In this week’s In-Ear Insights, we present the full session audio from our talk at the MarketingProfs B2B Forum 2019 on how to build better analytics dashboards and reports with data storytelling and analytics frameworks. Listen in for key principles, and if you’d like to get the slides and video from this talk, they’re available here.
Subscribe To This Show!
If you're not already subscribed to In-Ear Insights, get set up now!
- In-Ear Insights on Apple Podcasts
- In-Ear Insights on Google Podcasts
- In-Ear Insights on all other podcasting software
Advertisement: Generative AI for Marketers
Imagine a world where your marketing strategies are supercharged by the most cutting-edge technology available – Generative AI. Generative AI has the potential to save you incredible amounts of time and money, and you have the opportunity to be at the forefront. Get up to speed on using generative AI in your business in a thoughtful way with our new offering, Generative AI for Marketers, which comes in two flavors, workshops and a course.
Workshops: Offer the Generative AI for Marketers half and full day workshops at your company. These hands-on sessions are packed with exercises, resources and practical tips that you can implement immediately.
Course: We’ve turned our most popular full-day workshop into a self-paced course. The Generative AI for Marketers online course launches on December 13, 2023. You can reserve your spot and save $300 right now with your special early-bird discount! Use code: EARLYBIRD300. Your code expires on December 13, 2023.
Key takeaways you'll receive:
- Hands-on learning for the six major use cases. When you're done with the course, you'll have your first collection of working prompts based on your business and needs
- A plan and sales pitch to help you sell the use of AI internally to stakeholders
- A roadmap for who should be on your AI pilot team and your internal AI council
- Your first custom AI that you will build for your business (paid version of ChatGPT required)
- Tactical and strategic knowledge of generative AI to help you lead your team in the new AI landscape
- A priority matrix of the tasks you'll want to consider for AI enablement
Sponsor This Show!Are you struggling to reach the right audiences? Trust Insights offers sponsorships in our newsletters, podcasts, and media properties to help your brand be seen and heard by the right people. Our media properties reach almost 100,000 people every week, from the In Ear Insights podcast to the Almost Timely and In the Headlights newsletters. Reach out to us today to learn more.
Listen to the audio here:
- Need help with your company’s data and analytics? Let us know!
- Join our free Slack group for marketers interested in analytics!
What follows is an AI-generated transcript. The transcript may contain errors and is not a substitute for listening to the episode.
Today we’re talking aboutanalytics and dashboard specifically, I’m going to actually contradict Matt a little bit if you want the slides right now, you can go to where can I get the slides.com.
You will also see the links to some of the Data Studio dashboards that you can copy into your own Data Studio instance.
So if you want us to use some of the examples in this talk, you can get them there as well.
So there’s the aptly named Where can I get the slides calm? Let’s move on.
Here’s the major problem we have as marketers, b2b b2c, who cares what the name is, this is our problem.
We have no shortage of data.
This year as a civilization.
We will be creating approximately 40 zettabytes of data.
Now, a zettabytes kind of a hard number to wrap your brain around.
How many of you use Netflix just a quick show of hands? Okay, good.
So everyone knows on Netflix If you were to start watching Netflix and not stop to eat,sleep,do biological functions of various kinds.
And you started 55 million years ago, you would just use one zettabytes of data today.
That’s how much it’s going to create 40 of these as a civilization.
Now, as marketers, you also know that we’re not really good about making use of the data that we have.
So in the 2019 Pricewaterhouse SEO survey, they asked CEOs how critical are certain types of data, you’ll see all the top line data about your customers and clients preferences and needs that top yellow line 94% of SEO said that is essential.
That is critical information, financial forecasts brand and reputation, risks to the business.
SEO say we want to know what’s going on.
When asked do you get that? That’s what that bottom line is there that lovely yellow line that says 15% of SEO You are getting what they asked for.
It doesn’t take a rocket scientist or a data scientist to realize that you’re going to get fired if you keep giving the SEO, what they do not want.
Compounding that the state of dashboards and reporting today looks a lot like this.
We have lost the ability to communicate in any meaningful fashion.
We just sort of puke everywhere, as many numbers as possible with colors and hope that somebody can read this.
This is the takeaway from this entire session, you could walk out of the room right now.
If you take this to heart, a dashboard, without decisions, or decorations.
If you look at a dashboard, you look at a report of any kind and you can’t make a decision.
It’s worthless, throw it out, start over dashboards without decisions or decorations.
Thanks, it’s been great.
I’m just kidding.
The impact is no surprise, right.
We don’t communicate data meaningfully which means we don’t ever get anywhere.
We don’t make any money, we get fired.
This session had this whole data storytelling thing.
Well, we’ll cover a little bit about that day storytelling is pretty simple telling stories with data is the first part of that sentence is important.
What is a story in context of data? Here’s what it isn’t.
But yet, this is what we do to people, right? We just sort of pour data on their heads repeatedly and hope that they get it.
This is not storytelling, this is just injury.
Most stories have a logical flow of some kind, you know, there’s a beginning of the story and middle of the story and an End of story.
Maybe if you’re talking about data, you’re talking about taking a small piece of data and expanding the insights that you get from it.
Or maybe you start with something like why does something happen? What is it and how does it work? All of these are very simple formats, you know, all these formats.
If you were raised by caregivers who read us stories to go to sleep, you know almost all these formats there’s like seven total, Christopher Booker wrote a book 2004 on the seven basic story types, so you know how to tell stories you’ve been doing all your life.
Yet we completely forget this.
When it’s time to actually do our reporting.
Again, we don’t tell a story and without a story, nobody can know what’s going on and they can make a decision.
So here’s the if you have nothing else when it comes to data storytelling, if you’re not sure where to start, start with a three what’s what happened? So what? Now what? Every dashboard, every report, every analysis should answer these three questions in order What happened? So what now what what happened is pretty easy.
And most of the time, the existing analytics tools you have like Google Analytics, so your marketing automation software or your CRM can kind of similar to spit that out.
In terms of the time and energy and money that you spend.
This is where you should be spending your time court.
depend on what happened a quarter of it on why so what what does it mean? And then half your time should be on what are we gonna do about it? What decision Are we going to make? Because without that, you’re making declarations and Wall.
It’s great to be the arts and crafts department you want actually to get paid.
When you look at dashboards, that work sometimes there’s not a whole lot of data on them, but there’s a whole lot of writing.
Right Avinash Kaushik is famous for saying the higher up the dashboard goes, the less data is on and the more explanation, the more storytelling to be able to understand what happened.
So what now what, this is an okay dashboard to look at, um, it’s not the prettiest thing in the world, but in terms of being able to understand what to do next.
That recommendation section is so important, that is the now what.
And a dashboard is not the only form of data storytelling that you have access toWe’ll get to talking about your stakeholders in a bit.
But just keep in mind that people ingest information differently.
Right? We did some client work for a telecom who SEO wanted his reporting, eight and a half by 11 Word documents printed out and bound that every week why he was spending so much time on airplanes that he just put this in his bag and read it.
That was how he wanted his reporting.
And it worked for him totally fine.
Some people don’t want dashboards.
Some people just want a PowerPoint to look at.
So you’ve got to be able to create dashboards that can either adapt to a PowerPoint, or if that’s just what your company or your client wants, you give them what they want.
Some people need to feel like they they’re being guided through reporting.
This is an example I for one of our customers, I do YouTube videos walking them through their reporting is so they have a video to watch because that’s how they learn.
So think about that you have screen capture software like you know, Camp Asia, for example, you can screen cast your reporting on your explanation of it so that people can follow along.
This is really important.
If your data and your reports move around an organization if you just hand a report to somebody, and that’s the only person who reads it cool.
But in a lot of bigger organizations, that report that gets handed around and around and around, and like the game of telephone after it gets to the 15th desk, nobody has any idea what it is.
When you send the video with your narration, the 15th person can still watch the video and go, Oh, I see what you were saying.
And so that’s an important format.
How many of you listen to podcasts? Wow, like five years ago, no hands without now it’s wonderful.
Guess what? podcasts can be a form of reporting.
You can read out and explain and describe in audio format just with the voice memos app on your phone, and send that mp3 file to stakeholders like hey on your commute or in your at the gym, or when you’re cooking at home.
You can get the data you want and you know, in the format that works best for you.
So think about when we’re talking about storytelling with data.
It isn’t just put up a dashboard and walk away it is consider the needs of how people read and watch and listen to information and give them what they want.
There’s a concept in the discipline called neuro linguistic programming called throwing mattresses.
Everyone’s brain has a is like a door and every door is differently shaped.
There’s only so many ways a mattress is going to get through each door.
So you have to figure out what format audio video text gets through the doors, the people that you’re sending your reports too.
So keep in mind all these options, what storytelling medium works best for your audience.
So that’s the storytelling part.
Let’s talk about the data part.
How many of you have have or currently drive a vehicle? Okay, this number How many of you have ever tried and I don’t advise you to use driving a car just by looking at the dashboard and not out the window.
This is a bad idea for liability purposes, we have to say do not do this.
It’s a bad idea.
But as marketers, sometimes we think these dashboards are the be all end all of marketing, we put these things up, like, isn’t it beautiful? Let’s run all our marketing.
No, that’s not what it’s for, like the cards to help you make decisions.
But most, the most important decisions are not on the dashboard.
It’s the tree in front of you.
But you should stop driving at that point.
dashboard creation, reporting creation of any kind is like cooking, is the process of cooking, if you have ever cooked anything or tried, the process of creating a dashboard is is very much the same.
First, you got to decide who you’re cooking for, and what you’re cooking.
If you don’t know that, you’re not gonna you’re not going to make people happy, right? If someone is expecting breakfast tacos and you hand them sushi, it’s kind of the same thing but not really, at all.
This is the cardinal sin of Most dashboards, and I heard this this week at one of my customers, we made the one dashboard for marketing.
There is no such thing.
This is a fallacy.
This is a fool’s errand.
There is no such thing.
Look at your organization’s org chart.
Does the person at the bottom and the beige box need the same information as the person up top in that hot pink box? Of course not the CEO and the junior specialist need very different things.
But yet, we’ve come to think that we can create a magic dashboard the answers everybody’s questions.
As an exercise, look at your own work chart and create a grid of the roles in the organization horizontally and the team’s vertically.
Every box in that grid needs its own dashboard.
Because what the CMO needs is not what the VP needs is not what the director needs is not what the manager needs.
You’re going to need to create dashboards for each role.
If you want to be the as possible in your recording.dashboard creation is software.
Its software creation.
So one of the first things you do in software creation is doing what are called the user experiences a user explanations, what is the user statement of this person that you’re making a dashboard for? And it’s very simple three part sentence who, what and why? You asked somebody in the organization, why do you want a dashboard as cmo, the who I can see overall performance, the what? So that I can demonstrate marketing’s impact to the board the Y this user statement tells you exactly what needs to be on the CMOS dashboard, right.
Which means like you shouldn’t have Facebook likes on the CMOS dashboard, right.
As a marketing manager, I can see the performance of my team so that I can improve individual performance.
That’s what the marketing manager needs out of their users store.
Why they want a dashboard? What are your user stories? You need to ask everybody on that grid, what they want, who, what and why? To make your reporting as impactful as possible.
So we know what we’re cooking, we know who we’re cooking for.
Now, we got to find the ingredients.
Data is your ingredient.
And just like your kitchen, your refrigerator.
You need to know how to cook.
But you also need to have the right ingredients.
No matter how skilled a chef you are, if all you have is mac and cheese, guess what you’re having for dinner, you’re having mac and cheese, right? No amount of creativity alters the fact that that’s the only thing in your pantry.
So the first part of this is to figure out what data do you have access to? This is an example from Google Data Studio.
There are 178 different data sources that can connect to how many of those are in your organization.
Do you know have you done a catalog? Have you looked around See what’s in the pantry.
And this is only for stuff that you can connect to directly.
There’s a concept in marketing technology called data governance, which basically is who’s in charge of what, where and how much you’re paying for it.
Data Governance is a great excuse to be able to dig around within your marketing organization or your company overall say who’s got the data? Hey, sales, can we have access to the CRM data? A finance can we have access to the RP support? Can we have access to the call center data and find all the pieces identify what you have access to.
There’s also third party data outside data as well.
If you are at all concerned, for example, about talent management, there are reports put up by the US government paid for with for those who are citizens, your tax dollars, for those of you who are not citizens of America, you’re welcome.
of industries and they’re hiring and firing and the layoffs and things like that if that is something that can be predictive to your Marketing, you might want to think about those third party data sources as well.
So you’ve got the ingredients.
Now it’s time to prepare them.
This is the hard part of data dashboarding.
From a software development perspective, there’s an expression coin in 2006 data is the new oil.
And it’s a really good expression.
Because if you’ve ever seen crude oil, it’s useless.
It’s a black tar mess that does nothing.
You have to with crude oil extracted from the ground, refine it, and refine them and then deliver the final product, whether it’s gasoline or plastic cereal bowls to the market.
The same is true for dashboarding and your data, you have to extract the data from where it lives in your company, process it, which is computation, doing all your analysis, and then visualize it with dashboarding software.
Do not attempt to make your visualization tool, do computation.
Right? Every dashboard vendor on the planet so We can do it all.
You can maybe, but it’s like trying to do it all with a multi tool, right? a multi tool does a bunch of things pretty crappy, right? But the only thing it’s this particular ones good at is the the pliers.
That’s it as opposed to using the right tool for the right job.
Right, you are going to be much better off separating your computation from your visualization.
Find the best tool to visualize data easily, but don’t expect it to be able to do the hardcore analysis.
A lot of people will look at services like G to crowd and stuff like that.
There are dozens if not hundreds of dashboard tools, BI tools all these things and marketers pay up add these companies pay exorbitant amounts of money to have their their things listed.
When it comes to b2b marketing.
These are the three dashboard services.
I recommend that if you work for a dashboard service that’s not on here.
I haven’t tried your software for most marketers in most instances, most of the time.
Google Data Studio is your best choice for a couple reasons.
It’s free.most marketers use Google Analytics even though shops that have Adobe amateur have Google Analytics typically running alongside it because somebody is paranoid enough to switch a worry that they want to backup a check on.
So that’s there.
And looks like Google spreadsheets and Google documents, it looks familiar enough that you should be able to get started relatively quickly.
And there’s a ton of training for it.
If you’re in a situation, you have a lot of very complex data or multiple data sources and the organization is is messy or you’re in a highly regulated industry where you’re not allowed to use some cloud services.
Tableau is probably the best choice for you.
Because you can run it locally if you have things like HIPAA compliance and, and such.
And then if you want to integrate with artificial intelligence software, machine learning heavy compute, IBM Watson Studio is probably the best choice for those environments.
But for most people in this room, Google Data Studio is the probably the best place to start.
So you’ve got the ingredients, you’ve prepped them, you’ve done your computation.
Now it’s time to cook, cook up the dashboard.
Remember, dashboards are software.
So you have to go through the software development lifecycle.
Like any other software product.
If you want your dashboard to succeed, that means first talking to your users, getting those stories, getting their feedback.
That means wire framing stuff.
Do not just open Google day studio and start pressing buttons.
It’s a really bad idea.
Sit down with some or go to a hang those giant posts notes on the wall and start drawing out what you think based on the user stories people need, and test and test and test until you can get to a minimum viable product, something that satisfies those user stories.
This is an example of a dashboard I use for my own website.
I put the big numbers that I care about up top with green and red arrows because if that’s the only thing I look at at least have the information I need to make decisions.
And then if I want to investigate any of those numbers, I look at the different columns, the green, yellow red columns for more and more detail if I need two more information to make a decision, but that’s a good minimum viable product.
This, by the way, is one of the dashboards you can clone from the link I shared at the beginning.
Just because you cook it doesn’t mean somebody wants to eat it.
I found out the hard way when I tried making chocolate oatmeal.
So you have to do testing with what are called sponsor users or beta testers, whatever.
But ultimately, what you want to do is if if you have that that grid of roles in an organization you want to find, like say we’re doing the CMOS dashboard, you want to find that CMOS direct report and have them be the beta tester because they have a pretty good idea of what their boss wants.
Say, does this meet the boss’s needs? Are they going to get what they want out of this? Remember, they have to help you say yes This will help the CMO make decisions about budget or priorities.
That’s what the sponsor users do.
They tell you Yep, I can’t make a decision from this.
And then you roll it out, roll out the recipe, send out the dashboard to people.
And this is again, where most dashboards go off the rails.
We throw them out there, and we hope that it’s done.
And then we move on to the next thing on to do list.
Remember, dashboards are software, which means you need to provide tech support.
Just like any piece of software, you need to constantly be providing tech support iterating, taking into feet taking feedback in and providing new versions and updates to people as you need them.
So remember, there is a tech support component to a dashboard like any piece of software.
So that’s how to do it.
Let’s talk about how not to do it.
This six ways that your dashboards is going to completely fall apart on you and be useless.
Number one, role dashboards that decisions with decorations Then we get into visualization, there’s sick this what we call the 60 framework of good visualization clear, clean, complete, concise, sighted and conclusive.
We’re going to walk through each one of these.
First, your dashboard must be clear, you must know what you’re looking at.
Most people when they put together dashboards for the first time, they’re like that 12 year old who discovers the font menu and a word processor and of course, they try every single thought, right? We’ve all done that.
We do the exact same thing with dashboards, we Open Data Studio or Tableau or Domo or Power BI like look at all these toys.
This is great.
We make this like what the hell is this? I don’t know what to do with this.
I can’t make a decision from this but it’s somebody decided they want to play with all toys.
And that’s fine for learning.
But this is not what goes into production your your dashboard has to be clear.
How do you make it clear, know what the different charts and functionsdo.This is from Dr.
Andrew Abella, I believe at Carnegie Mellon, who has a little framework of how do you think about the different charts that are available in the dashboard? And there’s four kinds, comparison distribution, composition and relationship.
And each one of these types of charts comes with a question.
So when you’re asking, again, you need these questions to make decisions.
What could we improve? That’s a decision that you want to make from a dashboard.
Use comparison type charts, bar charts, for example, column charts help you compare one thing to another.
So if you want to show what could we improve, use this type of chart.
If you want to find what’s unusual, what’s sticking out what doesn’t belong? Use the distribution to find out why did something happen one day that didn’t happen again.
Can we repeat that? Use a distribution histogram, plot a bell curve, whatever.
When you want to tell how are things related to each other and you have multiple days use use something like us.
scatter plot to to identify those relationships.
And when you want to know how much of any one thing is there, use a composition chart.
Be very careful with pie charts.
Never ever, ever put two pie charts next to each other.
That is the worst possible way to use a pie chart.
Ideally, you never use them at all.
But if you’re going to use them use one at a time.
If you have to compare two things in composition, use stacked bar charts instead.
So these are the rules for making your dashboards clear.
You want to be able to answer what can we improve use comparison charts, what’s unusual use distribution? How are things related? Use relationship? How much of any one thing is there use composition that will vastly simplify that huge palette of tools into I know what this tool does.
So and I know what question I’m trying to answer.
Second, your dashboard has to be clean has to be well labeled well documented.
Do not make a dashboard that looks good.
Even though design and everything people love doing this, you see this all the time, right? This is a fun example from like some hollywood movie, right? That’s what people think a dashboard should be.
Except that you have no idea what the hell is going on here.
There’s like lights and blinking and all sorts of stuff.
It looks cool.
Can you make a decision from this? No, I don’t even know what’s going on.
Right? I play a lot of video games.
And this is a really interesting example from World of Warcraft.
This is from competitive people fighting online with each other using like all magic spells and things.
This is the interface of one of the top players in competitive World of Warcraft.
Look how much stuff there isn’t on here.
He has I think this is a he just the bare minimum that they need to make a decision very, very quickly before they know that Warlock kills them behind this behind them.
They slimmed things down and keep only the information is valuable and you put it front and center.
You want to think about the same practice.
With your dashboards, you know, I mean, no one’s going to physically kill you, I hope in your decision making but you want to approach it from what is the minimum amount of information I need to make a decision.
Get out your Marie Kondo and clean house.
Right? what belongs on the dashboard, KPIs.
What’s the KPI? Number four, which you will get a bonus or fired.
If you’re not going to get fired, it’s not that important.
Put it in a report, put it on another page.
But your dashboard should be only the things that you and or your boss care about.
One of the exercises that I recommend you do is you actually write out all the different metrics and how they tie together we call this KPI mapping.
And then just look at the was the one thing that you are held accountable for and what is the one number below that? That’s probably all that needs to be on your dashboard and nothing else.
So what is the one number you are held accountable for? And then what are the numbers the numbers immediately below that in the KPI mapping chart.
Do the exercise.
It’s not difficult.
Here’s an example of a perfectly useful dashboard.
Look, where’s all the cool stuff? There isn’t any.
This is the only number that someone’s being held accountable for.
And this is an indicator saying no red arrow or green arrow.
This is a okay dashboard, because you can make a decision from this you can go Hmm, revenues down half a percent, I’m going to update my LinkedIn profile.
Third, dashboards need to be complete, they need to be able to answer the questions you ask of them.A lot of times, the more crap you put on dashboard, the harder it is to determine what you want What question you want to answer.
You’re not sure what’s in there.
When you look, for example, this one here.
This is one that my SEO made, and it’s got everything she needs to just to make a decision about what to do with our marketing and she has put everything along the right hand rail for her.
And then she says okay, these are the things that I need to make decisions to answers the questions that she has.
And this has changed like five or six times over the last year.
She needs more in context and more information for the dashboard.
So does your dashboard answer the questions after it, you can sometimes go too far and take too many things off of a dashboard.
That’s okay, you can put the beacons back once you realize they’re missing, but start cleaning up and making sure that your dash was asked the questions answered of them.
your dashboard should be concise.
This is the opposite.
Jay Baer alluded to this yesterday in his talk, and we’re going to apply a slightly different version here.
Again, for those folks who are not from America, we have a drugstore called CVS.
This is a purchase I made.
The top is the required information that I needed, about my purchase.
The bottom is all the other shit they put on the receipt.
Right? And waste is like half a tree to do it.
This is the Opposite of concise, this is what you don’t want to do, do not let your dashboard become the bottom half of this right? You want just the information that belongs on their.
Fifth, our dashboard should be cited well.
So that when somebody does say, hey, I need to know more about that, you know where it is.
One of the reasons I love Data Studio is that along the top there doesn’t work.
There’s a little data selector that tells you what data source it’s using.
But even that, you’ll notice on the bottom left, there’s some small text but one of the charts providing extra explanation about here’s why these numbers look the way they do.
If there’s ever any question that somebody has, like, where’d this number come from? You need to make sure that those citations are on your dashboards so people understand it.
And six, conclusive, your dashboard has to answer a question.
If it doesn’t answer the questions, if you can’t make decisions, it’s no good.
This is one that I use for my publication.
Making page on my website.
I look at this and I can go, okay.
What do I need to do next? If no one has said, Hey, come speak at our event, I know that I’m not doing a very good job marketing myself.
Right? That’s pretty basic.
And then I can look at all the different things that go into that, like, how many times is my name searched? If my name is in the top searches on my website, I’m a terrible marketer.
And then I can tell where my traffic’s coming from.
So I have to be able to look at this and draw conclusions, make decisions and decide what to do next.
your dashboard should be just this simple.
Look at what’s important, what goes into it, and the Can you make a decision from it.
That is the 60 framework for building effective dashboards.
Right? These are the ways that your dashboards go wrong.
If you break these principles, they don’t work.
If you do it, right.
Right? Everybody’s happy and actually wore this outfit today.
dashboards without decisions are decorations.
Thank you very much.want help solving your company’s data analytics and digital marketing problems? This is Trust insights.ai today and let us know how we can help you
Need help with your marketing data and analytics?
You might also enjoy:
Get unique data, analysis, and perspectives on analytics, insights, machine learning, marketing, and AI in the weekly Trust Insights newsletter, INBOX INSIGHTS. Subscribe now for free; new issues every Wednesday!
Want to learn more about data, analytics, and insights? Subscribe to In-Ear Insights, the Trust Insights podcast, with new 10-minute or less episodes every week.