong foun­da­tion in data visu­al­iza­tion begins with under­stand­ing why and how to cre­ate an effec­tive bar chart. Per­fect­ing a bar chart is the foun­da­tion­al exer­cise of Search Discovery’s data visu­al­iza­tion work­shops, and a tool such as Domo is a great place to build them!

A sin­gle bar chart is the Swiss Army knife of visu­al­iza­tions; hor­i­zon­tal bars are per­fect­ly suit­ed to com­mu­ni­cate cat­e­gor­i­cal rank­ing or com­po­si­tion, while ver­ti­cal bars are suit­able for sin­gle met­rics trend­ing over time.

But they’re so sim­ple, you say! What fan­cy skills could be need­ed to con­struct a prop­er bar chart?

Some believe that effec­tive pre­sen­ta­tion of data is about what you include. For me, effec­tive­ness is pro­por­tion­al to what you take away so that the sim­plic­i­ty of your mes­sage clear­ly stands out.

And a big piece of that is remov­ing visu­al noise that you don’t need, oth­er­wise known as cog­ni­tive load. Cog­ni­tive load is defined as “the total amount of men­tal effort being used by the work­ing mem­o­ry.” And the work­ing mem­o­ry is “a cog­ni­tive sys­tem with a lim­it­ed capac­i­ty that is respon­si­ble for tem­porar­i­ly hold­ing infor­ma­tion avail­able for pro­cess­ing.” What does this mean exactly?

Think of your work­ing mem­o­ry as try­ing to jug­gle balls that rep­re­sent pieces of infor­ma­tion you’re try­ing to focus upon. Each bit of visu­al noise in your chart is adding a ball, and even­tu­al­ly, there will be so many in the air that you’ll drop them all.

Visual noise in your chart is like forcing your audience to juggle too many balls at once.

What’s the solu­tion for pre­vent­ing brainy ball-dropping?

Since I’m a clos­et crunchy gra­nola hip­pie, one of my favorite dai­ly prac­tices is detox­ing. And, I car­ry that prac­tice all the way through to my data visu­al­iza­tion work to ensure every pix­el pro­vides a purpose.

Luck­i­ly, Domo already incor­po­rates some data visu­al­iza­tion best prac­tices into their default bar pre­sen­ta­tion which def­i­nite­ly ele­vates it a few steps up from a more ubiq­ui­tous data viz tool. Even still, there are just a cou­ple of tweaks that will tru­ly allow your data sto­ry to stand out.

And in today’s post, I’ll show you exact­ly how to mod­i­fy the set­tings of bar charts in Domo to instant­ly improve clar­i­ty and com­pre­hen­sion with our Brain-Friend­ly Domo Bar Chart Checklist.

The Brain-Friendly Domo Bar Chart Checklist

No time to read the Domo Bar Chart check­list right now? Down­load for Domo cleanup to-go

 

This check­list will method­i­cal­ly go through each of the Chart Prop­er­ties menus with brief expla­na­tions under­neath. I’ve cre­at­ed an exam­ple using a the­o­ret­i­cal sales by mar­ket­ing chan­nel sce­nario. I chose the Sin­gle Hor­i­zon­tal Bar chart option and select­ed Mar­ket­ing Chan­nel as our dimen­sion and # Sales as my metric.

Ren­der­ing the graph with no mod­i­fi­ca­tions results in some­thing like this:

Let’s get this bar chart shipshape!

 

1. General

a. Font Size = Largest

i. Domo cur­rent­ly only allows three font sizes. I always sug­gest the largest size to accom­mo­date your most hard of see­ing end con­sumers. Reduce the font size only if your view­ers are view­ing the card data on Pages where the labels are forced into a diag­o­nal rota­tion. It is eas­i­er to eas­i­er to read small­er hor­i­zon­tal labels than larg­er diag­o­nal ones.

2. Sorting — There are two methods:

a. In the Gen­er­al menu, set Sort on Totals = Descend­ing.

b. OR Drag whichev­er met­ric the chart is dis­play­ing to the Sort­ing card and choose Aggre­ga­tion → Sum and Order →  Descend­ing (unless you are encod­ing a met­ric where low­er is more important).

c. NOTE: DO NOT cat­e­gor­i­cal­ly sort bars that rep­re­sent a time series, such as months. Time is best rep­re­sent­ed lin­ear­ly from left to right.

3. Bar Settings —> Height / Width Percentage = 75

a. The opti­mal width between bars is half of each bar width. In Excel, you would change Bar Gap Width to 50%. How­ev­er, in Domo, this set­ting rep­re­sents the width of the bar PLUS the gap between the next bar. So in order to arrive at the appro­pri­ate width, you set the total width per­cent­age to 75% (leav­ing off the % sign).

4. Grid Lines

a. Zero Line Col­or = Trans­par­ent  {UPDATE: Domo no longer allows you to set any col­ors to Trans­par­ent or a hex val­ue of your choice. Your pri­or set­tings should be grand­fa­thered in, but going for­ward you can­not make this change.}

b. Check Remove max/min lines (unless these lines pro­vide clear val­ue to you and the consumer)

5. Data Label Settings

a. Text =  %_VALUE, %_PERCENT_OF_TOTAL (if dis­play­ing composition). 

i. I direct­ly label each bar so that there’s no need for the audi­ence to inter­pret the bars against the axis or grid lines. Espe­cial­ly for com­po­si­tion graphs, I find both the absolute and per­cent of total val­ues to be of use to viewers.

b. Posi­tion = Out­side Right

c. Use Scale For­mat → Yes

6. Value Scale (X for horizontal bars, Y for vertical bars)

a. Max Grid­lines = 0

i. Grid lines serve the pur­pose of allow­ing the view­er to inter­pret each bar’s length, yet they add sig­nif­i­cant visu­al noise to the graph. I rec­om­mend remov­ing them to cre­ate a clean plot area and add data labels instead (stay tuned). At time of writ­ing, the only way I know of remov­ing all but the top grid­line is by set­ting the max to zero.

b. Val­ue Scale Max = 1.25 X high­est bar value

i. The most effec­tive bar com­par­i­son hap­pens when the max­i­mum bar val­ue is about 75% of the scale [source]. You can achieve this by cal­cu­lat­ing 125% of your max­i­mum bar val­ue and enter­ing that as the Max. Keep atten­tion on this over time if your cat­e­gor­i­cal or rank­ing val­ues dra­mat­i­cal­ly fluctuate.

7. Hover Text Settings

a. Text = %_VALUE, %_PERCENT_OF_TOTAL

i. I like to make my Data Labels and Hov­er Text con­sis­tent for rollovers; feel free to add rel­e­vant data points as desired.

b. Per­cent Val­ue Dec­i­mal Places = .0

i. I set the dec­i­mal places of any label to the num­ber required to pro­duc­tive­ly dis­tin­guish between two val­ues in the view. Once again, this is to min­i­mize extra­ne­ous visu­al noise. Think 12.4% and 12.9% vs. 12.395837% and 12.900478%.

8. Colors — Use neutral blue or gray palettes (avoid red and green)

a. Col­or is a cor­ner­stone of effec­tive data inter­pre­ta­tion. Red and green have cul­tur­al sig­nif­i­cance in our cul­ture and could poten­tial­ly skew a viewer’s per­cep­tion of the urgency or required atten­tion behind a data point. Domo did a great job of choos­ing an emo­tion­al­ly-neu­tral blue as the default col­or choice. Ide­al­ly, I would sug­gest turn­ing all bars gray and then using blue to sin­gle out a spe­cif­ic data point in your sto­ry. Since this func­tion­al­i­ty isn’t out of the box at time of writ­ing, I sug­gest choos­ing either blue or gray for all bars.

9. Make the Card Title succinctly informative.

a. I like to either clear­ly state the cat­e­gories and met­ric being dis­played, or phras­ing the card title in the form of a busi­ness ques­tion that the graph is answering.

10. Adjust the format of the Summary Number

a. I sug­gest con­sid­er­ing what num­ber would answer the key ques­tion being answered by the graph at the high­est aggre­gate lev­el. And remem­ber to use the most effi­cient scale abbreviation.

11. Adjust the Label to clearly described the Summary Number.

 

12. Voila! You’ve just created a beautifully simple, clean and easy-to-interpret Domo bar chart:
BONUS TIP: If you’d like to use col­or to empha­size one data point, hov­er over your select­ed data point and take a screen­shot. You’ll have an image that high­lights one bar in a stand­out blue with the oth­ers set to the back­ground like so:

Final Thoughts

It is imper­a­tive for a data design­er to cre­ate a seam­less visu­al­iza­tion expe­ri­ence for their audi­ences. This will enable your orga­ni­za­tion to make the most out of even the most advanced data visu­al­iza­tion plat­forms. I hope I’ve shown you today that the sim­plest tweaks can make a big impact on how your data sto­ry hits home with your exec­u­tive audience. 

And, if you’re ready to make the most out of your Domo invest­ment, or more infor­ma­tion on the Search Dis­cov­ery data visu­al­iza­tion work­shops, con­tact us and learn how we can help you make bet­ter deci­sions with your Domo data.

Your Search Ends Here!